$#%%@ Hits The Fan In GPL v VMWare

Everyone loves Linux and has the right to run, examine, modify and distribute the code“In October 2011, Conservancy received a GPL violation report on BusyBox for VMware’s ESXi products. Conservancy opened the matter in its usual, friendly, and non-confrontational way.

when Conservancy realized in late 2012 there might be a major Linux violation still present in VMware’s ESXi products, Conservancy representatives sought every industry contact we had for assistance, including those from trade associations, companies (both competitors and collaborators with VMware), and everyone else we could think of who might be able to help us proceed with friendly negotiations that would achieve compliance….
VMware’s last offer was a proposal for a settlement agreement that VMware would only provide if Christoph signed an NDA, and Christoph chose (quite reasonably) not to sign an NDA merely to look at settlement offer.”
under the terms of GPLv2, the licence permitting what would otherwise be violation of copyright. Why then does VMWare not provide its source code, a modification of the Linux source code, under the same terms as required by GPLv2?
” You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions:
a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change.
b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License.”

Creative Commons Attribution-ShareAlike 3.0 USA License see https://creativecommons.org/licenses/by-sa/3.0/us/#

Creative Commons Attribution-ShareAlike 3.0 USA License see https://creativecommons.org/licenses/by-sa/3.0/us/#


Years of negotiation have failed. WMWare would not even discuss a settlement without signature of an NDA (Non-Disclosure Agreement), for Pity’s sake. They weren’t negotiating seriously and now this goes to court where even more time, money and energy will be wasted in a case VMWare cannot win. They will be forced to cease distribution or rip out their code in some way. They may well have to pay for years of violation and profits in violation of others’ rights. If they did not want to distribute under GPLv2 they should not have used Linux. They got great value from using Linux. It’s the right way to do IT to use Linux so they should comply. They should have complied from Day One but did not. In spite of many of their loyal customers being placed in violation of GPLv2 by their action and pleading from those customers, VMWare stubbornly refused to comply. What were they thinking? They are offending their major suppliers and their major sources of revenue. That’s just not sustainable.

See VMWare Lawsuit FAQ.

See also, VMware sued in Hamburg, Germany court for failure to comply with the GPL on Linux

See also Statement in support of Software Freedom Conservancy and Christoph Hellwig, GPL enforcement lawsuit

See also, Christoff Hellwig defending the kernel against non-Free software as early as 2005…

Zachary Amsden, of VMware, wrote, back in the day“This is not an attempt to force a proprietary interface into the Linux kernel. This is an attempt to find a common interface that can be used by many hypervisors by isolating hypervisor specific idioms into a neutral layer. This new layer is just what is claims to be – a virtual machine interface, which allows hypervisor dependent code to be abstracted in a way that benefits both Linux and hypervisor development.”

I wonder what went wrong.

UPDATE Hellwig called VMware on this on LKML back in 2006:
“Until you stop violating our copyrights with the VMWare ESX support nothing is going to be supported. So could you please stop abusing the Linux code illegally in your project so I don’t have to sue you, or at least piss off and don’t expect us to support you in violating our copyrights. I know this isn’t you fault, but please get the VMware/EMC legal department to fix it up first.”

That’s from a thread started by VMware“I would like to propose an interface for linking a GPL kernel, specifically, Linux, against binary blobs. Stop. Before you condemn this as evil, let me explain in precise detail what a binary blob is.”

Ooh! They sunk my battleship! Not only does Hellwig have solid evidence of violation, he may well have proof of a guilty mind. In some courts that’s a hanging offence.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , , , . Bookmark the permalink.

637 Responses to $#%%@ Hits The Fan In GPL v VMWare

  1. DrLoser says:

    True.

    What? Then we agree. Let the record so state. Robert will no doubt catch up to our joint wisdom.

    Remember an OS (yes this can include a hypervisor)

    And here we disagree, with 100% honesty on both sides. Technically, I am not sure that the equivalence can be made. Legally, the issue has not even been approached.

    It’s a bit of a minefield here, oiaohm, my friend. Oh, and incidentally, don’t waste your time passing your “filter” on to the despicable Queef. It doesn’t seem particularly functional, does it?

    …is nothing without is core drivers. This is the big problem of ESXi vmkernel history is the lacks default drivers so cannot stand on its own two feet…

    A trivial observation where drivers are concerned, made even more trivial by the mention of “core” drivers. Whatever that might mean. I believe I covered this in Exhibits Two and Three, oiaohm.

    … without vmklinux.

    Which still has to be challenged in the German court, oiaohm.

    And, as I have pointed out: VMware could replace the driver interface with something else. (I am available. Six months of one person’s work, although I’m pretty sure they have it ready by now in the background.) And any lingering tainted bits of Linux in ESXi should be fairly straightforward to replace by FreeBSD or, well, anything at all, really.

    This is a pure case of independence.

    Whose? The FLOSS fanatics, who want to steal any bit of tainted code they can get their ignorant grubby little hands on?

    Or VMware’s?

    I’m guessing VMware here.

  2. oiaohm says:

    Robert
    Hmmm… and that software that runs those Linux drivers? Where’s the source code?

    True. Remember an OS (yes this can include a hypervisor) is nothing without is core drivers. This is the big problem of ESXi vmkernel history is the lacks default drivers so cannot stand on its own two feet without vmklinux.

    This is a pure case of independence.

    Something else MS Trolls will attempt to draw focus to the limited amounts of code that the Linux Developer has. German courts ruled that a Developer with limited amounts of code in a project can operate as proxy for all the other developers in the project. So only 1 Linux Developer has to spend the time defending the code. If other developer of the project disagree they are free to submit objections to the court. So far not a single developer has placed an objection to enforcement. Strangely enough not even VMWares own Linux Developers. What kinda suggests VMWare wants the court case.

  3. DrLoser says:

    Hmmm… and that software that runs those Linux drivers? Where’s the source code?

    Interestingly, Robert, it is more and more difficult to find this stuff in general. Unless you can apt-get it (or use the moral equivalent), you may need to hunt around.
    But, should you choose to ask for a VMware version of a Linux driver, I’m reasonably sure I could find it for you.

  4. DrLoser says:

    They get linked to some derivative of Linux then. FSC says VMware’s stuff tracks releases from kernel.org. There must be some derivations there. QED

    I notice that you use “QED” when you are particularly unsure of your position in the quicksand bog in question, Robert. No, it is not “QED.”

    Anybody who uses GPLv2 software and provides free access to the software and licenses the software according to the GPLv2 is presumably going to track the GPLv2 software via Github. It would be insane not to do so. Quod non est demonstrandum.

    ESXi is (unless otherwise proven) not linked in to any “derivative” of Linux whatsoever. Quod non est demonstrandum.

    It’s a hypervisor. It is not an OS.

  5. DrLoser says:

    This is the section that undoes point 1.

    Key point 2. Where in GPL does it define how big a section is. It does not.
    “identifiable sections” could be like 1 function when you go back to basic copyright law. Yes the second half is the viral bit. VMWare has without question used more than 1 functions worth of code.

    So what, Fifi?

    But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

    The “whole” has to be “a work based on the Program.”

    Since even German courts have not yet ruled on “complete corresponding source code,” let alone “fair usage,” I think it’s a bit of a stretch to imagine that they will see ESXi to be “a work based on the Program.”

    I mean, they might. It’s a legal lottery. But the odds are stacked against Nitwit and in favour of the Big Guns, which is to say the vastly better funded lawyers on VMware’s side.

    Sad, but true. On the other hand, I would personally be very pissed off if I’d built a commercial product of any value whatsoever, and some German twerp came along and demanded all of my source code simply because I’d used his implementation of the Radix Tree.

    In short, I do not personally see “a work based on the Program” here.

    The court may differ.

  6. DrLoser wrote, “they do not get linked to Linux, Robert”.

    They get linked to some derivative of Linux then. FSC says VMware’s stuff tracks releases from kernel.org. There must be some derivations there. QED

  7. DrLoser wrote, “VMware have duly complied with the GPLv2 license”.

    Hmmm… and that software that runs those Linux drivers? Where’s the source code?

  8. DrLoser says:

    Nvidia and AMD kernel mode drivers can prove they are not works based on Linux kernel since the binary blob part of driver works on Windows, OS X, Freebsd……

    Binary blob?

    BWAHAHAHAHAHAHAHA!

  9. DrLoser says:

    Separate works require at least Separate installers at least by all prior case examples.

    It Is So Engraved Upon Stone Tablets, Fifi. No, wait, it isn’t.

    Stop fantasising.

  10. DrLoser says:

    The authors of Linux didn’t write it so that VMware could modify it and distribute the modified work as non-Free software.

    And VMware have duly complied with the GPLv2 license, Robert.

    Now, if you want to talk about “taint,” then that would be a separate issue.

  11. DrLoser says:

    If someone in the real world modifies that Linux driver ABCco can’t throw that modification into those non-Free drivers because that’s not their code.

    But that isn’t what VMware are doing, in practice, Robert. You want a VMware modified Linux driver? Ask VMware or even the driver manufacturer.

    It’s GPLv2. You will be supplied with it.

  12. DrLoser says:

    Yes the Issue VMWare has done companies making routers have been toasted for as well. So why should VMWare get special treatment by being left alone???

    I have no idea at all, Fifi. Mostly because none of that makes any intelligible sense.

    Are we falling back to the previous German court decisions against companies that didn’t provide up-to-date source code, and a clearly recognisable license?

    Because, I’m afraid, that dog didn’t run back at the beginning of this conversation. And it sure as shit isn’t running any faster now.

  13. DrLoser says:

    Yes, they get linked to Linux. They are a part of Linux just as if they were embedded and not LKMs.

    No, they do not get linked to Linux, Robert. They get linked to ESXi. On the top of which, you are able to run Linux, with whichever drivers are necessary.

    They arenot LKMs. You can run an entire VMware hypervisor with no Linux kernel involved at all.

    You don’t seem to get this, for some reason, Robert. VMware is not an OS. It has no linked-in connection with the OSes above it (unless those OSes use VMware drivers to short-circuit things, and I believe they are all GPLv2 compliant).

    There is a very clear technical difference here.

    Now, will the courts agree? Let us wait and see.

  14. oiaohm says:

    They do have to reveal the source code if they contribute that LKM to Linux and if they distribute that LKM outside their organization but not otherwise.
    Not exactly but close Robert.

    They also don’t have to reveal the source code if they have separate work status. Nvidia binary blob is an example of this. But of course Nvidia code ships independent to the main Linux kernel with a wrapper code duel licensed.

    So no ship with Linux kernel or segment of and be provable as a separate work GPLv2 does not apply to the driver. At this point they can distribute that LKM outside their organisation and not release the source code. But this does become a problem for some ODM/OEM down the track when they ship products at times. The second half of the GPLv2 that undermines separate work status is evil. It what is catching allwinner and others

    Yes the LKM closed source and Linux kernel as shipped as two parts is legal. Glue them into a rom and o boy opps not legal any more. Yes you need to give uses a crappy driver that makes the hardware barely work at a min then ask the user to opt into the installation of the closed enhanced features so maintaining separate work status. Router makers have been getting this wrong and getting toasted 12 ways from Sunday.

    Yes the Issue VMWare has done companies making routers have been toasted for as well. So why should VMWare get special treatment by being left alone???

    Lot of closed source video/audio/… drivers under Android run in userspace on the right side of the Linus Clause. Userspace drivers is the other way todo closed source drivers on Linux and does not risk the nasty GPLv2 teeth.

    Yes the arguments that you cannot make closed source drivers for Linux is false. Legally you can. Using syscall exported functions and separate work status and everything is good.

    They are ways that VMWare could have done this properly and still had usage of the Linux drivers. It appears VMWare has been like the careless router makers. Has not really been careful enough to maintain the all important separate work status. Yes independent work is part of maintaining the separate work status.

    Independent work is simpler to prove and avoid court cases. Ok so the program works without the GPLvx code and you give user optin and opt out. So yes the two works are independent court case over unless the other part has been altering to match the GPLvx work. Please note VMWare in ESXi installers happens to install vmkLinux and not mention about installing it and not give an opt out option so its not an optional part.

    ESXi has been handling the GPL works badly. Complete failure to dot i and cross t on VMWare part.

    Mixing GPLv2/v3 code with closed source is not exactly simple process to get right. Its a very simple process to get it wrong.

  15. DrLoser wrote, “Are drivers an inherent part of the kernel? Not in any obvious way.”

    Yes, they get linked to Linux. They are a part of Linux just as if they were embedded and not LKMs. You can’t stick a LKM into some non-Free software any more than you could stick Linux, or a slice of its code in there.

    The only exception is that if some organization, say ABCco, creates some hardware and puts out a LKM for Linux and some drivers for other operating systems as makes sense in the world of x86, they could use the same guts in all those drivers regardless that they put that in a LKM for Linux. They do have to reveal the source code if they contribute that LKM to Linux and if they distribute that LKM outside their organization but not otherwise. They could still use the guts in a non-Free driver because it’s their code, but that’s not the case with VMware and all those drivers for various hardware found in the Linux kernel contributed by the world. If someone in the real world modifies that Linux driver ABCco can’t throw that modification into those non-Free drivers because that’s not their code. It’s the world’s code (to the extent that the world is authourized to read, run, modify and distribute it under GPLv2), not VMware’s.

  16. DrLoser wrote, “if you are going to distribute “the whole,” you need to carefully follow the GPLv2 licence on every single part of the distribution that was supplied to you under the GPLv2 licence.”

    … and every derived work of Linux just as you can’t write some software linking to a bunch of GPLv2 stuff and distribute it under a different licence. The idea is to keep FLOSS Free Software. The authours of Linux didn’t write it so that VMware could modify it and distribute the modified work as non-Free software.

  17. oiaohm says:

    Robert I forgot to point out 4 key things about this section of GPLv2 that apply to this case.
    These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works.
    This is the clause VMWare has to win on. Point 1.
    “separate works” This is VMWare main arguement. The they were in fact separate works on the same media. If VMWare proves this they win. There is no court ruling anywhere to make distributing on the same media in a unified installer separate works.

    Separate works require at least Separate installers at least by all prior case examples. Of course VMWare might be the first that you don’t have to use separate installers.

    This ruling could have some effects on Linux based Livecds if VMWare loses. Yes it depend how they lose this.

    But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
    This is the section that undoes point 1.

    Key point 2. Where in GPL does it define how big a section is. It does not.
    “identifiable sections” could be like 1 function when you go back to basic copyright law. Yes the second half is the viral bit. VMWare has without question used more than 1 functions worth of code. Arguing that something is driver or library makes no different to the GPL license terms. So VMWare using identifiable sections is 100 percent yes.

    Key point 3 “work based on the Program” is a path that VMWare can attempt to argue that the esxi kernel is not a work based on the Linux code. If there are as claimed direct examples of VMKernel altering when Linux kernel alters this arguement is screwed. Nvidia and AMD kernel mode drivers can prove they are not works based on Linux kernel since the binary blob part of driver works on Windows, OS X, Freebsd…… This same loop hole is using with driver/hardware firmware. To win point 3 you need provable independence. VMWare has provided no evidence of this at this stage but the case FAQ contains notes they have have evidence disproving this clause.

    Now if it is a library or driver this is not defined in the GPL license. Please note the license does not define how big a section is.

    Point 4. The reason Linux kernel GPLv2 is not viral to everything above it is the Linus clause.
    NOTE! This copyright does *not* cover user programs that use kernel
    services by normal system calls – this is merely considered normal use
    of the kernel, and does *not* fall under the heading of “derived work”.
    Also note that the GPL below is copyrighted by the Free Software
    Foundation, but the instance of code that it refers to (the linux
    kernel) is copyrighted by me and others who actually wrote it.

    Linus Torvalds
    This clause VMWare failed to maintain. Most libraries that contain a GPL license are LGPL. LGPL contains a clause again limiting how far GPL reaches.

    GPLv2 and GPLv3 are highly Viral with very little in the way of exceptions to avoid.

    The reality is the VMWare case is completely down to what you class a “separate works” as. Also down to what you must do to prove that a work is a separate work.

    Interesting point is the half second clause of GPLv2 is fairly clear that covers anything that you cannot prove independence.

    VMWare case is not complex all you have to do is read the GPLv2 license and stick to what it says and not go off on tangents.

    Due to how simple GPLv2 is to read I have never really understood why GNU has a GPL FAQ. Something else is the GPL FAQ in prior cases defendants have attempt to use it as defence only to be told it has no legal standing. The issue you agree to the License not the FAQ. Same applies to Microsoft guides to Licensing what ever is in the EULA overrides those Guides. You want to know if you are legal or not read the license anything else is a guide line nothing more.

  18. LinuxGentlesir says:

    oiaohm,

    Yup, so my opinion is these GPL lawsuits were weaker and the outcome was not particularly positive for the defendants:

    http://www.softwarefreedom.org/news/2007/sep/20/busybox/
    http://news.zdnet.co.uk/software/0,1000000121,39290971,00.htm
    http://www.softwarefreedom.org/news/2007/nov/20/busybox/

    If it is found that VMWare truly bundled a proprietary driver with the kernel, they have dun’ fucked up pretty badly. As you mention, pretty much the entire purpose of the GPL to to prevent proprietary software from linking to it. That’s like its key feature. If Linux author’s intended to permit such things, they would have used the LGPL or another license written for that purpose. But the key thing is: they didn’t. It’s clear that the intention of the copyright owners is NOT to allow the kind of thing VMWare did, and arguing that some technical measure gets you around a copyright license tends to not go well in court.

  19. DrLoser says:

    Perhaps even more interestingly, you could make a compelling case for several Linux libraries to be more “integral” to the Linux kernel than any driver at all.
    For example, you’re going to find it pretty damned difficult to build anything on top of the kernel unless you use glibc. Forks permitting.
    Unless you’re targeting a headless server, you’ve got a shed-load of X libraries you need to have.
    A Linux driver, OTOH? Meh. Swap one out, swap one in. No biggie.

  20. DrLoser says:

    And just to be nit-picky … Robert is far better at being nit-picky than I, but I hope to learn from the Master …

    the distribution of the whole must be on the terms of this License…

    That doesn’t really mean anything at all in terms of giving 100% of the code away for free, does it?

    All it means, on the surface of it, is that if you are going to distribute “the whole,” you need to carefully follow the GPLv2 licence on every single part of the distribution that was supplied to you under the GPLv2 licence.

    Which is where previous cases in front of the German courts have fallen down.

    But I’m not convinced that it applies to ESXi.

  21. DrLoser says:

    Actually, and to be completely fair to oiaohm’s unusually cogent analysis here, I should address myself to his GPLv2 cite:

    But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License…

    It’s rather difficult to see a modified Linux device driver as part of “a whole which is a work based on the Program,” isn’t it?

    Particularly since, although it would obviously be inconvenient, you could either swap something else in or else rewrite the thing from scratch yourself, plausibly with the help of pissed-off hardware suppliers?

    This one won’t wash, I’m afraid. The Radix Tree might. SCSI drivers might. But not this one, provided that the modified source of the drivers is freely available.

  22. DrLoser says:

    This complete VMWare case is really over this section of the GPLv2 license.

    Indeed true, Faithful Parrot.

    I am indebted to you for repeating my point verbatim.

  23. DrLoser says:

    Now, should you choose to state your opinion that Linux device drivers are per se inalienable because of GPLv2, you might have a point.

    It would be a point that risks seriously pissing off all those device manufacturers who, as you remember, were reluctant to spend any time writing a device driver specifically for Linux, only about six or seven years ago. But it would still be a point.

    In fact, it was a point I tried to discuss as Exhibit Three. I didn’t really hear much back then, except (I will admit) from oiaohm.

  24. DrLoser says:

    VMware is shipping drivers. Drivers are a component of the kernel. They are not libraries.

    Now, this is interesting. (And thank you both for returning to “interesting.”)

    Are drivers an inherent part of the kernel? Not in any obvious way.

    Let’s say, for argument’s sake, that I wrote a driver for Gizmo X specifically for FreeBSD. I actually have no idea how much of that effort would consist of interfacing it to the FreeBSD kernel, but let’s stipulate 5%.

    Now, let’s say I port that driver to the Linux kernel. Another 5%. And let’s say I port that driver to the ESXi “kernel” via Exhibit Two (which we all agree is a contested point). Another 5%.

    I don’t think that 5% of effort makes anything “integral,” as such.

  25. DrLoser says:

    People reading the GPL FAQ get the idea when it says stuff about not having to ship compliers and so that there is some exception the GPL license allowing.

    This, on the other hand, is questionable. (Although oiaohm makes a good fist of defending it.)

    It’s basically collateral damage. If you need to explain a legal obligation via an FAQ, then you’re probably not doing it right.

  26. DrLoser says:

    Robert technically that makes no difference. GPL license without exceptions is viral from Library just as much as it Viral from driver.

    That, by the way, is technically absolutely true. oiaohm is perfectly correct.

    Whether or not this stands up in court is another question.

  27. oiaohm says:

    VMware is shipping drivers. Drivers are a component of the kernel. They are not libraries.
    Robert technically that makes no difference. GPL license without exceptions is viral from Library just as much as it Viral from driver. People reading the GPL FAQ get the idea when it says stuff about not having to ship compliers and so that there is some exception the GPL license allowing. Get the GPL license attempt to find it and there is no such thing as an exception for this if you happen to ship the GPL part and your closed part as 1 bit.

    Why you don’t have to ship the source code of compliers and OS running a GPL work is the derivative work clause.

    However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable.
    This is from GPLv2 itself. Does not apply in VMWare case due to the unless and the reason VMWare ships vmkLinux and vmkernel as 1 thing. This is why I said about independent work status. There is only one clauses in GPL that allow you not to have to ship your source code or follow GPL and without question don’t allow you to ship the GPL code with your closed source code. Yes ship the GPL part as an independent work add on and everything good but this is not what VMWare has done.

    accompanies the executable This is the most dangerous part this void the clause as soon as the executable is shipped on the same media. The GPL work did not even have to be library or a driver before the do not have to give source code clause is no longer exists.

    So the only section of GPL license VMWare can challenge is the meaning derivative work under copyright law.

    These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

    This complete VMWare case is really over this section of the GPLv2 license.

    Note the words “independent and separate works” I have been using independent work repeatability for a reason. For something like VMWare case this is the only section of GPLv2 that applies. If VMWare has to win or lose on this section.

    Yes there were methods VMWare could have used to do what they wanted properly. Yes the other clause would have come into play if you installed ESXi then had to install vmkLinux afterwards but that would required ESXi having default drivers so it could operate without the vmkLinux parts.

    Lets say VMWare wins. This would end the arguement about usage of libredwg and other things open source and also allow Teigha and other closed source libraries to be used in open source projects.

    VMWare loses companies like Allwinner and other makers breaching GPL get a clear warning notice. Also you have remember the make of Xen and other competitors to VMWare product as also part of the Linux foundation. There is a lot on going disputes over how VMWare interacts with Open source. If this case settles it everything will operate better.

    Win or Lose this case is a win for different groups of FOSS developers. So trying to say we should be against this case is wrong.

  28. oldfart says:

    “VMware is shipping drivers. Drivers are a component of the kernel. They are not libraries.”

    You are entitled to your opinion, but it is the courts opinion that counts not yours. the german court has yet to try the case. So lets just sit back and see which way this goes.

  29. DrLoser wrote, “unless that component itself accompanies the executable”.

    VMware is shipping drivers. Drivers are a component of the kernel. They are not libraries.

  30. DrLoser says:

    Obviously talking about simple programming concepts is rather too abstruse for this blog.

    Full as it is of people who routinely examine and improve and redistribute FLOSS code. Which, you may have noticed, is overwhelmingly written in C or C++.

    So, back to the topic in the OP. (I like to revert to the topic in question. Fifi likes to gish-gallop. Robert loves to reminisce about the 1980s. We all have our little foibles.)

    I think it’s fair to say that the intellectual, moral, and possibly even the legal underpinnings of this case are based around the Gnu Public License, version 2.0. So let’s have a quick squiz, via the FAQ, shall we?

    If the libraries that you link with fall within the following exception in the GPL …

    However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable

    … then you don’t have to do anything special to use them; the requirement to distribute source code for the whole program does not include those libraries, even if you distribute a linked executable containing them. Thus, if the libraries you need come with major parts of a proprietary operating system, the GPL says people can link your program with them without any conditions.

    I suspect that the VMware lawyers are going to make hay with this one.

    What is a hypervisor, if it is not “a compiler, kernel, and so on of the operating system on which the executable runs?”

    If you’re going to issue a license which clearly privileges the compiler, the kernel, and the operating system, then how on earth does this “exception” not apply to a hypervisor?

  31. DrLoser says:

    Does anybody at all have a cogent comment on that?

    No?

    I suppose off-by-ones, rubbish logging, copypasta and complete lack of encapsulation are out of the question, then.

  32. DrLoser says:

    Incidentaly, oiaohm’s observations on the “reluctant necessity” for Dwarf-based C++ applications to use C++11x are, as always, totally misinformed miserable crap.

    I would have mentioned that earlier, but I figured we were all better off knowing about C++ exceptions in the first place. oiaohm’s drivel would, at best, have been a distraction.

  33. DrLoser says:

    DrLoser is on my ignore list.

    I was hoping for an interview with the Queen, Fifi, but I guess this will do me fine.

  34. DrLoser says:

    Strictly speaking, you can override terminate().

    But then we’re back to the party games of oiaohm’s fun little cite on “exception handling in C.”

    Don’t go there.

  35. DrLoser says:

    Now, back to exception-handling. In the teeth of all opposition, I am determined to educate you all in this simple domain. Who knows, you might even be able to “improve” FLOSS software once you understand this!

    The key to dealing with exceptions is to understand what they are intended to do. Let’s take a sort of generic approach. Let’s say “something” goes wrong, somewhere up the stack, we don’t necessarily know where. And that “something” could be a programmatic exception (“I am not able to deal with this particular state”) or an OS exception (“Out of memory”). It doesn’t really matter.

    Now, in all cases, there is an “inside” and an “outside.” I believe I tried to hint to oiaohm what goes on, by referring to the “outside.” oiaohm was sadly ill-equipped to think about the “outside.” Which is odd, because that appears to be, basically, where he lives.

    Here’s what happens.

    1. You throw an exception from the “inside.” Which can be practically anywhere. Which, incidentally, is why exception guarantees do not work in C++. (Another free gift.)
    2. Something happens.
    3. The exception is caught by the “outside.” But which “outside”?

    In the current state of affairs, vis-à-vis C++, step 2 is important in determining which “outside” catches the exception.

    Because what happens when an exception is thrown, you see, is that the stack gets unwound. And when the stack gets unwound, every destructor on that stack is called. And if a destructor allows a second exception to escape …

    … You’re up shit creek without a paddle.

    At that point, “outside” is no longer under your program’s control. You can’t save the state of, say, your CAD-CAM skyscraper design. You can’t warn the user. You can’t do anything.

    Because in these circumstances, “outside” is a call to terminate().

    Short, brutal, nasty, and without appeal of any kind. So don’t do it.

    Now, that’s the current state of affairs. It doesn’t apply to dynamic languages like Perl and Python. And it doesn’t apply to GC languages like Java and Scala and C#. But it does apply to C++.

    I could go a little further into the theory and talk about rollback semantics and possibly even functional languages and invariants. But I won’t. If anybody is interested in where C++ is going with this issue, I recommend (as previously advertised) Herb Sutter.

    Don’t thank me, Ma’am. It’s enough for me to know that the streets out there are safe from exceptions that somehow escape a C++ destructor.

  36. DrLoser says:

    I’ll just let that sink in, btw. GRASS is a FLOSS system dealing with GIS mapping.

    And it doesn’t even bother calculating the area of a polygon.

    Spiffy!

  37. DrLoser says:

    When a provider tells you how wonderful his software is but refuses to tell you just exactly how it works, bells go off…

    That’s a very good filter, Robert.

    Have you applied it to recent FLOSS candidates here like Musescore? Or QGis? Or GRASS?

    Being the leading substitute for ArcGIS, however, GRASS does not handle vector files of size close to 1 GB very well nor data table of size larger than 4 MB and does not have a function to calculate area (which is surprising).

    Admittedly, that last cite isn’t actually from GRASS themselves. Which is why it makes a pathetic attempt to be honest about the limitations.

    You don’t apply that filter to FLOSS at all, do you, Robert? In fact, you don’t even avail yourself of the cursory ability to examine the code — even though it is freely available.

    Not really a very useful filter at all, as far as I can see.

  38. DrLoser says:

    DrLoser quoth, “A technical workaround designed primarily to avoid getting a legitimate copyright license is illegal, flat out.”

    I hate to go all Monty Python on you, Robert, but that wasn’t one of mine. It was one of LinuxGentlesir’s.

    Good to see you criticising the opinions of one of your own, though.

  39. DrLoser quoth, “A technical workaround designed primarily to avoid getting a legitimate copyright license is illegal, flat out.”

    While that case may not have worked out, I believe both USA and Canada have passed legislation banning any technological measure to defeat copyright protection. While these laws are intended to preserve stuff like encryption of CDs, the same thing should apply when software is written to violate copyright by some obscure organization hiding the violation. I have a “filter” I use to judge software. When a provider tells you how wonderful his software is but refuses to tell you just exactly how it works, bells go off… It’s like Mom catching me chewing in close proximity to the cookie jar… VMware creating software that needs drivers from Linux to be useful but denying it’s violating the GPL comes to mind. “I’m only violating the GPL a little bit… Awww, come on!”. Yes, right. [SARCASM]

  40. DrLoser wrote, “I’m assuming that Robert is having more fun with his new tractor right now”.

    I spent a glorious afternoon at the rifle-range hammering targets with an 8mm Mauser and collecting hundreds of brass cartridges litterbugs had left behind. Count was ~600 9X19mm, dozens of .45ACP and a similar amount of .40S&W. Oh, yes, a few hundred .223 Remington brass. Not only did I have fun, we probably paid for our range memberships and gasoline used so far… I have a feeler gauge for the tractor and have only to finish checking the brakes and clutch before taking it out for a run. I expect to be ploughing Wednesday. Weather should be fine tomorrow afternoon to finish the work. I still have some assembly to finish on the roto-tilling attachment but that too is looking fairly straightforward. The ground is partially thawed so tilling will help it warm in the coming weeks of sunshine and mild temperatures. It’s about time to install my summer radials too. I’m staying busy.

  41. oiaohm says:

    Just so everyone reading this post knows due to DrLoser constant determination to use FIFI instead of Oiaohm to refer to me there will be no answers coming to the items DrLoser has raised if you wish for me to answer them please do your own posts. DrLoser is on my ignore list.

  42. DrLoser says:

    A half-arsed broken approach to programming that nobody here can even be bothered to download, let alone examine, let alone criticise, let alone spend their invaluable $0 per hour fixing, reporting, integrating and testing and so on?

    Let’s just say that MasterCard are not about to phone you Bozos any time soon for their latest set of commercials.

    Which is rather a shame, really. Robert would look pretty spiffy in a bikini.

  43. DrLoser says:

    Exceptions, whether escaping from a destructor, or simply being the subject of the Strong Exception Guarantee. I wonder what those could possibly mean?

  44. DrLoser says:

    Off-by-one. I wonder what that could possibly mean?

  45. DrLoser says:

    Ah well, yet another nauseous know-nothing brat.

    I was hoping for intelligent conversation on the issues. Perhaps a tete-a-tete with Robert, even.

    I’m assuming that Robert is having more fun with his new tractor right now. And given the alternatives of listening to cretinous sheeple like Fifi, LinuxGoodsir, Doug-nut, and so on …

    Digging through Good Honest Dirt has never sounded so appealing. More speed to your diesel, Robert!

  46. DrLoser says:

    A technical workaround designed primarily to avoid getting a legitimate copyright license is illegal, flat out. See American Broadcasting Cos. v. Aereo, Inc.

    Reversed and remanded, incidentally, Puppykins. For what it’s worth, Aereo appear to have won on that one.

  47. DrLoser says:

    A technical workaround designed primarily to avoid getting a legitimate copyright license is illegal, flat out. See American Broadcasting Cos. v. Aereo, Inc.

    And some dismal little fool claiming a “technical workaround” is deeply uninteresting, flat out.

    Ya wanna tackle Exhibit One, Exhibit Two, Exhibit Three, or even Propositions One and Two, Puppykins?

    It’s an intellectual step beyond drinking the Lessig Kool-Aid, I admit. And you’re pretty much incapable of dealing with legal reality.

    But feel free, Puppykins.

    Oh, and whilst your at it? That nominated MuseScore source file, please. Shouldn’t be too onerous. Just download it and pick one at random.

    Assuming that you actually care about software quality in FLOSS, of course.

    Which I rather doubt you do.

  48. LinuxGentlesir says:

    A technical workaround designed primarily to avoid getting a legitimate copyright license is illegal, flat out. See American Broadcasting Cos. v. Aereo, Inc.

    It doesn’t matter what Rube Goldbergesque scheme you come up with, the fact that you are doing it simply because you want to avoid following the terms of a copyright license makes it illegal.

  49. DrLoser says:

    Now, about those teeny tiny little code-review issues.

    1. Allowing an exception to escape from the destructor.
    2. Lousy logging.
    3. Copypasta.
    4. Complete inability to isolate properties or calculations via bog-standard isolation.
    5. Off-by-one statements.

    Any tiny little brainless ideas about these grotesque failures in what is little more than a FLOSS “C++” (failed, miserably) wrapper around a popular, yet questionable, FLOSS C library?

    I’m guessing that not a single one of you has a clue, do you?

    Certainly not Puppykins. Sorry, sweetie, but all you’ve got to offer is total ignorance and an inability to recognise that you really don’t belong here.

    Try again, Puppykins.

  50. DrLoser says:

    Anyways I’d love to watch Dr Loser get hit with some criminal copyright infringement charges.

    Sadly, Queefer, that ain’t gonna happen.

  51. DrLoser says:

    It doesn’t matter what “technical workarounds” you invent.

    Indeed not, Puppykins.

    Do you have a comment on “Exhibit One — The Radix Tree?”
    Or “Exhibit Two — the Linux device driver API?”
    Or “Exhibit Three — Linux device driver source, and the GPLv2?”
    Or even the yet-to-be-examined SCSI issue?

    You don’t, do you?

    Prove me wrong, Puppykins.

  52. LinuxGentlesir says:

    The author’s wishes are that if you make code that interacts with their code, you release your code under the GPL.

    It doesn’t matter what “technical workarounds” you invent. Technical workarounds designed to circumvent the law don’t usually go well with judges. Your intention is to get around a copyright license, that’s all that matters. Anyways I’d love to watch Dr Loser get hit with some criminal copyright infringement charges.

  53. Dr Loser says:

    Yes how to violate 60 000+ patents port port Linux drivers without GPL.

    Hee hee hee.

  54. Dr Loser says:

    Yet again, a non-starter, Fifi.

    But then again, the only real question when dealing with your babble is, at which point do you demonstrate your pitiable ignorance on the subject?

    Have you ever been right, even once?

    I’m searching my memory, over eight years or so, Fifi.

    No, you have not. Not once.

    Gold star for persistence, however.

  55. Dr Loser says:

    Go ahead and do it. Something highly nasty is that the patent grants to Linux kernel don’t come to you if you use a non GPL work.

    Have you adverted Mr Pogson of these “patent grants,” Fifi?

    I think you should.

    We all know how very enthusiastic Mr Pogson is about “patent grants.”

  56. Dr Loser says:

    Just to be horible most Linux drivers avoid ioctl control instead using /proc or module command line arguments. Using proc means file system io operations expected to a point.

    That’s because most authors of Linux drivers don’t give a toss about security, and leave it up to the kernel. Not a very good idea, IMHO. But in any case, completely irrelevant to a theoretical model of non-Linux driver APIs for device drivers. Because a non-Linux model would pretty much by definition avoid the security flaws and general nonsense involved with /proc.

    Remember, Fifi — this is a hypervisor we are talking about here. No need whatsoever to access /proc. Indeed, no ability to do so, either.

    Yet another non-starter, Fifi. You’ve got nothing at all, have you?

    But perhaps you do. In which case, I will abrogate my rights to complain about you forfeiting your puny little attempts to reply to me.

    Go ahead and speak your puny little mind, Fifi. I prefer ignorant gabbling stupidity above putrid feeble passive aggression.

    Sadly, in this case, it’s still a non-starter, isn’t it?

  57. Dr Loser says:

    Other than the subsystem of Linux provides generic functionality. So drivers in Linux can be majority incomplete. Like 90 percent generic functions coming from the subsystem.

    Squinting at this gibberish, I think I can get a vague understanding of what you are trying to say.

    The Linux kernel, being monolithic, includes a very large number of “generic drivers.” As indeed it should. It’s a successful business model.

    Nothing to stop them being adapted to a different API, though, is there? (Assuming GPLv2 compliance, via publication and so on. And assuming that my Proposition Two is accepted via the German Courts. But hey, why bother torturing your tiny little brain with these questions?)

    Again, Fifi, a non-starter.

  58. Dr Loser says:

    Don’t protect GPL future hardware vendors might not give patent grants at all. This would be perfect for Microsoft restricting a world to closed source binary blobs from vendors only.

    I am not proposing (in this theoretical model where VMware has to look elsewhere, because some Nitwit has been harbouring a hair up his ass since 2006) any such thing. A simple alternative API, composed of a “fabric” API (in telecoms terms) and a “data” API. That is all.

    A non-starter, Fifi.

  59. Dr Loser says:

    Forgetting of course the Linux internal kernel API changes when ever the Linux Developers fell like changing it. The changes can be to fix security faults.

    When is the last time the Linux device driver API changed? Refresh my memory, little one.

    A non-starter again, Fifi.

  60. Dr Loser says:

    Go ahead and do it. Something highly nasty is that the patent grants to Linux kernel don’t come to you if you use a non GPL work. Porting the Linux drivers is a great way to be sued back into the stone age by hardware companies if your end result is anything other than GPL.

    Unfortunately the hardware companies are the ones likely to benefit from an alternative interface, should Nitwit vs VMware for some unlikely reason deny them access via the Linux API.

    A non-starter, Fifi.

  61. oiaohm says:

    DrLoser basically it goes like this.

    You want to write drivers. You need at least some permission from the hardware maker that they will not sue you over patent violation. Part of getting that permission is agreeing to the hardware makers terms. In the case of a lot of hardware its GPLv2 or GPL work so you have valid patent license.

    You say to hardware making I wish to make a closed source driver. The hardware maker things hmm that closed source driver might make me look bad. No grant.

    Don’t protect GPL future hardware vendors might not give patent grants at all. This would be perfect for Microsoft restricting a world to closed source binary blobs from vendors only.

  62. oiaohm says:

    Complete plan will fall part.
    1. First, choose a Linux device driver. Any old driver will do. They’re all basically written in C and work towards the same interface.
    Not exactly C.
    2. Remember, kiddies, we are not allowed to use the “official GPLv2″ Linux driver API! (Well, we might be. The court has yet to decide on Exhibit Two. But let’s assume we’re not.)
    Forgetting of course the Linux internal kernel API changes when ever the Linux Developers fell like changing it. The changes can be to fix security faults.
    3. Let’s focus on what a device driver does at the API level. Why, it does the following things!
    3a. It provides the ability to receive input, in what I will for the sake of argument call an fstream. (It’s a little more various, but that would do for now.)
    3b. It provides the ability to send output, in what … repetition would be tiresome.

    Other than the subsystem of Linux provides generic functionality. So drivers in Linux can be majority incomplete. Like 90 percent generic functions coming from the subsystem. This is why VMWare made vmkLinux in the first place. This is why porting a Linux driver to another platform comes a bitch very quickly. If you don’t take the subsystem you are going to find yourself short of many critical things.

    4. It provides a control interface. Again, to generalise, this is basically the time-honoured *nix IOCtl.
    Just to be horible most Linux drivers avoid ioctl control instead using /proc or module command line arguments. Using proc means file system io operations expected to a point.

    And I could construct an architecture to do the same for every single last Linux driver in existence in a matter of, say, six months.
    Go ahead and do it. Something highly nasty is that the patent grants to Linux kernel don’t come to you if you use a non GPL work. Porting the Linux drivers is a great way to be sued back into the stone age by hardware companies if your end result is anything other than GPL.

    Yes how to violate 60 000+ patents port port Linux drivers without GPL.

    I am serous when I say you risk badly offending hardware companies. Hardware companies are willing to license a lot of Patents to GPL works but not to any other Open Source license or closed source work they did not produce. The less than 500 Microsoft claims Linux violates is nothing to the trouble you are proposing.

    VMWare had no choice on License or VMKLinux if they choose anything else they would not be looking down copyright volition right now but patent volition.

  63. DrLoser says:

    Yes indeed. There still remains the possibility that VMWare’s creator and his engineers were collectively lazy enough to just lift verbatim pieces of the… Linux…kernel…

    Actually, I am not discounting that possibility, oldfart. In fact, I’m assuming that it’s about the only leg that Nitwit vs VMware has to stand upon. But I’m quite prepared to admit that a large US company is ignorant enough to try it on.

    Let’s see, we’re up to near 600 posts on this particular disquisition.

    Probably about 100 of them are mine. I have tried, in a neutral sort of way, to analyse the issues. I have categorised the issues according to the Software Conservancy plaint. I have even divided the two possible consequences as Proposition One and Proposition Two.

    I have exercised what little knowledge I have of either software or the law to the fullest extent I can manage.

    And all I hear back from Robert, Fifi, et al, is basically the empty cackles of a monotonous parrot.

    Boy, is this fun!

  64. DrLoser says:

    Don’t protect GPL hardware vendors might lose trust in GPL for protecting their effort. Protect GPL risk offending companies like VMWare.

    What on earth are you babbling on about, Fifi?

    Robert, can you parse this drivel? What’s with the “Protect GPL risk offending companies like VMware” nonsense?

    In fact, Robert, do you believe a single word that Fifi says?

    And, if so, on what basis?

  65. DrLoser says:

    Robert the divide starts happening back in 1992-1993 when BSD had more market share than Linux so the divide is not market share. Yes early Linux was attracting more vendor support than BSD counter to market share at the time.

    And is completely irrelevant as always, Fifi. You really are an ignorant prat, aren’t you?

    I’m going to give you my Magic Recipe for converting a Linux Device Driver from the Linux Device Driver API to a FreeBSD Device Driver API.

    Fer Free! No Charge Attached! Your Cod-Piece Change is Intact! Also, You Are Free To Redistribute Blah Blah Blah.

    1. First, choose a Linux device driver. Any old driver will do. They’re all basically written in C and work towards the same interface.
    2. Remember, kiddies, we are not allowed to use the “official GPLv2” Linux driver API! (Well, we might be. The court has yet to decide on Exhibit Two. But let’s assume we’re not.)
    3. Let’s focus on what a device driver does at the API level. Why, it does the following things!
    3a. It provides the ability to receive input, in what I will for the sake of argument call an fstream. (It’s a little more various, but that would do for now.)
    3b. It provides the ability to send output, in what … repetition would be tiresome.
    4. It provides a control interface. Again, to generalise, this is basically the time-honoured *nix IOCtl.

    I could build all of that on top of a representative Linux device driver in, what, four weeks? It’s not difficult. It’s just a pointless expense.

    And I could construct an architecture to do the same for every single last Linux driver in existence in a matter of, say, six months.

    Special Sauce? There ain’t no special sauce here, folks.

  66. oiaohm says:

    That may be a factor but a larger factor is that the installed/user base of FreeBSD is less than GNU/Linux so more effort by all concerned is put into drivers for Linux.
    Robert the divide starts happening back in 1992-1993 when BSD had more market share than Linux so the divide is not market share. Yes early Linux was attracting more vendor support than BSD counter to market share at the time.

    This is the problem the evidence clearly backs the idea that a large percentage hardware makers like GPLv2 or Closed source solutions and does not like anything else. Now if the hardware vendors lose trust in GPLv2 this is going to be a problem.

    Linux kernel developers have a large amount of trouble even getting some companies to give up specs let alone drivers.

    Understanding the reality of this means that a GPL violation against Linux Kernel does have increased odds of ending up in court.

    Don’t protect GPL hardware vendors might lose trust in GPL for protecting their effort. Protect GPL risk offending companies like VMWare. Companies like VMWare are replaceable in time. Loss of hardware vendor support forced to reverse is a far worse problem.

    Now you have not heard a single Linux Kernel developer or Hardware vendor say that VMWare should be left alone. Lot say what I am saying this need to go to court and be settled one way or the other.

    No matter how the case is ends things will change for ever.

  67. oiaohm wrote, “FreeBSD lack of drivers come from it not being the License type Hardware vendors like.”

    That may be a factor but a larger factor is that the installed/user base of FreeBSD is less than GNU/Linux so more effort by all concerned is put into drivers for Linux.

  68. oiaohm says:

    “Again failing to ask yourself question. Why did the hardware vendors make Linux GPLv2 drivers instead of BSD ones.”
    And DrLoser answer
    Because they sell, you idiot.
    This is wrong in fact. There are such things as Linux BSD and MIT licensed drivers in the main kernel tree.

    It is the driver authors who decide the license. Most of Linux drivers licenses in fact come from hardware vendors directly. Yes they have chosen GPL license over the other license option.

    http://www.tldp.org/LDP/lkmpg/2.6/html/x279.html
    * “GPL” [GNU Public License v2 or later]
    * “GPL v2” [GNU Public License v2]
    * “GPL and additional rights” [GNU Public License v2 rights and more]
    * “Dual BSD/GPL” [GNU Public License v2
    * or BSD license choice]
    * “Dual MIT/GPL” [GNU Public License v2
    * or MIT license choice]
    * “Dual MPL/GPL” [GNU Public License v2
    * or Mozilla license choice]

    Yes 6 open source license options. Most hardware makers choose GPLv2.

    I’m proposing the thesis, and it is merely a thesis, that VMware could, if they wanted to, build the equivalent of vmxlinux from a FreeBSD basis and avoid an awful lot of legal hassle. And I have pointed out that there is no present reason for them to do so, unless the legal hassle becomes intolerable.
    Of course this brushes over the basic fact a lot of vendors don’t release BSD licensed drivers. Again using those GPLv2 licensed drivers would require VMWare keeping formal separation correct. FreeBSD lack of drivers come from it not being the License type Hardware vendors like.

  69. LinuxGentlesir says:

    Yeah well that’s the point of the court case!

  70. oldfart says:

    “The SFC’s requests for donations to initiate this lawsuit just got completely funded. VMWare is going to have to answer for GPL violations in a court of law!”

    First the alleged violations need to be proven sir. This is not a kangaroo court. Or does you need for personal sovereignty demand that VMWare just be judged guilty because you want it to be?

  71. LinuxGentlesir says:

    The SFC’s requests for donations to initiate this lawsuit just got completely funded. VMWare is going to have to answer for GPL violations in a court of law!

  72. oldfart says:

    “Pretty conclusive evidence that, I would say.”

    Yes indeed. There still remains the possibility that VMWare’s creator and his engineers were collectively lazy enough to just lift verbatim pieces of the… Linux…kernel…

    HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA…

    (gasp)

    I think not 😉

  73. DrLoser says:

    Any thoughts on the GPLv2 SCSI violations by VMware, by the way?

    No, I thought not.

    You people just don’t care, do you? I could replace every single one of you with a demented parrot that does nothing else but squawk “GPL violation! GPL violation! GPL violation!” all day long

    … and, frankly, the demented parrot would be a more useful amicus curiae than the lot of you put together.

  74. DrLoser says:

    Oh well: I hate to spoil Fifi’s fun.

    In short, a user world is not intended as a general-purpose mechanism to run arbitrary applications but provides only enough of a framework for processes that need to run in the hypervisor environment.

    Pretty conclusive evidence that, I would say.

    No doubt Fifi can come up with some other drivel. And I encourage that enterprise.

  75. DrLoser says:

    VMkernel is a POSIX-like operating system developed by VMware and provides certain functionality similar to that found in other operating systems…

    Aha! But it’s a self-admitted Operating System!.

    No doubt Fifi will run and run with this one. Whether or not he makes any cogent observations whatsoever, or even provides a single relevant cite …

    … That is for you and the jury to decide, M’Lud.

  76. oldfart says:

    My good doctor, I heartily recommend the following white paper from VMWare on th architecture of ESXi

    http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CCcQFjAB&url=http%3A%2F%2Fwww.vmware.com%2Ffiles%2Fpdf%2FESXi_architecture.pdf&ei=NM4eVeT6K4OSsAXg1IDYBw&usg=AFQjCNH8VIqnTMdAJuZRCqsg891oXZyOBg&bvm=bv.89947451,d.b2w

    VMWare specifically states

    Of particular interest:

    “VMkernel is a POSIX-like operating system developed by VMware and provides certain functionality similar to that found in other operating systems, such as process creation and control, signals, file system, and process threads. It is designed specifically to support running multiple virtual machines and provides such core functionality as:
    • Resource scheduling
    • I/O stacks
    • Device drivers
    Some of the more pertinent aspects of the VMkernel are presented in the following sections….”

    even a cursory review of this should make it clear that yopu are not going to run openoffice on this puppy!

  77. DrLoser says:

    Tech term wrong hyper-visors type1 and type 2 have kernels. ESXi is a operating system and a hypervisor. Xen is only a hypervisor.

    “Tech term” irrelevant. Not that you, oiaohm, are any sort of expert in that area.

    There may be a legal definition of the difference between a hypervisor and an operating system. I’m not aware of it. But, since you can run any OS you like over a suitably configured hypervisor, I suspect that the legal point of view is that the two are distinct entities.

    Perhaps I’m wrong. But it’s not a distinction I’d launch an underfunded major legal effort like Nitwit vs VMware upon.

  78. DrLoser says:

    Again failing to ask yourself question. Why did the hardware vendors make Linux GPLv2 drivers instead of BSD ones.

    Because they sell, you idiot.

    And the question is irrelevant, you idiot.

    I’m not proposing that VMware should force hardware manufacturers to write BSD drivers. That sort of bully-boy nonsense is typical of Linux evangelists, who seem to think that they have a God-Given Right to dictate to everybody else.

    I’m proposing the thesis, and it is merely a thesis, that VMware could, if they wanted to, build the equivalent of vmxlinux from a FreeBSD basis and avoid an awful lot of legal hassle. And I have pointed out that there is no present reason for them to do so, unless the legal hassle becomes intolerable. Which Nitwit appears hell-bent on making a reality.

    How would you interface the drivers to a FreeBSD-based vmxlinux? I’ve pointed out that I could write the architecture myself, in six months flat.

    But the details are not relevant. The point is, this is a realistic scenario.

  79. oiaohm says:

    DrLoser
    Now, the courts may say one thing or another. But do either of you loonies really want to discourage hardware manufacturers from producing GPLv2-compatible drivers?
    Again failing to ask yourself question. Why did the hardware vendors make Linux GPLv2 drivers instead of BSD ones. Number one hardware vendors hate their competitors and don’t want their competitors to use their source code without either paying for it or showing their code as well. Allowing VMWare to get away with breaking GPLv2 would have negative out come for Linux getting drivers from hardware makers. So the Linux world has no option bar to defend the license.

    Basically VMWare actions might have been good for their customers but it has been annoying the companies making the drivers this is serous-ally not a good thing why a court case is required to clear it up.

    They supply a hypervisor, not a kernel. Upon this hypervisor their customers can install an actual Operating System.
    Tech term wrong hyper-visors type1 and type 2 have kernels. ESXi is a operating system and a hypervisor. Xen is only a hypervisor. What is the difference Xen has almost no hardware drivers and depends on what ever OS is dom0 to provide them. Yes everything provided by vmkLinux in VMWare solutions come from DomO in Xen.

    VMware ESX/ESXi is the only hypervisor that is called by people type 1 that has a stack of drivers. Everything else called type 1 for a majority of drivers uses a hosted OS. ESXi design is really closer to a type 2 hypervisor than a type 1.

    DrLoser there is a theoretical hypervisor that no one has built yet called a type 0 hypervisor that has no kernel.

  80. DrLoser says:

    If corporations really don’t care, they should use FLOSS. It’s less expensive and is not out to “get” them.

    Which is why the world of FLOSS is disproportionately populated by very vocal “not for profit” Ambulance Chasing Lawyer Trust Funds such as, say, the Software Conservancy.

    Well, sure. What could possibly lead the managers, accountants, lawyers, technical leads, and developers to believe that “FLOSS is out to get them?”

    (Apart from the fact that this is the foundational principle of the license itself.)

  81. DrLoser says:

    As I say, it’s evident that neither one of you have put any real thought into this. You’d better hope that Nitwit loses the case on a technicality. Because, in that eventuality, everything will proceed as before, and the world of FLOSS will be unruffled.

    If Nitwit wins? Well, as I say, you’ve just pissed off a major corporation who was otherwise well-disposed to maintaining Linux driver standards. And was otherwise well-disposed to contributing to the kernel.

    Bad Move.

    And if Nitwit loses?

    Hard to see how the GPLv2 doesn’t come to grief in terms of legal precedent. Which is incidentally going to be the subject of my post on the GPLv2 FAQ.

    Neither one of you has a clue on that, either, do you? Probably because neither one of you has bothered to read the foundational documents of the license.

    Not even once.

  82. DrLoser says:

    DrLoser there is no where for VMWare to go away from Linux without hurting at least some of their customer base. This is why is so hard to claim independent work in this area.

    Should they choose to do so (which will be a business expense entirely unnecessary, but for Nitwit vs VMware), I assure you that it will not hurt their customer base in any way whatsoever.

    From the bottom up:

    1. The interface to the drivers will be remodelled. Actually, this is quite trivial. All you need is a control path and a data path. I could write that generic architecture myself, in six months flat at most.
    2. Some parts of vmxlinux might need to be re-written. Goodness knows which bits. Not a problem for a 14,000 strong company with the time given by inevitable appeals, etc … and don’t forget that even if vmxkernel is suddenly Public Domain there’s nothing to stop them using it. They just won’t bother developing it. Replacing vmxware with a bought-in like VxWorks (heavily modified, but reasonable) or even FreeBSD is not exactly a challenge.
    3. The management substrate is still there. It is not tainted in any way. And the management substrate is what VMware customers pay big money for.

  83. DrLoser says:

    Robert Pogson says:

    April 1, 2015 at 10:06 pm

    oiaohm wrote, “Without Linux kernel there is a huge stack of hardware they will not support.”

    Exactly. If they had no need of Linux, they would not have used it to produce a non-Free product.

    Unsurprisingly, neither one of you appears to have the slightest clue as to the VMware business model.

    They supply a hypervisor, not a kernel. Upon this hypervisor their customers can install an actual Operating System. Despite being a fully paid-up member of the Linux Foundation (and incidentally neither of you two are), VMware is entirely agnostic about that Operating System. They leave it up to their customers.

    And they’re perfectly capable of supporting any hardware in town, as long as there’s money to be made. They don’t need to rely on the Linux kernel, which sits above the hypervisor (any hypervisor).

    And now we’re back to Exhibit Two. The only relevant part of the Linux kernel, as far as hardware support goes, is the Linux device driver headers.

    As I have pointed out, and as both of you have failed to argue against, this is a business decision and not a technical decision.

    VMware have chosen to model their device driver layer on the Linux device driver API, not because it suits them (they could use any old thing), but because it suits their customers. Their customers have a faster time to market that way. And are still completely accessible from the Linux OS sitting above the hypervisor.

    Now, the courts may say one thing or another. But do either of you loonies really want to discourage hardware manufacturers from producing GPLv2-compatible drivers?

    Because that, amongst other unsavoury results, is likely to be the result of this court case.

  84. oiaohm wrote, “Without Linux kernel there is a huge stack of hardware they will not support.”

    Exactly. If they had no need of Linux, they would not have used it to produce a non-Free product. This is all about their dance to make it seem that they are not using Linux in their product while they are. Could an authour write a non-infringing book including half the chapters of another authour’s book without permission? Nope. Those drivers are part of Linux. It’s common practice to include non-Free drivers as LKMs with Linux but nothing suggests using Linux drivers as part of another kernel is proper. VMware’s lawyers either thought they were lucky or completely misunderstand copyright/Linux. It’s like SCOG’s law firm all over again.

  85. oiaohm says:

    And there has been not a single court case to rule when taint from a license begins.
    There are cases of where taint begins.
    http://www.timelex.eu/en/blog/detail/gpl-license-reaffirmed-by-german-court

    There have been two Fantec cases what one have you been reading DrLoser.

    The Hamburg court didn’t even rule on “complete corresponding source code.”
    Guess what the older case did in 2004 in the Landgericht of Munich did rule what that means. So the Hamburg court did not have to rule on that. Note a German court. FOSS has choosen legal system that suits upholding GPL.

    This is the problem with nitpicking DrLoser. Its not going to get you very far.

    VMware would very definitely change direction, and one would assume the new direction would be actively hostile to any possible taint of FLOSS code.
    You have to answer why did VMWare do it in the first place. Without Linux kernel there is a huge stack of hardware they will not support.

    Please also note busybox fork toybox under BSD license was out in 2006. Even after the legal mess over busybox vmware is still using busybox. The reality is most users of busybox have stayed using busybox.

    VMWare having to give up source code to ESXi really would not be end of world. End of world for VMWare would be a ruling forbidding the usage of Linux kernel code include having to push out a update to disable their current usage. 90-100 percent of all current installed ESXi using systems depend on drivers from Linux to operate. If you use the BSD kernel drivers you will only recover about 70-60 percent to operational state. Using Windows drivers 50-40 percent to operational state. Yes how to say ouch. A lot of people complain about Linux hardware support but really its massive in the server side just not so much in the desktop side. Making hyper-visors for the enterprise market equals dealing with Linux.

    Why is Windows numbers so low. Its 5 year old+ hardware and very new stuff. Lot of old hardware drivers are not maintained for Windows and the new stuff has Linux drivers first in most cases. Why having the complete OS source code makes life a lot simple to debug when you are trying to prove if a new bit of tech works or does not.

    Reality here a hyper-visor cannot be using unmaintained drivers because that would be a true security nightmare. Yes the Windows arguement for closed source drivers is not a big help with hyper-visors. So if not Linux Vmwares only choice is to use BSD.

    DrLoser there is no where for VMWare to go away from Linux without hurting at least some of their customer base. This is why is so hard to claim independent work in this area.

    I also typoed vmxlinux instead of vmklinux. VMK is the vmware supposed API of separation. Problem is it really does not read like separation.

    DrLoser basically arguing that this might scare off other companies from using Linux might be true. But the reality is VMWare has no option bar to ride this out and accept the ruling or otherwise they will lose massive amount of market share. Citrix behind Xen would gain lot of VMware was crippled in the market . Basically we can expect some of VMWare competitors to chip in the cash for the court case because a Win out right will return them a 1000 fold in income. A half win would allow there third party management tools to hook up to vmware products better.

  86. DrLoser wrote, “Large corporations pay VMware hundreds of millions of dollars for their management software fabric.None of those people give a damn what is underneath that fabric.”

    Strange. Just a few years ago we were reading garbage like corporations should not use GNU/Linux or face legal liability to SCOG or worse. So, when did corporations quit caring about what’s inside? If corporations really don’t care, they should use FLOSS. It’s less expensive and is not out to “get” them.

  87. oldfart says:

    “None of those people give a damn what is underneath that fabric.”

    (grin…)

    Exactly!

  88. DrLoser says:

    This is not all that unexpected if you consider that everyone involved is positively salivating at the possibility of getting the source code for an enterprise grade type one hypervisor opened up for all to see and mooch on.

    Well, except for Fifi, who seems to think that an invasive hypervisor like Xen is sufficient. (Curiously enough, the market appears to disagree with him.)

    Obviously, not a single one of these bozos would know what to do if the source code to vmxlinux landed in their lap. But, more interestingly, they have equally no clue as to what the professional Linux software engineers they suck blood out of would be able to do with it.

    So, let’s hypothesis a Total Failure for VMware. That’s right, let’s hypothesise a Total Win for Nitwit and the Software Conservancy.

    On that hypothesis, there would no doubt be a substantial monetary consideration — which we can ignore, because it isn’t relevant and VMware will be able to afford it.

    On that hypothesis, VMware would very definitely change direction, and one would assume the new direction would be actively hostile to any possible taint of FLOSS code. I can’t see this as a Win For Principles, personally.

    But, on a more mundane level, we’re not going to see hundreds of FLOSS developers queuing up to hack on vmxlinux, are we? I mean, what would the incentive be?

    And on an equally mundane level, it still leaves VMware’s management software fabric entirely untouched. Because, well, that’s inherent in the GPLv2. (Although it’s completely separate from my “you sunk my battleship” clause. Keep trying, kiddies.)

    And, you know what? Large corporations pay VMware hundreds of millions of dollars for their management software fabric.

    None of those people give a damn what is underneath that fabric.

  89. DrLoser says:

    Anybody here yet pledged, say, fifty dollars to the cause, incidentally?

    It’s quite an important cause. It’s a cause you all espouse, quite vehemently on occasion.

    And that $50 will be matched by a generous soul! Which makes it one hundred dollars!

    I can cite the address for donations if you like. So, which is it to be? Free as in Speech (Legal) — Software Hero For The Ages?

    Or Free as in Beer — Blood-Sucking Drone?

    Let me guess …

  90. DrLoser says:

    In the Fantec case on complete source code you need to read the ruling. As a developer it is a worry.

    I read it, Fifi. I quoted from it.

    I did more on either count than you have done. Cease babbling, please. I am a developer; it does not worry me. You are emphatically not a developer, so the question is without consequence. VMware couldn’t care less, either.

    Please note question of taint spread applies to all viral software licenses. There have been very few court cases to rule when taint from a license ends.

    And there has been not a single court case to rule when taint from a license begins.

    You didn’t really read the court ruling on Fantec at all, did you, Fifi? And you most certainly paid no attention to my relevant point, which as usual I have to repeat for the benefit of your dense pseudo-intellect:

    The Hamburg court didn’t even rule on “complete corresponding source code.”

    To make this even more plain, Fifi, it seems incontrovertible to me that VMware have indeed provided “complete corresponding source code.” Complete with “complete corresponding licenses.”

    Given that, and given that we still have the awesome spectacle of watching a German Court dealing with the as-yet uncontested issue of “fair usage” as it applies to the GPLv2, I suspect that Nitwit stands at something of a Legal Disadvantage.

    And once all that crap is worked through … and it will be … and it will be very, very expensive for the Software Conservancy, because that’s what Lawyers like to do … we’re still left with the “I’ve just sunk your battleship” argument.

    Much as I’d like to serve this argument up to oldfart, as requested, I think it’s only fair that I give you the chance to discover it for yourself, Fifi. Redeem yourself!

    It’s fairly prominently stated in the GPLv2 FAQ. Five hundred and sixty five lines. Fifteen thousand and sixty five words.

    (Mostly pointless self-serving blather.)

    Go ahead, little Fifi. Figure out the “other shoe.”

  91. DrLoser says:

    That having been said, I look forward to your dropping the other shoe. Because we know that none of the people here have a clue as to what you are talking about.

    Not true, oldfart. Not true at all.

    Fifi Sees All. Fifi is All-Knowing.

    Although, and perhaps unexpectedly, Fifi is only All-Knowing after the event. But to be fair to Fifi, he has a 100% record on this.

  92. DrLoser says:

    drloser you are almost there. Almost.

    Between ESXi kernel and vmxlinux there is a API. Is this enough separation to make ESXi kernel and vmxlinux a independent work.

    Is there, oiaohm? Is there, really?

    Indicate vaguely, with cites if you feel like it (although you may feel too affronted in some obscure way to provide them), what the API between “the ESXi kernel and vmxlinux” might be.

    I ask this as a humble supplicant. Because right up until now, I had no idea that such a thing existed.

    The independent work thing is inspired bullshit, btw. And completely irrelevant.

  93. oiaohm says:

    2. VMware contravenes the GPLv2 because vmxlinux incorporates GPLv2 code, which means that all the rest of the executable is tainted by the license and must be freely available.
    drloser you are almost there. Almost.

    Between ESXi kernel and vmxlinux there is a API. Is this enough separation to make ESXi kernel and vmxlinux a independent work.

    Train wreck? Yes, there’s a train-wreck coming straight down the tracks. But nobody here seems to have the faintest clue as to what that train-wreck might be.
    DrLoser this is so true. The question here is how far taint goes. Please note question of taint spread applies to all viral software licenses. There have been very few court cases to rule when taint from a license ends. So current VMWare vs Christoff Hellwig is a worry. Not just for GPL its also for viral licenses that state no commercial usage or restrictions on usage. This case will partly define how viral a viral software license is.

    In the Fantec case on complete source code you need to read the ruling. As a developer it is a worry. The courts hand down a no ruling so not for or against. Fantec had agreed to the court they would update the source code what they did. The problem in the Fantec case is that it came down that Fantec could not be held 100 percent responsible for their developers doing the wrong things. Yes of the FSF wanted to follow up they were free to sue the individual developers responsible at Fantec and Fantec was responsible to hand over their names. Yes screw up with a license as a closed source developer for a company you might get sued individually with your company throwing you to the wolves to save themselves.

    Please note the Fantec iptables closed source modules was also API separation arguement that they lost. The difference between the Fantec case and the current VMWare case is the link direction.

    Fantec modules was GPL work loading closed source without proper independence. vmkLinux and ESXi kernel is closed source loading GPL without proper independence. There is no ruling existing ruling to save or prosecute VMWare but prior rulings like Fantec give grounds to put the case before judge and get ruling who is right.

    DrLoser yes its the Fantec case and others that under mines the VMWare no mert arguement.

    Which is quite interesting, considering that there is a collective failure to defend the first proposition.
    Why is no one here bother with the other points is everyone else read the faq case.

    Furthermore, the technical details of VMware’s alleged GPL violation do not even mirror the typical scenarios that have usually been called “shim layers”. Conservancy’s analysis of VMware’s ESXi product, in fact, indicates that VMware rather flagrantly combined Linux code in their own kernel, and evidence seems to indicate the work as a whole was developed by modifying Linux code in tandem with modifications to “vmkernel” in a tightly coupled manner.
    DrLoser read this. Note they are talking about modifications inside the vmkernel lining up with modification caused in Linux. Please note vmkernel is not vmkLinux. So there arguement is that vmklinux is not a truly independent shim to vmkernel. No independent work status vmware should release the source code.

    The FAQ does not suggest they are going to bother about the independent parts you have been nitpicking over other than for evidence of alignment in changes between vmklinux and vmkernel.

    DrLoser why have we been mostly ignoring your points because you cannot do the simple thing of click on a link and reading the FAQ before commenting. We have not being fighting you points because for this case they are pointless as the case is not about that.

    oldfart got lost its not that we are drooling over the source code being released. The FOSS world already has a enterprise grade type one hypervisor in Xen and others. Please note Amazon, Rackspace and others use Xen. VMWare license expressly forbids benchmarking against competing products without permission.

  94. oldfart says:

    “Which is quite interesting, considering that there is a collective failure to defend the first proposition.”

    This is not all that unexpected if you consider that everyone involved is positively salivating at the possibility of getting the source code for an enterprise grade type one hypervisor opened up for all to see and mooch on.

    That having been said, I look forward to your dropping the other shoe. Because we know that none of the people here have a clue as to what you are talking about.

  95. DrLoser says:

    In other words, this case boils down to one of two possible Nitwit objections:

    1. VMware contravenes the GPLv2 on one or more of Exhibits One, Two, Three and Four. (Which, may I remind you, I have cribbed directly from the Software Conservancy cite in the OP.)
    2. VMware contravenes the GPLv2 because vmxlinux incorporates GPLv2 code, which means that all the rest of the executable is tainted by the license and must be freely available.

    Correct me if I am wrong, but I believe the Pogson/oiaohm/everybody else position hinges on the second proposition.

    Which is quite interesting, considering that there is a collective failure to defend the first proposition.

    And no, Robert, “it’s part of the operating system” just won’t do.

    Train wreck? Yes, there’s a train-wreck coming straight down the tracks. But nobody here seems to have the faintest clue as to what that train-wreck might be.

    And I repeat: I am a fair and reasonable man. The train-wreck in question is quite obviously going to be based on Proposition Number Two. I’ll even help you out further: Proposition Number Two is precisely why VMware’s lawyers claim that Nitwit vs VMware has “no merit.”

    Although they will undoubtedly be quite happy bleeding the Software Conservancy dry by arguing the inconsequentials on Proposition Number One.

  96. DrLoser says:

    However, I will give oiaohm great credit for actually trying to come up with a cite that supports the GPLv2 position vis-à-vis the VMware case. As a matter of fact, he’s dug up a very solid one. It covers a lot of relevant stuff.

    Bit of a shame that nobody else on this Site For Linux Enthusiasts can be bothered to put in as much as a tenth of the effort that oiaohm makes, but whatever. Let’s deal with what’s provided to us here.

    Here is something scary German counts any customer who had bought VMWare could enforce the GPLv2 license as well due to the fact they were being denied their right to modify as much code as they should have been. That is another case that a company lost in germany.

    Unfortunately, not a very scary one at all, if you are VMware. The cases are clearly not comparable:

    The software available for download for the Fantec product was reviewed during a “Hacking for Compliance Workshop” in Berlin organized in 2012 by Free Software Foundation Europe. The hackers discovered that the source code provided by Fantec did not include the source code for the iptables software and that the source code for some other components did not match the versions used to compile the binary code of the firmware.

    My emphases. Now, I happen to think that this attitude is needlessly nit-picking (unless of course the modified iptables code is actually unavailable, which is a clear violation of the license). However, it’s the sort of nit-picking for which the GPLv2 is expressly designed, and the Court obviously agreed with the license.

    Sort of.

    On June 14, 2013, the district court of Hamburg found that Fantec violated the obligation in the GPLv2 to provide to its customers the “complete corresponding source code” of the software. The court affirmed a violation of the GPLv2 license conditions because the iptables code was not contained within the source code provided by Fantec. However, the court did not rule on the second argument that the source code was not up to date. Consequently, the decision does not provide significant guidance on the definition of the term “complete corresponding source code.”

    Crumbs, that’s even worse than I thought. We haven’t even got to the concept of “fair use” yet, and we’re already stuck on the definition of “complete corresponding source code.”

    Which, I might add, refers exclusively to GPL’ed modules like iptables. And not to external elements compiled into the same executable, which would have to be the claim with vmxlinux.

    Frankly I don’t see this as a Good Sign for Nitwit vs VMware. But, as I say, I am a generous and open-minded soul, and unlike Fifi I do not withhold cites for reasons of pure childish personal spite.

    In fact, I’ll even give you one that I propose to use, but which has not yet been featured in the present discussion. I assume you have all read and digested the GPLv2 FAQ.

    And I’m pretty damned sure that the VMware lawyers have. There’s one particular answer in there that, to use Robert’s phrase, sinks Nitwit’s battleship.

  97. DrLoser says:

    Some of the german rulings have advantages. Like devices using GPL works the maker cannot punish you for modifying the GPL work and even making the product do extended functionality.

    Which has nothing to do with anything at all, does it, Fifi?

    You never saw a Google link you didn’t feel like spewing out, regardless of reason, did you?

  98. DrLoser says:

    DrLoser Hellwigs busybox code was used against another company. The German court ruled that the small fragment of code was enough to enforce the license.

    Fifi the point of that quote was that Busybox has nothing to do with the case whatsoever. And that the Software Conservancy initially thought they could put one over on VMware using Busybox as an excuse. And that they then realised they could not. And that they went searching for some other excuse. And that Hellwig provided them with one … which I am inclined to think, upon analysis, is not necessarily much better.

    In other words, this particular Not For Profit positively stink of Ambulance Chasing.

    Do I really need to repeat my point over and over? If I’m talking to you, Fifi … evidently I do.

  99. oiaohm says:

    http://www.computerworld.com/article/2498725/enterprise-applications/german-court-rules-users-can-modify-free-software.html

    Some of the german rulings have advantages. Like devices using GPL works the maker cannot punish you for modifying the GPL work and even making the product do extended functionality.

  100. oiaohm says:

    DrLoser Hellwigs busybox code was used against another company. The German court ruled that the small fragment of code was enough to enforce the license.

    This is disturbing to me because it is clearly a scattergun approach to license infringement. It’s akin to both tort law abuse and its nasty little cousin, ambulance chasing.
    Vmware was in fact infringing many FOSS license. The VMware open source releases happened after the Busybox case and discoveries.

    BSA uses the same method they get in the front door then audit.

    DrLoser I was not using a current link. I remember the threat when only a small percentage of work in a GPL work could in fact do enforcement that was proven in german courts when.

    http://www.itworld.com/article/2732025/open-source-tools/gpl-enforcement-sparks-community-flames.html

    Its not like flames worry us. DrLoser your arguement is the license should not be enforced because it might give open source a black eye. You are missing that its both ways. Companies that agree to release source under GPL so their competitors cannot nick off with it also get upset when GPL is not enforced.

    http://mjg59.dreamwidth.org/10437.html

    It was this post back in 2012 before the VMWare current case got under way that got me digging. In the westinghouse case in the USA was argued there was not enough code and the court ruled on the side of SFC. There is also a German case ruling the same way.

    Dr Loser something to remember the GPL cases are showing us what will count as tainted by a licensed work. Yes how far reaching academic licence restriction and other thing will be in court.

    How do I know? I don’t really. But a very large American corporation with pots of money and very well-prepared lawyers are about to pick every single nit they can find out of one or both of these otherwise lightly examined licenses.
    Remember this is a German court not a USA one. This is critical. Microsoft vs Samba. Microsoft tried the nit pick method. Not a good idea Lawyer end up in jail. Nitpicking does not work.

    http://opensource.com/law/13/7/fantec-german-foss-compliance

    GPL is very heavily tested license in German courts.

    GPL and LGPL are not lightly examined licenses in German counts. DrLoser your agreements would hold if this was a USA court.

    Here is something scary German counts any customer who had bought VMWare could enforce the GPLv2 license as well due to the fact they were being denied their right to modify as much code as they should have been. That is another case that a company lost in germany.

    Yes not enough code to not having any code have all been tested in german courts and both proven not base less. To enforce in a german count either
    1) have at least 1 line of copyright code in the work.
    2) bought the product.
    If you have done neither you cannot enforce the license.

    Yes when dealing with licenses it pays to remember countries like Germany your customers can enforce what ever licenses your product use.

  101. Dr Loser says:

    Now, if I were oiaohm, I would claim Flawless Victory here. Because apparently I have beaten everybody else into the ground, one way or the other. (Don’t worry — Easter is coming up soon, and I’ll have time to consider SCSI.)

    However, I am not.

    That being said, I need to remind you that
    a) I have no skin in this game. It is purely an intellectual discussion (leaving aside personal snipes, which do not affect the issues in Nitwit vs VMware).
    b) I think this case is a really, really bad case to bring if you want to “defend” the GPL.

    So, let’s concentrate on the GPL. Specifically, GPLv2, which is most obviously the subject of legal debate.

    Although, interestingly enough, the LGPL is very likely to feature in this case.

    How do I know? I don’t really. But a very large American corporation with pots of money and very well-prepared lawyers are about to pick every single nit they can find out of one or both of these otherwise lightly examined licenses.

    And since nobody here has come up with compelling evidence why Exhibits One, Two and Three are really all that legally obnoxious, we regrettably come down to the standard (GPLv2 and LGPL) conclusion:

    We want all of your software, because you used some of ours!

    … which is the implicit threat behind all versions of the GPL.

    Unfortunately, and I’ve been reading up on it as you will (if you don’t waste time just collecting headlines like Robert and oiaohm) …

    I don’t believe this is necessarily a defensible case.

  102. DrLoser says:

    I was wondering what oiaohm was drivelling on about when he mentioned BusyBox, which is clearly irrelevant to the current case. My initial assumption was that he had picked up on my long-ago (on this thread) comparison to the BusyBox lawsuits of the last decade. But it turns out that I might be wrong, and that oiaohm has managed to google up a genuine link to the present case before the court.

    I find this quite disturbing, actually. (Not the fact that oiaohm is an ignorant googling maniac who doesn’t understand what he reads … that’s pretty much a given.) The disturbing thing is this:

    There are a few claims about what Hellwig wrote that aren’t quite right, in particular, “Hellwig and others have written a set of tools called BusyBox. … Hellwig has a number of copyrights easily seen in the BusyBox versions.” Hellwig hasn’t written much BusyBox code that I’m aware of. While Conservancy did initially begin our enforcement against VMware because of a BusyBox violation, we then found a violation in the kernel named Linux…

    This is disturbing to me because it is clearly a scattergun approach to license infringement. It’s akin to both tort law abuse and its nasty little cousin, ambulance chasing.

    Software Conservancy: We can’t get you on BusyBox infringement? Damn! But we can probably dig something else up on you. Let’s see … somebody somewhere wrote a nastygram to VMware in 2006 … let’s dig that one out of the archives … Aha!

    On the whole, I prefer not to deal with legal scumbags paid through tax exemptions when they come up with bullshit like this. In fact, I would prefer them to risk their own money, if they’re going to come up with a suit.

    But what the heck, nobody is as Pure as the Driven Snow.

    If all you have going for you is Ambulance Chasers, Tort Lawyers, and “Charitable Tax-Free Foundations,” I don’t suppose you can be all that choosy.

    You are, however, encouraged to permit a small smidgeon of doubt into your mind. It’s just possible that these people do not have the best interests of FLOSS in mind.

  103. DrLoser says:

    I don’t need a cite for the following, oiaohm. I don’t need a lengthy discursion on the legal ins and outs, the principles of accountancy involved, the number of sun-spots affecting microwave radiation in the Van Allen belt in 2015, or any single one of your various exotic pointless trivia.

    All I need is an estimate from personal experience. Even a vague estimate.

    Yes those other Authors have fixed security faults and other defects in the driver. So VMWare is profiting from their work so do owe them payment.

    On an estimated basis, oiaohm, how much have you paid any provider of freely-available FLOSS updates for “security flaws and other defects since, let’s be generous, 1990?

    Me? I’ve paid nothing. I’m not proud of this, but it does seem to be the going rate.

  104. DrLoser says:

    vmxnet3 has been modified by many different authors. So the copyright head by it self is not enough to say its VMWare work pure you need the git log of modifications.

    Everything under the “VMware” copyright line is the bog standard text of the GPLv2 license, oiaohm. Do you not recognise the bog-standard text of the GPLv2 license, oiaohm? I’m beginning to think that you have never read the bog -standard text of the GPLv2 license, oiaohm.

    The point of my cite, oiaohm, is that the bulk, indeed the important part, of the vmxnet3 license is the bog-standard text of the GPLv2 license.

    And that there is therefore little or no evidence to suggest that VMware is not complying with the GPLv2 license in re vmxnet3.

    I regret having to repeat this simple message about five times, but if you’re going to be dense and not listen the first time, my only recourse is to sledgehammer repetition.

    VMware is not in any obvious way infringing the GPLv2 when it comes to vmxnet3.

    Everything else you had to say on the matter is therefore pointless babble, as always.

  105. oiaohm says:


    a) Some pitifully small rip-off of a Radix Tree implementation
    b) Misuse of the Linux device driver APIs
    c) Misuse of actual Linux device drivers
    d) Misuse of Linux SCSI code

    This is forgetting the busybox case.

    The complete work of busybox was protect by a author than had less than 20 lines of code in the complete work.

    Proff of your code used inside a project allows you to enforce the copyright of the complete project.

    VMWare says that VMKlinux is a GPLv2 project since Christoph Hellwig contains code in that project he can enforce the license of that complete work.

    DrLoser this is why breaking down into parts does not work. You are forget the busybox precedent.

  106. oiaohm says:

    DrLoser in Germany copyright can be void by the courts. Its an very extreme case and you have to have done a lot wrong. What causes a Author to lose copyright the laws against profiting from a crime do it. The result is your authorship is removed and the state/country where the work was created existing rights become the party 100 in-control of the work. The official Author of the work becomes Anonymous/Public Domain. Yes after this you are not legally allowed to even attempt to claim ownship as this would be attempting to profit from the work.

    Yes a court can officially delete you from being a Author of your own work so removing all your copyright law control over your Work.

    DrLoser I did not say transfer you dug up a quote that copyright cannot be transferred this is true. Copyright can be deleted by the courts. The country where the work was created has rights from the very day the document was created. So there is no transfer when the counts take control of a copyright. The law allows removal under very particular cases such as profiting from a crime. Copyright cases can turn very nasty.

    vmxnet3 has been modified by many different authors. So the copyright head by it self is not enough to say its VMWare work pure you need the git log of modifications. DrLoser you quoted the wrong thing. You need to quote the authors to the file as well that is in git log and there are a lot modifications to that driver that are not vmwares. Just to be fun the file you have brought in as example if you go and read the git log on it has over 200 different authors. Only 2 of those are VMWare. The rest own to other companies.

    DrLoser the file you have brought up is extremely tainted so VMWare cannot ship as it as a closed source item so has to obey GPL license of the other Authors. Yes those other Authors have fixed security faults and other defects in the driver. So VMWare is profiting from their work so do owe them payment.

    Network Plugin Architecture was questioned on legality as well.

    Just because you are the open source author does not mean can do actions that breach GPL work itself. Remember GPL license agreement the author and the end users agree to uphold.

    Network Plugin Architecture to go mainline needs the upcoming court ruling to know if it legal.

  107. DrLoser says:

    I’m sure Fifi will drag this up eventually, but just to spare everybody six days of pointless waiting … here’s the source code for the vmxnet3 Linux driver.

    Oh, but every last one of you is too damned lazy to read beyond the header. And obviously you all want proof of GPLv2 compliance. Well …

    /*
    * Linux driver for VMware's vmxnet3 ethernet NIC.
    *
    * Copyright (C) 2008-2009, VMware, Inc. All Rights Reserved.
    *
    * This program is free software; you can redistribute it and/or modify it
    * under the terms of the GNU General Public License as published by the
    * Free Software Foundation; version 2 of the License and no later version.
    *
    * This program is distributed in the hope that it will be useful, but
    * WITHOUT ANY WARRANTY; without even the implied warranty of
    * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
    * NON INFRINGEMENT. See the GNU General Public License for more
    * details.
    *
    * You should have received a copy of the GNU General Public License
    * along with this program; if not, write to the Free Software
    * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
    *
    * The full GNU General Public License is included in this distribution in
    * the file called "COPYING".
    *
    * Maintained by: Shreyas Bhatewara
    *
    */

    Have a nice day, suckers.

  108. DrLoser says:

    Further on that thread, Robert. I assume you have fully read and digested it. But for the benefit of those less able than you (and they all know who they are):

    VMware has been working on various device passthrough technologies for the past few years. Passthrough technology is interesting as it can result in better
    performance/cpu utilization for certain demanding applications. In our vSphere
    product we support direct assignment of PCI devices like networking adapters to
    a guest virtual machine. This allows the guest to drive the device using the
    device drivers installed inside the guest.

    That “guest virtual machine” would be Linux, Robert. Thus the need for a plug-in to the Linux kernel. Perhaps you have a deep desire that the Linux kernel is to be crippled when it comes to talking to a hypervisor? How interesting.

    We have reworked our existing Linux vmxnet3 driver to accomodate NPA by splitting the driver into two parts: Shell and Plugin. The new split driver is
    backwards compatible and continues to work on old/existing vmxnet3 device
    emulations. The shell implements the API interface and contains code to do the
    bookkeeping for TX/RX buffers along with interrupt management.

    tl;dr. Particularly if you have no clue what you are talking about.

    The “Linux vmxnet3 driver” is, surprise, surprise!!! a Linux driver! Inside the Linux kernel! Like every other Linux driver!

    Oh, and:

    The IHV plugins are required to be distributed under GPL
    license
    and we are currently looking at ways to verify this both within the hypervisor and within the shell.

    That was back in 2010, Robert, according to your own cite.

    Perhaps you can point out how tardy these dreadful people have been?

    Or perhaps not.

    You seem to take a magpie attitude to verifiable truth, don’t you?

  109. DrLoser says:

    We know they derived stuff from GNU/Linux because they made contributions to Linux and told the world they had modified Linux in “tests”. They have to prove they didn’t distribute those modifications.

    This is an insane proposition, Robert. Utterly insane and without merit.

    I’ll take the example of a bit of Gnu I know very well, to whit, Bison. As you all know, Bison is the equivalent of the old *nix yacc, as flex is the Gnu equivalent of the old *nix lex.

    It so happens that Bison/Flex is a far superior combination to Yacc/Lex. I know this. I have tried both. I ended up writing an enhanced emulator for the VOS “init” system that relied on Bison/Flex as the generator. All good so far, and if anybody wants to see the code, I can provide them with it — although since nobody uses VOS any more, it’s almost certainly useless.

    However, I am quite the expert when it comes to Bison/Flex. And I have experimented with “Hairy Bison.”

    Hairy Bison quite probably marks the point in time when Richard Milhous Stallman went completely off the rails. It’s a disastrous mistake. But in a good cause: possibly you can improve on the Bison template.

    I run a few tests with my own ideas. I try out several possible improvements. None of them work satisfactorily.

    Because they are TESTS.

    You would have to be categorically insane to insist that the GPL requires you to publish each and every set of code that you have written in order to test something acquired under that license.

    Categorically Insane.

  110. DrLoser says:

    We intend to upgrade the upstreamed vmxnet3 driver to implement NPA so that Linux users can exploit the benefits provided by passthrough devices in a
    seamless manner while retaining the benefits of virtualization. The document
    below tries to answer most of the questions which we anticipated. Please let us
    know your comments and queries.

    Still nothing at all to do with the present case, Robert.

    Are you intentionally trying to cripple Linux users of VMware? Just because you, personally, don’t use a hypervisor, it doesn’t mean that other people with a Linux IT corporate base are prepared to forgo the thing.

  111. DrLoser ragged about the uncertainty of legal procedings and wrote, “you seem to think that only one of these mangled bodies will be left on the track”.

    Unlike the law of contracts where there are two sides to every story, copyright is much more clear. The court can demand to see the code and see for itself whether the work is derived. If VMware does not show the code, they lose. If VMware shows the code and there is substantial copying/modification, they lose. We know they derived stuff from GNU/Linux because they made contributions to Linux and told the world they had modified Linux in “tests”. They have to prove they didn’t distribute those modifications. That’s going to take a lot of work and will fail.

    e.g. read this thread in LKML:“RFC: Network Plugin Architecture (NPA) for vmxnet3”

    Notice what VMware was trying to do, use Linux drivers that had nothing to do with Linux, “Dmitry Torokhov wrote:
    > > Yes, with the exception that the only body of code that will be
    > > accepted by the shell should be GPL-licensed and thus open and available
    > > for examining. This is not different from having a standard kernel
    > > module that is loaded normally and plugs into a certain subsystem.
    > > The difference is that the binary resides not on guest filesystem
    > > but elsewhere.
    >
    > Forget about the licensing. Loading binary blobs written to a shim
    > layer is a complete pain in the ass and totally unsupportable, and
    > also uninteresting because of the overhead.”

    Major kernel developers have never supported the tricks VMware was playing, trying to misuse Linux and the GPL.

  112. DrLoser says:

    They get to use all that lovely Linux code without having to give back their modifications to it. That’s wrong.

    It might very well be wrong. That would be a different case, in a different court, Robert.

    I’m a little tired of specifying individual bits and pieces of the Software Conservancy case, so let’s make it as general as possible, with regards to your particular complain in this case.

    Pick a single bit of ESXi code, derived from GPLv2 code (you are allowed to stipulate the Linux kernel or otherwise), which falls into the following category:

    * Code (lovely, hideous or just plain … well, plain code. It makes no difference) appropriated by VMware to build ESXi but the modifications for which are not publicly available

    That’s a pretty fair test, isn’t it? I haven’t really placed any limitations on your choice at all. And if I have, you can leave the limitations aside.

    Just one single item of GPLv2 code, modified by VMware or their associates, however inadvertently, that is not publicly available in its modified form, complete with a reference to the relevant license.

    Ten … nine … eight … OK, ignore specious comments from Fifi … seven … six … five … four … three … really, ignore the pointless little prat … two … one …

    There might be one out there, Robert. And you are diligent enough to search for it via the various sources you use when researching articles for this site.

    And if you find one, I will be the first to congratulate you.

  113. DrLoser says:

    Uh, a mess of Linux drivers. They are part of the Linux kernel.

    No, Robert. Drivers plug in to the Linux kernel. They are not an inseparable part of the Linux kernel. Particularly not if they are third party drivers. This is not to deny that any given Linux device driver is subject to the GPL, presumably version 2.

    They plug into a modified Linux kernel.

    Not in this case. Not with ESXi. Not, in fact, with any similar hypervisor architecture. You want to know why? Because a hypervisor is not an OS kernel. That is why. How often do we need to repeat this to you?

    Now, you can choose to attempt a proof that vmklinux is a “modified Linux kernel.” But, absent the bits and pieces from Exhibits One, Two, Three and Four, there really is no such proof whatsoever.

    You’ve tried the “it boots from Red Hat” gubbins. That hasn’t worked since the original ESX. You’ve tried the “it needs Grub to boot” gubbins. That hasn’t worked either.

    You are left with

    a) Some pitifully small rip-off of a Radix Tree implementation
    b) Misuse of the Linux device driver APIs
    c) Misuse of actual Linux device drivers
    d) Misuse of Linux SCSI code

    All of which are legitimate grounds for complaint.

    But for some disastrously insane reason, you are still bleating on about a “modified Linux kernel.”

    Good thing for Hellwig that you are not going to be called up to support him in court.

  114. DrLoser says:

    I am just a bystander to this train-wreck.

    Oh, we all are, Robert. Me more than most. You care about the GPL: I do not. Neither of us cares about VMware, Hellwig, the Software Conservancy or pretty much anything else about the case.

    But we both share the morbid fascination (part of the human condition) of observing “train wrecks.”

    The only difference is that I will be able to look on with equal fascination on the mangled body of either a) VMware or b) the GPL (Hellwig, Software Conservancy, whatever).

    For some bizarre reason, for which you have provided precisely zero arguments other than “I likes the GPL, so I does,” you seem to think that only one of these mangled bodies will be left on the track.

    I have to warn you: this here is legal proceedings. Even if you are morally, technically and in all other ways “in the right…” you are still liable to a crushing disappointment. And since you haven’t made a convincing argument that you are in the right on Exhibits One, Two and Three, you are quite likely to be unpleasantly surprised.

    Never mind, there’s always Exhibit Four, Linux SCSI. Your thoughts on this are welcome.

  115. DrLoser wrote, “ESXi does not plug anything at all into the Linux kernel.”

    Uh, a mess of Linux drivers. They are part of the Linux kernel. They plug into a modified Linux kernel. Where’s the source code of that kernel?

  116. DrLoser says:

    Do you think it’s OK to modify the Linux kernel simply by plugging in a module?

    Do you still labour under the misapprehension that this has any relevance at all to the present case, Robert? God knows, we’ve explained this at least five or six times.

    ESXi does not plug anything at all into the Linux kernel.

    IT IS A HYPERVISOR. It is not a version of the Linux kernel. The Linux kernel runs on top of the hypervisor.

  117. oldfart says:

    “By not providing the correct information you have just dug yourself into a super deep hole.”
    I am not playing your game little man. Say whatever you want, the words of a demonstrated and admitted liar and fraud are just noise.

  118. DrLoser says:

    Yes a court can officially delete you from being a Author of your own work so removing all your copyright law control over your Work.

    No, it cannot. At least, not in Germany.

    Das Urheberrecht kann in Erfüllung einer Verfügung von Todes wegen oder an Miterben im Wege der Erbauseinandersetzung übertragen werden. Im übrigen ist es nicht übertragbar.

    Ignorance is no defence in front of the Law, oiaohm.

    Which is a pity, really. Because, were it to be so, you would be the easiest person to defend in the whole wide world.

  119. oiaohm says:

    Yes a court can officially delete you from being a Author of your own work so removing all your copyright law control over your Work.

    Most people have no clue how dangerous taking a copyright case to court can be. There is a major reasons why most copyright cases are settled out of court.

  120. oiaohm says:

    http://www.wsdot.wa.gov/NR/rdonlyres/4B167F66-B393-496F-A4A0-0BBD69F8F938/85821/20120709SettlementAgreementSummaryFINAL1.pdf
    OldFart it does not have to be a open source either. The nastiest ruling a court can do against a software company is that their Source code to their product is no longer theirs.

    It does not pay to mess around in court and piss a judge off in a copyright case.

  121. oldfart wrote, “you should suspend judgement until the court rules.”

    That’s an option. However, I’ve read GPLv2 and know how Linux works, more or less. Do you think it’s OK to modify the Linux kernel simply by plugging in a module? Why does VMware require Linux modules? Do you think a Linux module can act like an operating system without modification of the Linux kernel? Why would VMware do this if they didn’t want to have their cake and eat it as well? I may suspend judgment but I won’t suspend rational thought. VMware is cheating. One can argue how serious the violation is but I don’t see any way it’s not a violation. oldfart who claims to support the rights of authours should be in Hellwig/SFC’s camp.

    OTOH, VMware’s motivation is clear. They get to use all that lovely Linux code without having to give back their modifications to it. That’s wrong. The court can come to no other conclusion. “wrong” is not a legal term but the court will find 100 other ways to state the obvious, you can’t modify Linux, distribute it and not show the code.

    In such a trial, the plaintiff can insist to see VMware’s stuff build and run without using Hellwig’s code in the versions VMware has distributed. VMware will fail to do that and lose the case. It’s called “discovery” and no amount of hand-waving can avoid it. In Germany, courts won’t put up with delays either. They are punctual. We could know the outcome within a few months.

  122. oiaohm says:

    oldfart the fact you have claimed to be a VMWare product user not knowing about the source code release is prove of your incompetence it was in the memos VMWare sent out to their users. Or do you happen to use VMWare products by copyright infringement. Basically either you are person who does copyright infringement or you incompetent oldfart. By not providing the correct information you have just dug yourself into a super deep hole.

    Robert in prior posts on the blog had mentioned the solutions he used as well. Yet for some reason Oldfart you went off with the baseless presume that Robert would gain or lose by VMWare source code release or not.

    Sorry Oldfart you arguement had absolutely no base in fact or logic.

    Robert
    Why is the case in court if they were releasing enough? Hellwig and FSC have evidence of illegal copying or they would not have brought the case.
    Robert the problem here is derivative work and independent work define has not been fully tested in any count. VMWare argument that there product is combined work of two independent works. Hellwig is that it is a derivative work.

    Depending on what way you interpret the define both could be right. But VMWare is kinda behind attempt to prove independence.

    Legally you don’t have the right to use GPL unless its terms are followed. One broke term and you are using the GPL code without a license.

    VMWare and Hellwig has got down to 1 sticking point. Problem here is VMWare did in fact ship for years in breach of GPL before starting to come into compliance to avoid a case. So VMWare is no angel.

  123. oldfart says:

    “Hellwig claims VMware ignored that problem by distributing Linux code as their own”

    But Hellwig is not the court. That being the case, you should suspend judgement until the court rules.

  124. oldfart wrote, “Robert Pogson still doesn’t get any new freebies.”

    I want nothing from VMware. I don’t use their stuff. Never have. KVM works for me as did Virtual Box before Oracle went after FLOSS. I would be happy if VMware stripped out their illegally used FLOSS and suffered the consequences. I would be happier if they published their modifications of Linux as required by the licence but I am just a bystander to this train-wreck.

  125. oldfart says:

    “When the commodity of the case is source code the damage is many times higher than normal if the case is fort incorrectly.”

    Of course you can’t prove any real experience to back up this, so it becomes pure speculation on your part.

  126. oiaohm wrote, “VMWare has been releasing code for ESXi to a point. The question is if they have released enough.”

    Why is the case in court if they were releasing enough? Hellwig and SFC have evidence of illegal copying or they would not have brought the case. “enough” means complying with GPLv2 wrt Linux source code. The stuff VMware has published is stuff that was put into Linux with Linus’ blessing. VMware was not allowed to put in their non-Free software hooks. Hellwig claims VMware ignored that problem by distributing Linux code as their own. That was the wrong thing to do and the court will decide.

    Hellwig is not sue-happy. Neither is SFC. They want copyright respected.

  127. oldfart says:

    “Now sorry I have done a master job at tricking you.”

    HAHAHAHAHAHAHAHAHA.

    Lying again aren’t you?

  128. oldfart says:

    “oldfart the reality is you are trolling Robert calling him a cheapscate providing no valid information to the arguement. ”

    The valid information is provided by Robert Pogson himself time and time again in his blog posts. He has even boasted of his cheapness. I feel no need to provide anything else.

    “Why was it me who had to show Robert that VMware was releasing source code. That should have been your post if you were the expert you are claiming to be.”

    It should have been Robert Pogsons business to do his somework. Instead he went of (as usual) half cocked, And you who can’t resist butting into any conversation (also as usual) used your (most likely) freshly googled information to “correct” him.

    As far as your estimation my expertise is concerned, since you are a demonstrated and admitted liar and fraud, it is nothing but noise on paper.

    And one does not get upset by noise, especially the small squeaky ones from the rodent that you are.

  129. oiaohm says:

    oldfart where are you cites to challenge Robert. I see 1 rule for me and a different rule for you.

    Now sorry I have done a master job at tricking you.

    But it does not have to – nice to dream eh cheapskate.
    Just because you talk about a possiblity does not mean you are a cheapskate.

    Worst case is VMWare fights incorrect and the judge decide to take payment out VMWare hide in the form of the source code to all their tools.

    Cease and Desist order would still a quite minor punishment. Historically legal discovery order in GPL cases has been used to get courts hands on source code under NDA of court that is then undo to release the source code to public as payment of costs.

    When the commodity of the case is source code the damage is many times higher than normal if the case is fort incorrectly.

    oldfart the reality is you are trolling Robert calling him a cheapscate providing no valid information to the arguement. Why was it me who had to show Robert that VMware was releasing source code. That should have been your post if you were the expert you are claiming to be.

  130. oldfart says:

    “The law of copyright can result in cease and desist order from the court which would shut down VMware”

    But it does not have to – nice to dream eh cheapskate.

    ” Customers will be prompted to switch to KVM, Xen or even that other OS, diluting VMware’s business no matter the outcome.”

    And when they do, VMWare will be there with tools that allpw their management suites to manage Xen, KVM, or even Hyper-V based VM’s.

    Its all good.

    And Robert Pogson still doesn’t get any new freebies.

  131. oiaohm says:

    http://www.v-front.de/2011/10/vmware-finally-released-open-source.html

    Robert Pogson you have missed this. VMWare has been releasing code for ESXi to a point. The question is if they have released enough. This is why VMWare can legal argue mitigation against too extreme of punishment.

    Now if VMWare was still 100 percent not conforming they would be toast.

    Allwinner with all the stunts they are doing to try to hide the fact they are using FOSS if case was brought would be really open to have the book thrown at them.

  132. oldfart says:

    “Like in responses here there are 22 cases of DrLoser not being logically sound.”

    And because you have been so effective in faking being a liar and fraud, I still don’t believe anything that you say.

    Idiot.

  133. oldfart says:

    “Like in responses here there are 22 cases of DrLoser not being logically sound.”

    And because you have been so effected in faking being a liar and fraud, I still don’t believe anything that you say.

    Idiot.

  134. oldfart says:

    “Really how better to trap idiots like DrLoser or Oldfart than say I was a Lier /Fraud. Causes morons like DrLoser or Oldfart to blindly attack me proving they don’t really know the topic.”

    I am only pretending to be a liar and idiot – Nice try Sir, but no cigar.

  135. oiaohm wrote, “this basically would say to the judge that I am not here because I don’t want to follow the law but I was just not sure what I should have been doing.”

    Unlike M$’s licence, GPLv2 is rather simple. If you incorporate GPLv2 code and distribute it you must also distribute the source code. VMware has not done that. Hellwig and others have them cold if they can show Hellwig’s code embedded in VMware’s secret stuff. VMware has no licence to distribute Hellwig’s stuff as non-Free software.

    Hellwig is the tip of the iceberg. If FSC can convince other Linux authours to join the suit the evidence will be overwhelming. Dependence on Linux drivers is pretty convincing evidence at any level. If VMware’s argument is that all they have copied is an API why do they need Hellwig’s code? That’s lethal to VMware’s position. Why did they distribute Hellwig’s code for years while they knew he objected with how they did it? What court would allow that for violation of M$’s EULA, for instance? Linux is a complete work and distribution of any part, including drivers without compliance with GPLv2 is wrong and illegal.

  136. oiaohm says:

    Robert Pogson the fact VMWare has been releasing the code they are in a really good location to argue they were not sure of legal interoperation so they choose the side that was most protective to their share holders as they were legally required.

    Now if this case had be brought while VMWare was not releasing any of the source code it would be another matter.

    It would be moron level stupid for VMWare to choose not to follow GPL rules if the court does rule against VMWare.

    Basically VMWare will only get badly hurt if they handle the legal process badly. Really the damage to VMWare should be basically nothing win or lose as long as VMWare does not decide to fight to the point where cease and desist gets used.

    VMWare could even PR turn around a Win or a Loss. Heck if I was VMWare I would have been so tempting to be completely insane and given the attack the 50 000 to start the case so I could say publicly I am a good guy and I just need a legal ruling. Why this basically would say to the judge that I am not here because I don’t want to follow the law but I was just not sure what I should have been doing. Its a fairly sneaky way to under mine damages claims.

  137. oiaohm wrote, “Only reason I can see for VMWare holding out is question over legal option. Profit damage wise is not going to be great. GPL damages are going to be basically nothing.”

    Those statements are only true if they end up complying with GPLv2. The law of copyright can result in cease and desist order from the court which would shut down VMware. The plaintiffs do not have to accept money or any other compensation unless that is in their claims to the court. In the end VMware will have to comply and the final settlement could well be public harming VMware’s reputation if nothing else. VMware by trying to skate around GPLv2 will have shot itself in the foot. All they had to do was release the code as Xen did. Customers will be prompted to switch to KVM, Xen or even that other OS, diluting VMware’s business no matter the outcome. No one likes to deal with a company that talks out of both sides of their mouth: “We love FLOSS” and “We won’t reveal the source code”. VMware still has the option to rewrite everything to use no FLOSS but that would be very costly and likely reduce the quality of the product in the short term. The best solution is to comply with the licence of the software they use, just like everyone else.

  138. oiaohm says:

    Really how better to trap idiots like DrLoser or Oldfart than say I was a Lier /Fraud. Causes morons like DrLoser or Oldfart to blindly attack me proving they don’t really know the topic.

  139. oiaohm says:

    oldfart the reality is you never bother checking if DrLoser arguements are logically sound.

    Like in responses here there are 22 cases of DrLoser not being logically sound.

  140. oiaohm says:

    oldfart LOL I am not exactly a FOSSTARD there is no real evidence of that all.

    Basically you are off spitting out another lie about me.

  141. oldfart says:

    “But the reality is DrLoser is noise as well. Every single point DrLoser brings is normally not backed by any cites.”

    Perhaps, but Doctor Losers points are often logically argued and sound plausible, whereas many of your points should like the fantasies of a FOSStard.

    You cause your own messes sir.

  142. oldfart says:

    “So the correct answer is I am no neither parties side.”

    So you argue for the sake of argument. Now there’s a surprise.

    “Oldfart reality you have spread more lies about me than I have ever said.”

    Nope. As a self admitted liar, you have basically screwed yourself in that regard. I and others merely pointed out that fact.

  143. oiaohm says:

    Oldfart
    “TMR guys come in here with the idea GPL steals…”

    So in the end it is your speculation that VMWare will get a slap on the wrist, and having worked to excise all the toxic GPL’d code from vSphere, will simply walk into the sunset saying:

    “So long and thanks for the fish…”

    Whose side are you on exactly?

    Number the TMR guys called me a FOSS guy even that in the past I had attacked FOSS guys for anti-closed source actions without valid ground.

    The reality is I have never been for or against FOSS or for or against closed source without grounds in every case.

    So the correct answer is I am no neither parties side. Also long as everything VMWare is doing is legal they can keep on doing it. If they get out of using GPL by the legal method that is perfectly acceptable to me.

    Mind you VMWare may be still using the GPL code for drivers in future except now with proper provable independent work status. This would also be fine to me as well.

    Does VMWare have to excise all the GPL code to get legal. Most likely no.

    For someone so called in the industry you lack the basic skill of neutrality.

    Oldfart reality you have spread more lies about me than I have ever said.

  144. oiaohm says:

    oldfart calling me noise fine I have already accepted that.

    But the reality is DrLoser is noise as well. Every single point DrLoser brings is normally not backed by any cites.

  145. oldfart says:

    “TMR guys come in here with the idea GPL steals…”

    So in the end it is your speculation that VMWare will get a slap on the wrist, and having worked to excise all the toxic GPL’d code from vSphere, will simply walk into the sunset saying:

    “So long and thanks for the fish…”

    Whose side are you on exactly?

  146. oldfart says:

    “oldfart by the way when you were making up crap about ESXi console I brought cites in here proving you were telling lies. So if I am lieing present the cites proving I am liering otherwise shut up.”

    It doesn’t work that way. I am not shutting up any more than you are. As far as cites are concerned, I leave them as an exercise for you just as you leave finding your cites as an exercise for me.

    Whats sauce for the goose…

  147. oldfart says:

    “oldfart basically since you want to call me a lier and fraud I no longer have no reason to bother attempting to provide cites or defend self because you will call those lies as well.”

    Fair enough. But then again neither do I or anyone else have any reason to take you seriously. You just become noise.

    Can you live with that sir?

  148. oiaohm says:

    TMR guys come in here with the idea GPL steals.

    Reality due to re-licensing it does not. In fact project that have gone from open source to closed are very hard to find the source code to 10 years down the track. Why when the re-licensing completes and they are no longer shipping the binaries that required the GPL source they are able to get rid of providing it completely and mirrors of code don’t last for ever either.

    http://sourceforge.net/projects/rewind/ This is a classic example from the Wine project. We know the address use to be the MIT licensed fork for Wine it self. Look what is there now. 10 years later and the open source code under the different license does not exist any more.

    https://www.transgaming.com/opensource
    Yes transgaming the competitor to crossover(the ones behind wine) end up changing to LGPL. So even a funded competitor can be crushed by not having the core developers and made todo the will of the core developers.

    10 to 15 years providing good customer service the code you were forced to release under open source will most likely no longer exist in an accessible location.

    Remember only original authors can make or give permission to make a closed source program using the code under GPL license as long as it no longer tainted by other GPL authors works.

    Open source projects cease to exist all the time. GPL means if you competitors get your code they also have to show you how they change it if you are one of their customers. But since the code is your code you don’t have to show your competitors.

    Something to remember about GPL. GPL only requires you to provide source code to your customers. You cannot block you customers from providing to anyone else. But this also means your customers are first to get new features.

    Reality here is GPL is not that toxic. Nothing in GPL will force you to return any income from sales of the program. Incorrectly use Visual studio and the like and result in causing a not commercial usage you technically have to return you income even support income so really give your program away for free.

    Your customer will always want updates and new features right. So breach of GPL fix the issue and use re-licensing to return to closed source production. If user wants new features they will have to by the closed source right? So you just have to out compete you competitor for a few years.

    Legal system is slow so you get 12+ months to have a stack of new features in the closed source that will not be in the open source. So keeping your completive advantage.

    Basically if GPL harms you as a software company your software has to be stagnant. Anything that results in non-commercial makes GPL look angelic. Managing a GPL issue correctly should see very little interference to operations even better in 10 years time you have a 90 percent chance that the released source code will no longer exist.

    Long term open source required commercial entity support without it the open source project will in most cases disappear.

    Yes GPL is something to be careful with but its not really the borg. There are NDA and Microsoft licenses way worse.

    To be correct GPL a borg license is kinda correct when you remember people like caption Picard were borg infected and after a complex operation got themselves borg free. So calling GPL borg like could be correct. But borg don’t allow individual profit making. GPL does allow individual profit making.

    The ways GPL does not block profit making when you understand copyright is so cool is not funny.

    I really don’t see how VMWare will be harmed at all having to release ESXi kernel open source. Its not the VMWare tools. If the user wants the user interface of VMWare they will still have to pay VMWare.

    Only reason I can see for VMWare holding out is question over legal option. Profit damage wise is not going to be great. GPL damages are going to be basically nothing.

    Really the end result of the VMWare case will mostly be that GPL is not as harmful as most think. Yes GPL can be inconvenient but that is about it.

  149. oiaohm says:

    oldfart by the way when you were making up crap about ESXi console I brought cites in here proving you were telling lies. So if I am lieing present the cites proving I am liering otherwise shut up.

    Since I am winning in cases that I believe you are lieing to put forwards cites the mirror applies.

  150. oiaohm says:

    oldfart basically since you want to call me a lier and fraud I no longer have no reason to bother attempting to provide cites or defend self because you will call those lies as well.

  151. oiaohm says:

    oldfart in fact since I am a lier and fraud its up to you to put forward the arguement why the cite is invalid.

    All you prove to me repeatability is you will challenge valid cites anyhow.

    I have kept my word about not giving DrLoser cites.

    The reality is I have to defend nothing.

    By the way Oldfart everyone who I have used a insult name against has called me oiaham or something else first. Not once have I been insulting to someone who has not been insulting to me first.

  152. oldfart says:

    “http://stackoverflow.com/questions/5419923/can-gpl-be-re-licensed”

    And this is authoritative why sir?

  153. oldfart says:

    “DrLoser you are still not entitled to ask me for cites. You are still calling me fifi. As long keep that up I will not be giving you cites. Cites will be only given when I see it as benefit to other readers.”

    You still don’t get it. Without cites you have nothing but your word. and as a demonstrated and self admitted liar and fraud your work is worth nothing.

    As far as you new name is concerned, you have earned it by thinking that you are entitled to call people names without consequence.

  154. oiaohm says:

    http://stackoverflow.com/questions/5419923/can-gpl-be-re-licensed

    For those interested re-licensing is covered here. This is the section of copyright law people who claim GPL stole their work want to forget. GPL only takes your work if you are using someone else work in the first place.

  155. oiaohm says:

    DrLoser
    One single cite on what the Software Conservancy evidently believes is a legally important contravention of the GPLv2 as regards SCSI use in VMware’s ESXi product, please.
    This is what I need as normal I thought you were being incompetent.

    The mark out of Christoph Hellwig work only is required to prove Co-Author status. If the section of work taken by VMWare did not contain a single line of Christoph Hellwig work he would not be able to enforce the license.

    The arguement with VMWare is about if the VMK API in fact provides enough separation so that ESXi kernel and the VMKlinux kernel are two separate works.

    VMWare believes that the VMK API is enough separation, Christoph Hellwig and Software Freedom Conservancy believe not. Also due the the ammount of strait pass threw you do have to question.

    This is a case we need a legal ruling. The prior cases of people wrapping GPL works behind libraries suggest what VMWare has done will not hold ground.

    Everything in yellow in relationship to everything in red in the diagram is what the case is about. Is everything in yellow and red independent works or are they derivative works. If everything in red and yellow is derivative works VMWare is legally required to release the source code.

    DrLoser you are still not entitled to ask me for cites. You are still calling me fifi. As long keep that up I will not be giving you cites. Cites will be only given when I see it as benefit to other readers.

  156. DrLoser says:

    In fact I will open that up to everybody.

    One single cite on what the Software Conservancy evidently believes is a legally important contravention of the GPLv2 as regards SCSI use in VMware’s ESXi product, please.

    Pace the mental dwarf oiaohm, I am committed to making my own observations on the subject, with or without assistance.

    However, it would be awfully polite of all the Bar Room Lawyers out there, if only a single one could express the slightest opinion on the subject. Possibly even with reservations. I don’t profess certainty, and I don’t expect it of anybody else.

    Oh, and, Dougie? You’re excused. Class finishes early for you. (I’m sure you heard this quite often in your teenage years.)

    VMware wants their free cake, but refuses to give back to society..booooooo.

    It’s amazing that anybody ever bothers bringing any sort of case to court, isn’t it? If you live in a state of child-like entitlement like Dougie, anything that takes more than ten seconds to figure out is just not worth the effort, is it?

  157. DrLoser says:

    Put yours up I will see how many mistakes you have made.

    That is the most pathetic and wimpy little “challenge” I have ever heard in my life, you contemptible little dweeb.

    All you ever do is to wait for a proposition and then randomly google it, isn’t that right? You are a feeb and a drone and a complete waste of space.

    This one time, oiaohm, this one time, I have given you the opportunity to construct an argument based upon the ESXi malfeasance regarding SCSI device drivers and possibly Linux stacks.

    And you are failing miserably to stand up for your side!

    You know what the really sad thing about all this is? I was, am, and will be willing to figure out some sort of position on Nitwit vs SCSI ripoffs. It’s tricky, but I’m prepared to put the hard yakka in. I was, am, and will be prepared to do so, regardless of your infantile challenges. I already know that anything you claim is going to be specious. I’ve been watching you do that for the best part of eight years.

    You, Fifi? Not. You just sit on your fat bottom waiting for somebody, anybody at all, to come up with a proposition.

    And then you make your usual asinine attempt to Gish Gallop your way out of it.

    I’m minded to suggest that you are a worthless, self-entitled, ignorant fool and despicable time-waster … but of course you can always prove me wrong with a single cite on the SCSI thing.

  158. DrLoser says:

    DrLoser by the way there are historic examples of open source projects being relicensed closed source. Just like there are examples going the other way.

    The second is trivial. The first is uncited.

    More pointless imaginary faff from Fifi, I think.

  159. oiaohm says:

    DrLoser
    It’s an honest question. And when I bring my analysis to bear, you can’t say you weren’t warned.
    Put yours up I will see how many mistakes you have made. That blog entry you pulled was so bad its not.

    Really with your skills at doing research and miss reading like claiming I said revoke when I did not. Its going to be great fun to see how much you have balls it up.

  160. oiaohm says:

    DrLoser by the way there are historic examples of open source projects being relicensed closed source. Just like there are examples going the other way.

  161. oiaohm says:

    DrLoser I am not using the word revoke. Please find where I used the word revoke or revoked other than here.
    http://mrpogson.com/2015/03/05/hits-the-fan-in-gpl-v-vmware/#comment-261306

    So I could release a program/book/what ever copyrighted under a million different licenses and it would be legal as long as I was the author/producer of the work and the work is true independent work. The way out of GPL is have agreements from all authors of the code you are using and provable independent work status. Same method is used when open source projects change license.
    I said the way out of. Not that the license would be revoked. There is a method that VMWare will not have to obey the terms of GPL forever.

    Reassignment allows the future work to be a different license to its prior forms I lay out the rules here. Of course if someone in this case wish to use the prior form without the new features under the old license they can.

    When the wine project changed from MIT to LGPL the project Rewind was formed and died. Lack of resources will see you run into the ground.

    As rewind with Wine demonstrates without the skilled coders a complex code base will become worthless. 3-4 years down the track the ESXi code of today will be fairly much worthless. So as long as VMWare can get provable independent stats quickly its not a major problem. Note VMWare will only have to release its source code up until the date independent work status happens. So the source code the courts force VMware to release if VMWare does the right things now could be 12 months+ old.

    But we were talking about the opportunity for an invidual to revoke their copyright, once assigned to the Impersonal Borg That Is The GPL. It’s not possible, is it?
    For VMWare no its not really going to be possible. Please note wine project could not revoke the MIT grant either. This is the catch with copyright if the grant is legal revoking it is basically impossible. Doing a process of relicensing so you don’t have the follow the terms of the old license in future is possible.

    The revoking options this applies to all OSI licenses and most copyrights(exception being contract work were doing following can be blocked by the contract you signed. Signing of CLA and NDA come to mind).
    1) Claim copyright infringement that someone licensed the work under the License without formal permission todo so.
    2) Prove unsound mind at time you released the source code.
    At that point the license can be revoked and then the revoked status has to be published.

    Basically all the same stunt you would use to get out of having to keep a contract. Its not that GPL cannot be revoked the price of doing so is quite nasty.

    Please note ideas of 90 trials and so on in most Civil courts don’t hold grounds. The majority of revoking clauses in copyrights when taken to court turn out not to hold.

    Microsoft FUD pushed the idea that there is no way out of GPL. Completely disregards the legal process of relicensing. Yes relicensing does not revoke the prior license just makes all future works not have to obey the terms of the prior license.

    The way out of GPL is legally relicense in most cases. Relicensing require approval of the Authors and you have independent work status.

  162. DrLoser says:

    Anybody got a clue on GPL copyright infringement via VMware usage of SCSI code?

    It’s an honest question. And when I bring my analysis to bear, you can’t say you weren’t warned.

    Go to it!

  163. DrLoser says:

    Consonant with my policy of completely ignoring any babble whatsoever after you have made your first cretinous argument, Fifi:

    DrLoser copyrights licenses are not allowed revoking clauses in most cases.

    Except that your original claim was that an individual is entitled to revoke his or her copyright license. You remember? Books and so on.

    Not being a lawyer, I have no opinion on revocation of copyright licenses in general. I do have an opinion on doing so as a German Citizen, and I will here repeat the basis for that opinion:

    Das Urheberrecht kann in Erfüllung einer Verfügung von Todes wegen oder an Miterben im Wege der Erbauseinandersetzung übertragen werden. Im übrigen ist es nicht übertragbar.

    But we were talking about the opportunity for an invidual to revoke their copyright, once assigned to the Impersonal Borg That Is The GPL. It’s not possible, is it?

    It isn’t so much “revoking copyright,” which is a strange concept. It’s more “reassigning copyright to the original author.”

    Not really possible under the GPL, as far as I can see. And intentionally so, for good or bad.

  164. oiaohm says:

    First of all, oiaohm is talking about the potential copyright infringer here, which wasn’t actually the point.

    So was I. If as a potential copyright infringer you can prove independent work status you will win. VMWare might win that point. But from every thing we can see the odds of winning that is low.

    DrLoser copyrights licenses are not allowed revoking clauses in most cases. You can change from a particular license for future works. This is not a GPL thing.

    Not necessarily. In fact, there’s a probable case for them to be found in default for Exhibit One, as an example. But that doesn’t mean much in terms of the overall plaint: I’d expect the Court to make a specific order for damages and restitution, re radix_tree.c.
    German law/Civil Law comes into play here you are still thinking USA law.

    Radix_tree and the scsi are only in fact require prove that he is one the combined work Authors as long as he can prove that he can legally protect the all the other parts the work unless VMWare can declare Author ownerships.

    Remember to protect the complete project under German law he only had to have 1 line of source code in it to protect the complete work. He has way more so he has standing to protect the complete GPL work.

    The case will be resolved over the combined work not individual parts unless VMWare can prove they Authored all the other parts(ya right thats not going to happen might be able to mitigate a little but its not going to be complete).

    This is why with Open Source in Germany the damages are more likely to have to be paid in source code since source code is a more fair payment to all Authors involved. VMWare loses badly they could be forced to open source the complete product. The extra area code being payment of damages. Basically the real worst case we don’t need a second case of. There is a prior open source case where damages was paid in source code.

    And which part of the Visual Studio Community EULA suggests that “the generate files are not to be used for derivative work actions,” Fifi?

    My point is there is not a clause.

    1 there is no clause in the main license to separate the generated files from the main license. 2 the generated files don’t have a license inserted on them saying they are not licensed by the main license.

    No matter how obvious it is that a multi-million dollar closed-source company would be using a completely different license for Visual Studio.
    DrLoser remember is not the version of Visual Studio you are running allown but the version branded in the files as well. Just because you are running Paid for Visual Studio does not allow you to break the license of a GPL work right. Same applies for a Paid for Visual Studio using a Visual Studio Community generated file.

    Multi-million dollar closed-source company is normally aware they have to perform licensing auditing including looking for the evil I am describing as they absolutely don’t want to lose their investments.

    The big trap with Visual Studio over other things is the UUIDs. A generator making stuff directly from your work the output by copyright is normally yours. The exception is when it contains other content you did not create that is not just like formating. UUID are badly enough actively generated.

    http://www.gnu.org/licenses/gpl-faq.html#GPLOutput
    The reality here is you have entered a Fair Usage arguement you can lose if generated are not tagged as independent. Interesting point here is the example.

    http://www.gnu.org/software/bison/manual/html_node/Conditions.html
    Bison these days explicitly stats that the generated is not covered. This is what all open source generators using licenses that may force some nasty term on a closed source has done. This takes it out of Judge/Jury to rule fair usage and into a legal certainty.

    The legally safe presume is the products license covered the generated unless properly stated otherwise or tested in court. MS Office contains a lot of pointless macros that it auto inserts into your work. Are those macros in fact generated from your work.

    Please note using fair usage does not mean you have not committed copyright infringement it means you have had to use a legal excuse equal to the dog ate my homework and hope the counts agree. When you need to have the courts not to agree to fair usage is a bad run of judges/juries..

  165. DrLoser says:

    To quote the wonderful Cuba Gooding Jr, Robert:

    Show me the violations!

    On one of your excellent and unimpeachable graphs, please. A logarithmic graph would do awfully well should you choose to defend the absurd proposition of “many orders of magnitude.”

    I feel fairly confident that you are not going to do that, are you? Not through any failing of your own.

    Let’s face it, Robert.

    The evidence is not there.

  166. DrLoser says:

    Assuming there is some relation between use, use-cases and application in business/commerce, there have been many orders of magnitude increase in opportunities to use FLOSS without respecting GPL or other FLOSS licences.

    And assuming there is very little relation, there have not been.

    Perhaps you could consult StatCounter, Robert?

    Hard Cold Facts are always useful in these arguments, I feel.

  167. DrLoser says:

    You do realize that 20 years ago, there was mostly GNU and darkness roamed over the face of the deep. Then came Linux and the use of FLOSS grew from ~100 buddies of Linus…

    Yes I do, Robert. All very sweet. Practically a Hollywood Frat Party, isn’t it?

    Also, totally irrelevant.

    The prevalence and sheer volume of GPL violations has increased by many orders of magnitude in the nearly two decades that I have worked on enforcement of the GPL.

    The Magnificent Mr Kuhn is not referring to the exponential growth of Linux.

    He is referring to some sort of exponential growth of GPL violations, which sadly exists only in his strangely-shaped, yet very distinguished in an obvious technological way, head.

    I’m slightly disturbed that you can’t tell the difference on a cursory examination of the quote, Robert. Tell me, have you consulted an Eminent Medical Practitioner (one would always recommend one who is self-described as such, possibly from Loyola, possibly not) as regards your monumental Confirmation Bias?

    It’ll make you walk with a limp, I tell you. Thinking with a limp, well, let’s face it, that’s pretty much a given in your case.

  168. DrLoser quoth, “by many orders of magnitude in the nearly two decades”.

    You do realize that 20 years ago, there was mostly GNU and darkness roamed over the face of the deep. Then came Linux and the use of FLOSS grew from ~100 buddies of Linus working on the Linux project until today when a billion or more are using */Linux and it is widely used commercially. Assuming there is some relation between use, use-cases and application in business/commerce, there have been many orders of magnitude increase in opportunities to use FLOSS without respecting GPL or other FLOSS licences.

  169. DrLoser says:

    No. I wish to be fair to the Great Kuhn, Supreme Mongol Leader!

    Working with Base 2, I suppose that a progression from zero to, say, 1024 is “many orders of magnitude.”

    But it’s not “magnitude as we know it,” is it, Captain?

  170. DrLoser says:

    Apparently there are dismally few self-styled “Distinguished Technologists” who can understand the mathematical meaning of “many orders of magnitude,” incidentally.

  171. DrLoser says:

    I don’t suppose that anybody read my cite from Glyn Moody, because you’re all intolerably lazy, so I’ll just add a couple of grace-notes:

    Commenting generally on the issue of GPL enforcement, Bradley M. Kuhn, President and Distinguished Technologist of Conservancy, stated: The prevalence and sheer volume of GPL violations has increased by many orders of magnitude in the nearly two decades that I have worked on enforcement of the GPL. We must make a stand to show that individual developers and software freedom enthusiasts wish to uphold copyleft as a good strategy to achieve more access to source code and the right to modify, improve and share that source code.

    In other words, the particular legal niceties of this case (I have referred to them as Exhibits One, Two, Three and Four) are to be completely disregarded. Apparently there is no legal difference in Distinguished Fat-Head’s opinion between VMware and, say, a BusyBox infringement.

    Regrettably, and fairly obviously, there is.

    In other words, if an example is not made of VMware and its intransigence in this case, the danger is that GPL violations will multiply uncontrollable, undermining the very foundation of free software.

    Well now. That depends upon whether VMware have a reasonable argument for “intransigence,” doesn’t it? It has nothing at all to do with other GPL violations, whether or not they might “multiply incontrollable.”

    I’m trying my best to analyse this case on its own merits, and to do so analytically. I’ve been avoiding the sort of nonce diversions so beloved of Fifi and Robert.

    I’ve tried to argue Exhibits One, Two and Three. I have yet to argue Exhibit Four. All four, as far as I can see from the Software Conservancy diagram, comprise the case.

    Does this mean that I think VMware are right on any of the four “Exhibits?”

    Not necessarily. In fact, there’s a probable case for them to be found in default for Exhibit One, as an example. But that doesn’t mean much in terms of the overall plaint: I’d expect the Court to make a specific order for damages and restitution, re radix_tree.c.

    Let’s not get too frothy here. This is obviously a stupid case for the Software Conservancy to bring, because they can’t win very much at all. They certainly won’t win control of the ESXi source code, for example. Nor will they put VMware out of business, thus benefitting whichever shady corporations are selling Xen and other VM solutions such as, er, VirtualBox (hello Oracle!).

    Nothing at all personal, because I really don’t care. (Apart from the stupefying lack of real argument on this site.)

    But, in all honesty, if I were a FLOSS evangelist, and had I been asked …

    … Having looked into the detail, I would have said — Leave the Nitwit Hollwig to his own legal doom!

  172. DrLoser says:

    Round-up call for comments on ESXi and abuse of Linux SCSI components, incidentally.

    All comments welcome. As I say, it’s murky, and I appreciate any and all views.

    Perhaps Fifi would care to step out from behind the Visual Studio EULA lamp-light and actually get down to brass tacks?

    (Storage Area Networks, Fifi, Storage Area Networks!)

  173. DrLoser says:

    One last piece of blather to point out.

    This is the reality that there is something out their way nastier than GPL. Microsoft licensing on IDE is not like using QT Designer,Eclipse , Kdevelop, Netbeans…. Why all these other things include a declare in license that the generate files are not to be used for derivative work actions.

    And which part of the Visual Studio Community EULA suggests that “the generate files are not to be used for derivative work actions,” Fifi?

    Not a single one. I can confidently assert that this is never going to happen.

    Now, if you wish to continue your peculiar fetish of defending multi-million dollar closed-source companies against the iniquities of the Microsoft Visual Studio Community License, I can’t stop you.

    No matter how obvious it is that a multi-million dollar closed-source company would be using a completely different license for Visual Studio.

    And perhaps I shouldn’t even try to stop you. What’s it to me?

    If you insist on sounding like an ignorant imbecile, I’m not the one to stand in your way.

  174. DrLoser says:

    Because its copyright contract you don’t need a revocation clause.

    A author of a work is legally allowed to license it under as many licenses as they like as long as it does not infringe on some else work.

    So I could release a program/book/what ever copyrighted under a million different licenses and it would be legal as long as I was the author/producer of the work and the work is true independent work. The way out of GPL is have agreements from all authors of the code you are using and provable independent work status. Same method is used when open source projects change license.

    This particular oiaohm argument is, however, both sensible and very much relevant to the current OP. It’s a nice theory (although I suspect that oiaohm dreamed it up on the spot), but I’m not convinced that it has any basis at all in reality. The relevant statement is here:

    The way out of GPL is have agreements from all authors of the code you are using and provable independent work status.

    First of all, oiaohm is talking about the potential copyright infringer here, which wasn’t actually the point. The point I am making is that I cannot see a way out (other than edge-cases like German Copyright Law, cited lower down, in specifics, in German, and I cited the relevant clause, in German) for Person X to assign a GPL License to his “work” (probably code, but let’s not limit) and then to revoke it. Once it’s out, it’s out.

    Now, Fifi’s suggestion that a potential GPL copyright infringer can talk to everybody in the universe and get them to agree is patently absurd. Insofar as it is any freedom at all, it is clearly not one that will ever happen.

    But oiaohm’s suggestion:

    Because its copyright contract you don’t need a revocation clause…

    … seems to be the main reason why nobody in their right minds would sign up to the GPL.

    “Because it’s copyright contract” for whatever copyright clauses you choose, an individual can, indeed, choose to revoke the freedom of distribution granted under that copyright.

    I’m not at all sure you can do that with the GPL, though. You’ve basically castrated yourself. The moment somebody “forks” your work, and by “forking” I include “copying bitwise from one medium to another,” your original “copyright” assertion is no longer revocable.

  175. DrLoser says:

    DrLoser the sad reality is GPL did not force the bankrupt.

    I never said it did, Fifi. I merely pointed out that, of the four, it was the one already due for bankruptcy.

    The rest of your idiot post is therefore forfeit. I am not going to waste my time on any individual post of yours that starts with an evident cretinism.

    To be fair, though, I’m quite happy to address your other posts.

  176. oiaohm says:

    Westinghouse Digital Electronics is interesting. Because GPL does not restrict against commercial usage it was only the source not being released was the issues that case was limited to 90 000 in damages.

    DrLoser the sad reality is GPL did not force the bankrupt

    http://en.wikipedia.org/wiki/Westinghouse_Electric_%281886%29

    Westinghouse cut free any part of itself with legal troubles and let them go under. This is one of the problems with enforcing copyright. You have to attack the partent company or they will just sacrifice a sub company.

    Every time it looks like GPL has caused a company to go bankrupt if you follow it back the parent company has decided to give up a sub company than obey. GPL or any copyright enforcement rarely harms the core companies.

  177. oiaohm says:

    DrLoser
    Once you’ve signed up to any version of the GPL, it’s difficult to see how you can sign out again. I mean, there’s no revocation clause, as far as I can see.
    Because its copyright contract you don’t need a revocation clause.

    A author of a work is legally allowed to license it under as many licenses as they like as long as it does not infringe on some else work.

    So I could release a program/book/what ever copyrighted under a million different licenses and it would be legal as long as I was the author/producer of the work and the work is true independent work. The way out of GPL is have agreements from all authors of the code you are using and provable independent work status. Same method is used when open source projects change license.

    So it is possible for ESXi to in future be free of GPL restrictions by getting provable independent work status to the Linux kernel so allowing them to change license back to closed source.

    You can see GPL forcing you to release your code as punishment for not keeping independent work status. Of course the updates to that code you were forced to release as punishment can be kept secret once independent work status is restored.

  178. oiaohm says:

    I Am Enterprise! I have ~250 PC desktop seats! I earn $1 million+ per year! My business depends upon closed-source software!

    FEAR ME! For my business depends upon the Visual Studio Community License!

    You’re being just the eensy, weensiest bit of a Dick here, aren’t you, Fifi?
    No you are being the dick. You have missed the problem its not the 250 PC or the $1 million+ per year income mostly. It is the 250 users that is the problem.

    All generated files by Visual Studio Community and Visual Studio Express…. are under the license of the product. So you license of a open source project code can turn out to be OSI license + Visual Studio license what ever version.

    More than 1 developer(this could happen from using a open source work) and more than 250 sales and you are screwed as a closed source developer. Now if I have modified one of those Visual Studio generated files I can become a co author so enforce the terms of the Visual Studio license on you. Now if I am super creative and mix the right versions of Visual Studio into my open source program you can generate something so bad its not funny. So a BSD license project with Visual Studio generated files can result in nasty no-commercial usage absolutely if you build it with Visual Studio or anything that uses the Visual Studio generated files.

    Basically Visual Studio Generated files should be feared due to Viral license. So if you fear GPL then you so also fear Visual Studio Licenses. If you don’t you are a bias little bugger who does not know what they are dealing with.

    This is the reality that there is something out their way nastier than GPL. Microsoft licensing on IDE is not like using QT Designer,Eclipse , Kdevelop, Netbeans…. Why all these other things include a declare in license that the generate files are not to be used for derivative work actions.

    LSILogic, BUSLogic & PVSCSI all three are Linux provide you find the same items used by KVM and Xen to their virtual machines. Yes vmware windows SCSI drivers happen to work with KVM and Xen because the underside is basically the same.

    Even stuff like vmdk file reading/writing support is part of the Linux kernel.
    http://www.cyborgworkshop.org/2014/07/05/mount-a-vmdk-in-linux-without-having-vmware-installed/ So vmdk support limits are the same on ESXi. KVM and Xen why they are all using the same open source code that is part of Linux. The raw device pass threw limits again straightly match Linux.

    DrLoser everything in that blog link about scsi if you new what you were looking at suggests that VMWare is using anything other than Open Source for the scsi support.

    Busybox is still used in ESXi for console and for starting the services of ESXi. Busybox provides ESXi equal to init, bash+ more. Basically remove Busybox and ESXi cannot boot without some major alterations.

    DrLoser German counts decide max damages for the accuser in advance. In this case its 100 000 and that money has to be in trust before they can bring the action to court. That is the Max vmware and the count can get if they win to cover costs. Issue here is 20 000 will go in court costs. So its not 100 000 for VMWare its only 80 000. VMWare might spend more in legal.

    Yes it s a civil court system the loser has to pay for the courts time as well as plaintiff.

    Yes 80 000 is the max no matter how much VMWare or the Open Source guys drags it out. So even if VMWare wins by money they can still lose.

    Please note Oldfart said he wants VMWare to lose. If VMWare loses the taint from Visual Studio becomes enforceable. Any sane commercial developer should want VMWare to win at least partly.

  179. DrLoser says:

    Never fear, though, Pog. I know how much all you blood-sucking drones hate being called “cheapskates,” particularly when it comes to “giving back” to FLOSS.

    Through assiduous exploration, I have found a way for you all to do that!

    Join the Glorious Victory March of Nitwit vs VMware! Contribute! Give money! You will never have a better chance to prove that FLOSS is worth more to you than money … more than oxygen … more than Life Itself!

    Bad actors have become complacent because they think you don’t care. A strong show of public support for Conservancy and Christoph’s position will help our legal case and demonstrate the interpretive context for it. Please donate to our campaign to enforce the GPL.

    And the really spiffy thing about this Mega-Deal is that somebody else (identity unknown, but presumably not Bill Gates) is prepared to match each and every donation up to $50,000!

    That’s a cool $100,ooo for the Plaintiff!

    So, what are you waiting for, folks? Break out your pocket-books! Shake the moths off from that dusty collection of dollar bills! Give! Give! GIVE FOR FREEDOM!

    I can’t imagine how I got here first, before Dougie beat me to it.

  180. DrLoser says:

    Of course, Till Jaeger can be discounted altogether, because he doesn’t have his own Wikipedia page.

    I’d have gone with Miss Piggy, myself. Nobody disses Miss Piggy.

  181. DrLoser says:

    Whiling away the long, lonely hours whilst I wait for somebody to point out quite how dreadful the VMware violations re SCSI are, I figured I might as well look up a rather more important actor in this little soap opera. That being, Till Jaeger, who is sadly not a “Distinguished Technologist” owing to the misfortune that he is a professional lawyer, rather than a bottom-feeding marketroid.

    What Till Jaeger is, however, is that he’s the lawyer in charge of this particular case. And he has chops when it comes to defending the GPL. Well, I say that, but although he is often quoted as working on “several successful suits for GPL violations,” I can only really find the one: Harald Welte vs Skype.

    That said, the man seems to know his stuff, unlike most of the rank amateurs involved in this case. But as to that “stuff,” you’re probably not going to enjoy hearing what he has to say about the typical GPL enforcement case:

    The typical GPL infringing product does not contain any notice about the fact that GPL software is contained. Consequently, no license text and no source code are provided. Such products are the main focus of enforcement activities. Sometimes a GPL notice is provided, but the license text of the GPL and the source code are not available, or the notice points to a webpage that might contain this material or nothing at all. Although a situation where the necessary information is provided on a webpage is in general no reason for enforcement activities, one should keep in mind that the District Court of Munich has decided that the mere referral to a webpage does not comply with section 3 of the terms and conditions of the GPL (version 2).[7] This example demonstrates that careful in depth consideration of all obligations of the GPL is required for compliant use of GPL software.

    Doesn’t seem to apply here. That ISO you cited, Robert? Essentially it’s a bunch of tgz files, each one of which scrupulously includes the relevant GPLv2 or LGPL or I presume Apache or MIT license.

    Never rely on consummate professionals like Till Jaeger, Robert. They will only let you down.

  182. DrLoser says:

    He’s in Wikipedia, for pity’s sake. I think that qualifies for “distinguished”.

    Ah, the “distinguished” Miss Piggy. The “distinguished” Jeffrey Dahmer. The “distinguished” Vanilla Ice. I see your point, Robert.

    Unfortunately, not one of these Modern Greats were ever awarded the title of “Distinguished Technologist,” least of all by an organisation that they, themselves, founded.

    The man is about as “distinguished” as a non-malignant mole on my left shoulder. (To be fair to my mole, it makes no claims either way. Please do not boil it. It is not a frog.)

  183. DrLoser wrote, “No delusions of grandeur whatsoever there, then.”, in reference to Bradley M. Kuhn being entitled, “Distinguished”.

    Education: “Kuhn attended Loyola Blakefield, followed by Loyola College in Maryland, graduating in May 1995 with a summa cum laude Bachelor of Science in Computer Science.”

    Work: 15 years or so dedicated to Free Software.

    He’s in Wikipedia, for pity’s sake. I think that qualifies for “distinguished”.

  184. DrLoser says:

    Commenting generally on the issue of GPL enforcement, Bradley M. Kuhn, President and Distinguished Technologist of Conservancy, stated …

    No delusions of grandeur whatsoever there, then. I mean, “Distinguished Technologist” is hardly the sort of certification you find on the back of a bubble-gum wrapper.

    Is it, oiaohm?

  185. DrLoser says:

    Here, for example, St Glyn babbles on about BusyBox:

    Hellwig, a key Linux kernel developer and one of the earliest members of Conservancy’s GPL Compliance Project for Linux Developers, has publicly denounced VMware’s misuse of GPL-licensed code since 2007. In 2011, Conservancy discovered that VMware had failed to provide nor offer any source code for the version of BusyBox included in VMware’s ESXi products (as required by BusyBox’s license, GPLv2). Conservancy began in early 2012 negotiations with VMware to seek compliance on all GPL’d components in the ESXi project. Progress was slow through 2012 and 2013.

    Possibly because BusyBox is no longer relevant to ESXi? (If indeed it ever was.) It doesn’t seem to feature in the current case, for some reason.

    Well, we can leave the thorny question of GPL contravention vis-à-vis SCSI for the moment, unless anybody has a bright idea.

    Is there any good reason why Nitwit and his bottom-feeding cohorts at the “Software Conservancy” have chosen to wait until 2015 to bring suit?

    I mean, the Software Freedom Law Center managed to win a slam-dunk victory on the rather pathetic issue of BusyBox. And only one of them was bankrupt at the time.

    Strangely enough, that was Westinghouse Digital Electronics. Not, as I had assumed, either

    1. Monsoon Multimedia or
    2. Xterasys or
    3. High Gain Antennas

    All of whom, as far as I can tell, have been bankrupted by the proceedings and no longer exist. Still, there’s nothing like encouraging free enterprise, is there?

    It became necessary to destroy the town to save it. (Major Booris, in Vietnam.)

    That’s literally what you lot believe, isn’t it?

  186. DrLoser says:

    And it would be unfair to ignore the contribution of Glyn Moody here. Who can ignore Glyn Moody? The man is a bloviating useless windbag! But difficult to ignore.

    Nonetheless, he raises several issues that y’all might have missed. So, in the interest of fair discussion, I encourage you to build on his various theses.

    Why you lot can’t look this stuff up for yourself is beyond me.

  187. DrLoser says:

    But never let it be said that I can be diverted by hysterical laughter at the obvious stupidities of others on this thread. Not at all. I take this stuff seriously. The future of the GPL is at stake, goddamnit!

    With which in mind, I wish to start the discussion on SCSI violations off via this cite:

    This is an interesting question, and something which popped up in some recent discussions. This is nothing more than a math exercise considering that a VM can have 4 SCSI controllers and 15 devices per controller, but the numbers are still quite interesting. Let’s look at what options you have. The following is true for all the Virtual SCSI Controllers that are supported on VMs at this time (LSILogic, BUSLogic & PVSCSI).

    Now, those “virtual SCSI Controllers.”

    Any opinions on how they breach the GPL?

  188. DrLoser says:

    I Am Enterprise! I have ~250 PC desktop seats! I earn $1 million+ per year! My business depends upon closed-source software!

    FEAR ME! For my business depends upon the Visual Studio Community License!

    You’re being just the eensy, weensiest bit of a Dick here, aren’t you, Fifi?

  189. DrLoser says:

    I must say it has been truly amusing to watch Fifi come up with increasingly inventive “interpretations” of EULA in his flailing attempts to win brownie points in the argument at hand…

    To be fair to oiaohm, Robert has been at least as guilty of flailing around.

    And at least oiaohm addressed the topic of the OP several times. I’m not sure that Robert is up to that.

    SCSI, anybody?

  190. DrLoser says:

    Everyone else is screwed by MS piss poor define of enterprise.

    Once again, Fifi:

    If none of the above apply, and you are also not an enterprise (defined below), then up to 5 of your individual users can use the software concurrently to develop and test your applications.

    Don’t blame me. It’s your cite.

  191. DrLoser says:

    Well you certainly took your time getting here Dottore.

    It’s a fifteen day old thread, oldfart. I am a patient man. I like to build my arguments one step at a time. Exhibit One, Exhibit Two, Exhibit Three and so on.

    Seems a long while since we began this cozy little fireside chat with the Dougster accusing a large US corporation of behaving in the same rather disgusting way that he does, though, doesn’t it?

    VMware wants their free cake, but refuses to give back to society..booooooo.

    Perhaps everybody’s favourite uneducated idiot is prepared to jump in at this point and explain the rather complicated details behind the Software Conservancy’s plaint on SCSI?

    I darned well hope so. Because, of all of this mess, I’ll confess that the SCSI bit makes no sense whatsoever to me. It might very well be the linch-pin of the case … but it’s murky enough, and I am desperate enough, that I feel compelled to turn to an under-educated snake-oil salesman to explain the details to me.

    How about it, Dougie?

  192. DrLoser says:

    So about the only people who can use the Visual Studio Community license is the academics, OSI developers and true solo developers. Everyone else is screwed by MS piss poor define of enterprise.

    And what have the Romans ever do for us?

    Imbecile.

  193. oldfart says:

    “You are a canting worthless imbecile, oiaohm. There are far more important possible license violations out there. Go get them and stop being a worm.”

    Well you certainly took your time getting here Dottore. I must say it has been truly amusing to watch Fifi come up with increasingly inventive “interpretations” of EULA in his flailing attempts to win brownie points in the argument at hand, not to mention how he kept ascribing actions to me even after repeated reminders that I had no intention of following his improbable path.

    Once again our Fifi demonstrates his credentials: Proven and self admitted liar and and fraud, and master of the imbecilic interpretation.

    Its all good.

  194. DrLoser says:

    Saying something is your own when it someone else trademark is attempting to steal the trademark.

    Bollocks.

    If I used Visual Studio to examine/modify/improve/distribute my Open Source Software, this is precisely what I would claim in order to comply with the Microsoft Visual Studio EULA. And the Microsoft Visual Studio EULA is designed to let me do so.

    You are a canting worthless imbecile, oiaohm. There are far more important possible license violations out there. Go get them and stop being a worm.

    F’rinstance. SCSI and Nitwit vs VMware. As I say, it seems to involve Storage Area Networks — your specialty.

    Your thoughts, please? I mean, I hate to drag this damn thread back on topic. But the topic is Nitwit vs VMware.

  195. DrLoser says:

    Nope. The authours have at least one capability you don’t. They can choose to change or to maintain the licence.

    Leaving aside my earlier quote of German Law (UrhG), at which point I mentioned that, in some jurisdictions, you are not actually allowed to give copyright away, this seems to me to be a bit of a stretch, Robert. Even by your standards.

    Once you’ve signed up to any version of the GPL, it’s difficult to see how you can sign out again. I mean, there’s no revocation clause, as far as I can see.

    It would be a fascinating legal battle, though.

    Stallman: I now wish to reclaim my rights to Emacs, Your Honour. It is a very popular piece of software, and I belatedly realise that I could be rich beyond my wildest dreams if I made it closed-source.
    Impartial Judge: Well, you can’t, can you? Emacs is licensed under the GPL.
    Stallman: Temporarily, yes. I never meant for anybody to take any of that bollocks seriously. I’m sick and tired of living in an MIT broom-cupboard. I want to live in a penthouse apartment overlooking Park Avenue.
    Impartial Judge: Unfortunately, you can’t. Copyright is copyright. You should have thought of that earlier.
    Stallman: I hereby cite Robert Pogson, Your Honour. He is an Expert on business practises, legal affairs, and copyright infringements. In his spare time, he boils frogs for fun.
    Impartial Judge: Is that relevant?
    Stallman: Not really. I’m totally against boiling frogs, except when they won’t have nasal sex with me.
    Impartial Judge: No, I meant, was any of the rest of that relevant?
    Stallman: It seems to be the only way I can get my copyrights back. Robert says I can, so I’m going with Robert on this one.
    Impartial Judge: You’re even more of a fool than I thought you were, then. Next case please!

  196. oiaohm says:

    Gnu emacs (to take but one example) is just as much “my own application” as it is anybody else’s … providing, of course, that I abide by the license.
    DrLoser you did not get what I said did you so I have to be blunt.

    Saying something is your own when it someone else trademark is attempting to steal the trademark. Only the owner can claim they own it. GNU Emacs the only ones that can claim to as their own application is GNU Project. DrLoser you are not the GNU Project.

    DrLoser I am not the moron on this. You are. Trademarks be very careful using the word own. Using the word own around a trademark incorrectly is a very quick way to land in court if the wrong people hear you.

  197. oiaohm says:

    DrLoser I state the community version for a reason. Visual studio comes in Community and Express. Yes different license traps. Currently not sitting at a Windows machine.

    Microsoft Visual Studio Solution File, Format Version 11.00
    # Visual C++ Express 2010

    Also you are kinda being blind.

    There is other license declare in the file. Only 1 line declaring what made it. Yes the # at the start of line is a comment. This is also the declare of what License applies.

    Since there is no license declare you have to expect the license of the creating applications for License grant. Yes “# Visual C++ Express 2010” is the down right license declare. 1 single simple to over look line.

    If you cannot find a copyright notice or matching License you legally don’t have the right to distribute the work.

    There is no legal requirement to have the word License or Copyright in a copyright declare only that it be present.

  198. DrLoser says:

    Basically DrLoser don’t say something like this again. Because you just suggested someone steals a trademark.

    No I did not, you moron.

  199. DrLoser says:

    DrLoser “Your Own Applications” was defined in a court case over game engine licensing. It has a legal precedent behind it. Same with “Your Application” when referring to a Organization. There is no wiggle room here.

    No wiggle room? In that case, you are honour-bound to cite the case, in order to prove that you are not simply the fantasising liar that oldfart mysteriously seems to believe that you are.

  200. DrLoser says:

    DrLoser you failed to understand the problem. The MSBuild file itself includes the License declare.

    Feel free to quote one, Fifi, but you are sadly out of date. (Not that it matters, because I’ve already taken the pains to explain to you how the Visual Studio EULA works.) These days, it doesn’t.

    To take a scrupulously fair approach, I have just built “Hello world” in Visual Studio C++ 2010 (the “Express” edition). Here is the sln file:

    Microsoft Visual Studio Solution File, Format Version 11.00
    # Visual C++ Express 2010
    Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "Hello", "Hello\Hello.vcxproj", "{E5CAE039-5689-42DA-9C0E-8CA2F659017F}"
    EndProject
    Global
    GlobalSection(SolutionConfigurationPlatforms) = preSolution
    Debug|Win32 = Debug|Win32
    Release|Win32 = Release|Win32
    EndGlobalSection
    GlobalSection(ProjectConfigurationPlatforms) = postSolution
    {E5CAE039-5689-42DA-9C0E-8CA2F659017F}.Debug|Win32.ActiveCfg = D
    ebug|Win32
    {E5CAE039-5689-42DA-9C0E-8CA2F659017F}.Debug|Win32.Build.0 = Debug|Win32
    {E5CAE039-5689-42DA-9C0E-8CA2F659017F}.Release|Win32.ActiveCfg =
    Release|Win32
    {E5CAE039-5689-42DA-9C0E-8CA2F659017F}.Release|Win32.Build.0 = Release|Win32
    EndGlobalSection
    GlobalSection(SolutionProperties) = preSolution
    HideSolutionNode = FALSE
    EndGlobalSection
    EndGlobal

    No EULA visible, I see.

    I then built the app, as follows:

    msbuild Hello.sln /p:Configuration=Debug /p:Platform=”Win32″

    No EULA visible, I see.

    At this point I would advise a normal human being not to believe everything they dredge up on Google. But then, that would be futile in your case, wouldn’t it, Fifi?

    So using MSBuild file with another complier/IDE suite still sees you hit with the same taint problem.

    There isn’t a “taint” problem. “Taint” problems are far more frequent in FLOSS. But even if there were a “taint” problem, all I can say, Fifi, is “t’aint so.”

    Great fun have to remake your own build file.

    BWAHAHAHAHAHA!

    That’s so very difficult when you’ve got gmake available, isn’t it, Fifi?

    You’ve never once written a make file in your life, have you?

  201. oiaohm says:

    DrLoser “Your Own Applications” was defined in a court case over game engine licensing. It has a legal precedent behind it. Same with “Your Application” when referring to a Organization. There is no wiggle room here.

    Gnu emacs by Organizational own-ship define own to GNU project. You are just a user/developer of it. GNU can only appear in the name of a project owning to the GNU project. Why GNU is a trademark of the GNU project.

    Gnu emacs (to take but one example) is just as much “my own application” as it is anybody else’s … providing, of course, that I abide by the license.
    Basically DrLoser don’t say something like this again. Because you just suggested someone steals a trademark. Its not your own application at all. FOSS projects are copyright and trademarks.

    Welcome to the trap FOSS Projects truly do have owners and they are not you unless you do a particular things.

    If you fork GNU Emacs under a new name then its yours kind of. See how this could get horrible mess for a solo developer.

  202. DrLoser says:

    Various yard sales???

    And, why specify “private”? Governments are some of the largest businesses around. Yes, they do have budgets and shareholders. They have bureaucracy too.

    I specify “private” for the same reason that I emphasise “customers actually stumping up their own cash for your product,” Robert. That would seem to me to be the core requirement for business experience.

    However, I accept your experience with

    a. Versatile Farm Equipment, a division of Ford, back in the 1980s
    b. Quality Bed and Spring co, back in the 1960s

    Those there qualify as perfectly reasonable sources of “business experience.”

    A tad limited, when it comes to producing commercial software in the 21st century, but you do have a solid basis to work from.

    Which is interesting, really. Looking back on your business career, how much do you feel that Quality Bed and Springs in the 1960s and Versatile Farm Equipment in the 1980s would have benefitted from FLOSS?

    (Yup, obviously, Gnu wasn’t really a “thing” in the 1960s, but it’s still interesting to speculate. And there were all sorts of “free software” things going on in the 1980s, so a division of Ford might reasonably be assumed to take advantage of the goodness.)

  203. DrLoser wrote, “Gnu emacs (to take but one example) is just as much “my own application” as it is anybody else’s … providing, of course, that I abide by the license.”

    Nope. The authours have at least one capability you don’t. They can choose to change or to maintain the licence. They can even set up a dual licence or deny any access to the code. That won’t happen, of course. That would be pointless.

  204. DrLoser says:

    Notice the words “your own applications”. This means anything from a OSI project that is not 100 percent your own work is not covered by the Individual License.

    A judge would laugh that one out of court, Fifi. Not that Microsoft would ever bring it to court.

    I don’t think you quite understand the concept of “open source,” anyway. per definitionem, Gnu emacs (to take but one example) is just as much “my own application” as it is anybody else’s … providing, of course, that I abide by the license.

    That is, after all, the whole point of the GPL. You don’t seem to understand that, for some reason. Luckily for FLOSS, Microsoft are rather more intelligent and up-to-date than you are.

  205. DrLoser wrote, “which one of your various career paths included working for an organization in the private sector that depended on customers actually stumping up their own cash for your product?”

    • various yard sales,
    • Versatile Farm Equipment, a division of Ford, back in the 1980s,
    • Quality Bed and Spring co., back in the 1960s,

    And, why specify “private”? Governments are some of the largest businesses around. Yes, they do have budgets and shareholders. They have bureaucracy too.

  206. oiaohm says:

    DrLoser
    Yes you would, oiaohm. There’s nothing to stop you using an MSBuild file with any other compiler/IDE suite. Not that one would do that: one would use Visual Studio, in which case precisely the same licensing terms apply.

    DrLoser you failed to understand the problem. The MSBuild file itself includes the License declare. So using MSBuild file with another complier/IDE suite still sees you hit with the same taint problem. Great fun have to remake your own build file.

    Yes you could be using SharpDevelop instead of Visual Studio and have this taint issue sneak up on you. This is the nightmare user who used Visual Studio Community tainted the file yes it writes in what License the generated file is under being Visual Studio Community. Yes the source codes yours but the file to make it is Microsoft. Not everything is truly your work going into your source code base at times. Yes MSBuild file generation by Visual Studio is a derivative work problem.

  207. oiaohm says:

    DrLoser forbid was the word I was going for.
    Individual license. If you are an individual working on your own applications to sell or for any other purpose, you may use the software to develop and test those applications.

    Notice the words “your own applications”. This means anything from a OSI project that is not 100 percent your own work is not covered by the Individual License.

    If none of the above apply, and you are also not an enterprise (defined below), then up to 5 of your individual users can use the software concurrently to develop and test your applications.
    Some miss read this. Note the section you are in “Organization licenses. ” This does not include sub contractors or individuals working as a group. Only 5 individual users owning to the same Organization.

    Remember to be a Organization you have to have a government registration.

    OSI projects are fairly safe because of the different Organization bodies you can own to so granting you the right to use the MS Tools for OSI stuff.

    Remember “your applications” in the Organization section applies to the Organization you have to be affiliated with. Like anything hosted on github is a github application. So github company is the Organization you are the github user. Same with hosting on sourceforge…..

    Simple for OSI project to meet the requirements. Not that simple for closed source developers who don’t want to risk sharing their code not so much.

    Also that million dollar thing sound cool. Other than being super vague.
    more than one million US dollars (or the equivalent in other currencies) in annual revenues
    So you make 1 million in revenue you might next year to you have to pay now? It is in fact as clear as mud.

    more than 250 PCs or users
    Also clear as mud. 250 users that log into your commence web site? Yes 250 PC is clear. 250 users is not.

    If none of the above apply, and you are also not an enterprise (defined below), then up to 5 of your individual users can use the software concurrently to develop and test your applications.
    Basically this clause is a brick. Most software developers want to sell more than 250 copies and have more than 250 users of their product. Yes has to be fairly expensive software to bring in 1 million dollars worth of income from selling 250 copies.

    So about the only people who can use the Visual Studio Community license is the academics, OSI developers and true solo developers. Everyone else is screwed by MS piss poor define of enterprise.

    Microsoft is really good at writing licenses that trick you that they have in fact given you the right todo something when you don’t have it at all.

  208. DrLoser says:

    I guess you did not read the MICROSOFT VISUAL STUDIO COMMUNITY 2013 stuff either Oldfart this also can bite you in the case of a BSD or equal work with a MSBuild file generated by Microsoft Visual Studio Community. Hello fun that file is covered by Microsoft Visual Studio Community clauses so since you would not the the primary license holder you would not have the right to use it in a project that was not OSI.

    Yes you would, oiaohm. There’s nothing to stop you using an MSBuild file with any other compiler/IDE suite. Not that one would do that: one would use Visual Studio, in which case precisely the same licensing terms apply.

    But it’s awful sweet of you to care so much about the welfare of giant closed-source corporations. When did this particular infatuation take a hold of you?

  209. DrLoser says:

    Read my cite again this time think you are attempting to make a closed source work and that license raises is ugly little head. 1 team work on a closed source application forbidden by that license.

    No it is not.

    Quote the precise paragraph, please. And please try to spell “forbidden” correctly. I will not help by correcting you next time.

  210. DrLoser says:

    What’s slightly bizarre about that OSI clause in the EULA is that, if you go to my cite, you will find practically every “open source” license in existence listed. Except for the Microsoft one.

    Of course, that’s included, too. But it’s not heavily advertised by Microsoft, for some reason.

    Never mind. The whole Open Source Initiative thing was cooked up by Bruce Perens, who is a witch and a heretic and an apostate and should be burned at the stake.

    (Or, in Robert’s more enlightened view, simply hanged by the neck until pronounced dead. Although I still feel the corpse should be buried at a cross-roads with a wooden stake through its rotten little heart, just to make sure.)

  211. oiaohm says:

    DrLoser OSI developers most of the Microsoft crap is not harmful to. It those like Oldfart attempt to keep the tradition of closed source development who are in the firing line due the the Microsoft Licensing crap.

    Read my cite again this time think you are attempting to make a closed source work and that license raises is ugly little head. 1 team work on a closed source application forbin by that license. This means you now only have the option of OSI. MSBuild file contains this taint even after you regenerate. Microsoft with sneaky underhanded viral traps.

  212. DrLoser says:

    oldfart wrote, “You know nothing of business …”

    I’m not entirely sure that quoting from some random piece of FLOSS puffery is a convincing refutation, Robert.

    Although I’m open to the idea that you have some sort of business experience to bring to the table. Tell me, which one of your various career paths included working for an organization in the private sector that depended on customers actually stumping up their own cash for your product?

    That, I think, would qualify as relevant business knowledge.

  213. DrLoser says:

    I was curious to see what dark horrors oiaohm had uncovered in the relatively uncontroversial Visual Studio EULA:

    I also love some other nasty Microsoft clauses.

    Individual license. If you are an individual working on your own applications to sell or for any other purpose, you may use the software to develop and test those applications.

    Well, that’s pretty darned devastatingly shocking. NOT.

    Any number of your users may use the software to develop and test your applications released under Open Source Initiative (OSI)-approved open source software licenses.

    My heart bleeds, bleeds it does, for all those poor Open Source Initiative programmers who have been locked in to this egregious … wait a mo … they’re not locked in at all, are they? This is a terrible, terrible, onerous EULA.

    And it only gets worse if you fall under the definition of an Enterprise, to whit more than 250 seats, or $1 million per annum revenue.

    What swine these people are!

    But, back to noddy-land:

    So if you get any assistance at all MICROSOFT VISUAL STUDIO COMMUNITY 2013 cannot any more be used to make a closed source program.

    Not evident in this particular cite, oiaohm. What are you babbling on about this time?

  214. oiaohm says:

    Microsoft makes software coding and documentation writing a game pot luck of am I still legal.

  215. oldfart wrote, “You know nothing of business cheapskate.”

    How To Use Linux To Help Customers Go to Market Faster

    “If you have an idea and a business model to base it on, going the open source route can result in a huge time saving in getting your product to market. In fact, your project may become one that can save a great deal of development and permit you to get the product to market faster. This is especially true if you need to modify open source software – you have already done your homework and can show the cost benefits of using the open source software.”

    See Expert Mysql

  216. oiaohm says:

    I guess you did not read the MICROSOFT VISUAL STUDIO COMMUNITY 2013 stuff either Oldfart this also can bite you in the case of a BSD or equal work with a MSBuild file generated by Microsoft Visual Studio Community. Hello fun that file is covered by Microsoft Visual Studio Community clauses so since you would not the the primary license holder you would not have the right to use it in a project that was not OSI.

    Reality documentation and code sides are both out to get you all because of Microsoft annoying licensing.

  217. oiaohm says:

    oldfart sorry what is the independent work the manual or application. You can end up in the result that they are both derivative works from each other.

    My license for office would the same full blown office profesional license that is used by everyone to write technical documentation.
    LOL. you completely miss it. Remember what I said that its the version of MS Office to produce the documents is the issue not the version you are using alone. Also not use by everyone to write technical documentation is using full blown office professional License. Some are in fact legally using MS Office academic. You will find OSI based stuff at times have documentation written by MS Office academic. So the BSD code base might be fine for your closed source but the documentation about it was not. Notice how the manual and the code in some projects ship strangely independent. Guess why.

    Unless of course you are going to now say that the entire Microsoft developer community has put their code rights in jeopardy by writing about it.
    Of course moron Oldfart would go this line. The complete Microsoft developer community put their code bases at risk all the time not understanding the basics of derivative works and how many land mines out there.

    Oldfart I guess you do receive doc files by email. Taint marking from MS Office academic can appear in your .doc file even if you paste something in then straight up delete it. Entrapment is so simple.

  218. oldfart says:

    “Really here is you presuming editing code. Someone types up some notess you copy them into your code base as comments. I never said you would be editing code inside MS Office did I. This is where you are making stuff up completely. You might be editing your manual in MS Office.”

    So now according to the sage of the outback, one can lose the right to code by editing a technical manual. You are being an idiot in the name of winning brownie points Fifi.
    And this is over and ablve the fact that You keep assuming that I will somehow be using versions other than the Standard commercial licenses product. My license for office would the same full blown office profesional license that is used by everyone to write technical documentation.

    Unless of course you are going to now say that the entire Microsoft developer community has put their code rights in jeopardy by writing about it…

    (this should be good 😉

  219. oldfart says:

    “Not having to produce all the code in a product helps get to market sooner at lower cost, huge benefits in the world of IT. T”

    You know nothing of business cheapskate. All you know is that you bask in the warm glow of your freebie and will do anything to keep it going.

    Fortunately the world is smarter than Robert Pogson thinks it is.

  220. oldfart says:

    “The world of GNU/Linux will just leave them in its dust.”

    Only as long as commercial businesses choose to support it because they think that they can live with its terms. A bad loss by a business a major as VMWare should disabuse them of that fantasy.

    And if the business men don’t get it, as the good doctor has pointed out, their lawyers will.

  221. oldfart says:

    “I think this reflects a closed mind. ”

    And I think Robert Pogson that all of your protestations of the virtues of sharing are nothing more than a smokescreen to hide the fact that you are a cheapskate who has found nirvana in the form of FOSS. You wish to continue to bask in its glow so now you doing anything and everything you can to work towards this end. You unashamedly represent your ancient experiences with now obsolete versions of microsoft OS’s as run on inadequate hardware as if they still represent the state of the art. And you continue to engage in emotional blackmail in your attempt to bully people into using software that not only may does not meet their need, but which will require them to work harder to reach the goals that they bought their computer for in the first place.

    “Musicians share inspiration on a grand scale. Very few find satisfaction playing their own music inside a barrel to keep others out.”

    Successful and popular Musicians also make a living of their creative output by selling their creations. But of course Robert Pogson cheapskate, who seems to think he is entitled to pay for as little as possible, says their desire to make money is bad, and that they should learn to share. How nice!

    What really needs to happen is that Robert Pogson should learn is to keep his hands out of other peoples pockets!

    There will always be entitled cheapskates like you Robert Pogson, who think that the world needs to cater to your idea’s of what goods should cost and how long you should be able to continue to use them before you have to pay for new. FOSS may have started out as in idealistic coop, but it has long since been hijacked by the cheapskates like yourself who mooch off its benefits on one side, and the businesses who game the licenses of what is freely given to enhance their ability to make a large amount of money.

    Frankly, I would love VMWare to lose this one and pay big, because IMHO that is the only way that people will realize that the GPL is not to be played with,
    One can only hope that the outcome of Nitwit vs. VMWare

    The world does not and has never worked that way.

    “In a competitive market such works can’t compete with freely-shared works like GNU/Linux.”

    Sure they can – they do every day. Remember Robert Pogson, the world of commercial software is much bigger than one company. Microsoft is not the world, and a saturated market is not a dead one. In the end money talks and the realities of FOSS beyond the fact that it is free speak for themselves.

  222. oiaohm says:

    There just isn’t a good business-case for secrecy.
    This is not exactly true Robert Pogson share trading, information that is required for a patent applications and a few other operations internal secrets are important. Most of these have a limited time usage. Remember patent application in some place the first to file wins.

    But I do agree there are a lot of cases that internal secrets are being kept secret with no commercial gain reason.

    Secrecy leads to doubts about integrity, a killer.
    This is truly becoming a big problem.

  223. oiaohm wrote, “Find yourself in a location where you have to release details that for commercial gain you need to keep secret is bad.”

    That might be correct in a monopoly based on secrecy but the world is thriving on openness. Secrecy leads to doubts about integrity, a killer. The world of IT is becoming rapidly more open as open standards work for everyone. There just isn’t a good business-case for secrecy. If you have to keep secrets to succeed, you’re moving too slowly and will fail in a free market.

    I’ll give an example. I once designed and built my own PC. It was built on Motorola 68K and used a VMEbus and crate for a backplane. By the time it was ready to fire up is was a couple of steps back on Moore’s Law, bus bandwidth, RAM and all it had for software was a UNIX in ROM. I never fired it up because it was pathetically obsolete. I still have it on a shelf somewhere, reminding me how IT should be done, with open standards and FLOSS. I think it was two years from conception to completion, at least twice too slow to be good IT. M$ took 8 years to produce Vista and shot itself in the left foot. M$ then shipped 8 and shot itself in the right foot. Some of the GNOME/KDE/systemd folks are trying to do the same with FLOSS but it’s a lot harder. The world of GNU/Linux will just leave them in its dust.

  224. oiaohm says:

    Robert Pogson Please note I said GPL can be harmful. Its not aways harmful.

    Yes it can enrich as well. Find yourself in a location where you have to release details that for commercial gain you need to keep secret is bad. But enrich by having better development can help.

    The restrictions in Microsoft licensing make GPL look angelic so often it is not funny. GPL does not prevent you from working in a team or putting restrictions on team sizes ….

  225. oiaohm wrote, “GPL…can be harmful to business”

    The GPL can also enrich a business with huge leverage. Not having to produce all the code in a product helps get to market sooner at lower cost, huge benefits in the world of IT. This is why M$’s client division is sputtering. They try to do everything in-house and it does not work well. The world is leaving them behind. That’s why they missed the boat with mobile while being overwhelmed by malware. The malware artists got way ahead of M$. That’s why M$ is starting to give away binaries. Their old way of doing things can’t compete in a fast-paced free market.

  226. oiaohm says:

    https://www.visualstudio.com/support/legal/dn877550

    I also love some other nasty Microsoft clauses.

    Individual license. If you are an individual working on your own applications to sell or for any other purpose, you may use the software to develop and test those applications.
    So if you get any assistance at all MICROSOFT VISUAL STUDIO COMMUNITY 2013 cannot any more be used to make a closed source program.

    Like a person like me who needs english proofed read or you are wanting translations performed. Yep you cannot using community.

    Next you take something produced by MICROSOFT VISUAL STUDIO COMMUNITY to your company it must be under a OSI license. Note this block using a non OSI license on that code.

    oldfart you like complaining about GPL. GPL is not blocking like this. GPL only if tainted do you have to release source code and it does not block you relicensing you own code closed source.

  227. oldfart wrote,

    • “…compromise my creator rights…”
    • “The fact that by releasing source code I am providing any would be competitor with a blueprint for undercutting the value of what I am creating is apparently for you, neither here nor there.”

    I think this reflects a closed mind. Doing everything by one’s own efforts is commendable, heroic, legendary, but it is not optimal. We are social beings. Our strength/capability is greatly extended by sharing the load with others. That’s what corporations are all about. That’s why corporations merge/trade/cooperate. It’s rarely optimal to cut off sharing.

    Musicians share inspiration on a grand scale. Very few find satisfaction playing their own music inside a barrel to keep others out. The same is true of computer programmers. Modern software is usually too large/complex for any individual to own all of it. Take the Linux kernel for example. One guy did create the first draft which barely did the job but he was able to inspire many thousands of others to join in the task. Linus gave up nothing of his creative work yet he was able to build on the contributions of a good chunk of humanity. That’s the right way to do IT. There’s no merit in a single individual or corporation even keeping some work in-house and closed. That’s inefficient, inflexible and likely to be far from state of the art. In a competitive market such works can’t compete with freely-shared works like GNU/Linux. That’s why M$’s client division is in decline and M$ is set to give away the binaries and update them forever for $0. M$ has seen the light, almost, but oldfart and some others are living in the inner sanctum of a deep cavern.

  228. oiaohm says:

    Did you consider the possibility that I might be responding for my self and not as part of a team? the very nature of my answer should have told you that I understood the problem and the s0lution. But since you wanted brownie points, you assumed that I was incompetent. That assumption makes you the incompetent in this case, not I.
    Oldfart the answer does not change even if you are attempting to be a individual operator. You might depend on a Library developed by someone else who screws up.

    This really is quite entertaining. You are an incompetent buffoon if actually believe that I would be editing code inside Microsoft Office?
    Oldfart how do you make help files for MS Windows programs.

    Really here is you presuming editing code. Someone types up some notess you copy them into your code base as comments. I never said you would be editing code inside MS Office did I. This is where you are making stuff up completely. You might be editing your manual in MS Office.

    So when we clear away all the irrelevant blather, your contention is that the GPL is better for that the microsoft Academic license (which I already established that I would never work under in the first place) because it “only ” requires me to release source code for all to see. The fact that by releasing source code I am providing any would be competitor with a blueprint for undercutting the value of what I am creating is apparently for you, neither here nor there.

    The reality is being forced to release the code GPL does not stop you from correcting your ways and understanding independent works and derivative works so allowing you to product future works correctly that expand on your prior works functionality and to sell commercially.

    Remember anything that is your own copyright work you can multi copyright.

    So it is possible for VMWare even if they are forced to release ESXi kernel open source under GPLv2 to close source it again in future after its been cleaned of taint.

    GPL does not prevent you charging for your product while you are correcting the screw up. Non-commercial means exactly that while you are fixing the taint you cannot bill clients even to provide support.

    GPLv2 is correctable. Microsoft Academic license damage is start over from scratch.

    I am providing any would be competitor with a blueprint for undercutting the value of what I am creating is apparently for you, neither here nor there.

    Remember GPL mirror license effect you as master author can make your source code closed again once you undo the tainting problem. Your competitor under cutting you has not such protection.

    GPL is possible more competition but you still can have a income. The products containing the nasty derivative work non-commercial usage is a death sentence.

    Oldfart choose Zero income because you risked usage of something with derivative work non-commercial thrown on you. Or GPL where you are legally still allowed a income just have to fight harder with some competitors for a while.

    Please note academic usage is not the only Microsoft licensing way with restricting. Microsoft Partner licenses derivative work limitations are kind a little worse only for your own companies usage so make something and attempt to sell to client breach of license again.

    Microsoft derivative work clauses can block you from even selling support for your product. Non-commercial product does not have support. Yes this really sux because your competitors can sell support for the program you made and there is absolutely nothing you can do about it.

    GPL is moderately bad. Remember GPL yes you have release source code. But you can still charge for the binary and update and support.

    Breaching Non-commercial derivative work clauses is very much the monploye board game of go to jail do not pass go do not collect any income.

    Basically if I had a choice of licenses to breach I would prefer BSD, MIT … where is just add a few decelerations and I am fixed. If I breach GPL ok this is livable. Might be painful and I am testing my customer loyalty. But there is no way to live after Non-commercial or restricted own company only usage derivative work breach.

    Please note partner version of MS Windows contains the taint clause of restricted for training or own internal usage.

    This is why I hate Microsoft. Too many licenses. Too many ways to screw up.

  229. oldfart says:

    “These GPL test cases on what derivative work covers and what it does not will be helpful to work out how viral MS Office Academic Licenses and other viral licenses requiring horible restrictions are.”

    But since I am staying away from the licenses you spoke of for my personal development tasks, your contentions are irrelevant and you are only showing your incompetence in continuing to exposit them.

    “Basically every GPL case indirectly applies to how well you can protect hide in future one way or another if someone does things wrong with other viral licenses Oldfart. Yes you might need to be call them as examples of why enforcement should be impossible.”

    Actually Fifi, I won’t need to do anything. As you well know I intend to stay away from GPL’d code completely. And since I am speaking of personal development efforts, my control of what is used is absolute.

    and you are, as usual, full of crap.

  230. oldfart says:

    “Sorry not at all your statement was wrong. Provide software was only half the problem. Anyone who has really done it would have included enforce usage.”

    My statement was fine. It was your assumptions in responding that are wrong.
    Did you consider the possibility that I might be responding for my self and not as part of a team? the very nature of my answer should have told you that I understood the problem and the s0lution. But since you wanted brownie points, you assumed that I was incompetent. That assumption makes you the incompetent in this case, not I.

    “This is truly possible it what the license agreement says. You are presuming that Microsoft does not transfer enforcement rights to a completely troll company opps they already have in BSA.”

    But since we have established that I would be taking special care to stay away from products that could compromise my creator rights, your interpretation (as stupid as it is) is moot.

    “The derivative work clauses on MS Office Academic is that vague is viral. So you copy paste from the .doc made in MS Office Academic into your code base you have just tainted it with a non-commercial usage as its now become a derivative work.”

    This really is quite entertaining. You are an incompetent buffoon if actually believe that I would be editing code inside Microsoft Office?

    “Oldfart I would be fairly sure at times you have breached the MS Office Academic viral license because you were not aware of it.”

    And you would be wrong. I would not edit code inside microsoft office – I work in programming editors/IDE’s when I code.

    “Yes it one thing to call GPL bad you have to remember its not the only Viral class licensing out there. Some are extremely bad. Some of the worse are by Microsoft. GPL is only a moderate for risk when it comes to derivative works clause containing licenses. Having non-commercial usage stamped on everything is fairly much a day ruiner. GPL just stamps you must release source code but you can keep on using it”

    So when we clear away all the irrelevant blather, your contention is that the GPL is better for that the microsoft Academic license (which I already established that I would never work under in the first place) because it “only ” requires me to release source code for all to see. The fact that by releasing source code I am providing any would be competitor with a blueprint for undercutting the value of what I am creating is apparently for you, neither here nor there.

    As far as I am concerned, you are not only a proven and self confessed liar and fraud, but you are a fool if you think that anyone is going to take this kind of crap seriously as a defense of the GPL.

  231. oiaohm says:

    These GPL test cases on what derivative work covers and what it does not will be helpful to work out how viral MS Office Academic Licenses and other viral licenses requiring horible restrictions are.

    Basically every GPL case indirectly applies to how well you can protect hide in future one way or another if someone does things wrong with other viral licenses Oldfart. Yes you might need to be call them as examples of why enforcement should be impossible.

  232. oiaohm says:

    Nope, you made up assumptions of what I would or would not do based on your desire to win brownie points in a debate. It does not work that way.
    Sorry not at all your statement was wrong. Provide software was only half the problem. Anyone who has really done it would have included enforce usage.

    I now understand why Doctor Loser finds you so entertaining! Imagine – I could compromise rights to code that I developed in a programming language just because somebody typed up notes in Office Academic.

    Is this actually your contention, Fifi?
    This is truly possible it what the license agreement says. You are presuming that Microsoft does not transfer enforcement rights to a completely troll company opps they already have in BSA.

    Every derivative work from the MS Office Academic is for non-commercial usage. Microsoft does include giving away for free as non-c0mmerical usage.

    The derivative work clauses on MS Office Academic is that vague is viral. So you copy paste from the .doc made in MS Office Academic into your code base you have just tainted it with a non-commercial usage as its now become a derivative work.

    Oldfart this is the nasties and the most common viral derivative work license regularly broken. When BSA comes to your door to audit and finds this you find yourself looking at huge bills for no apparently good reason. Yes each use of a Academic based document is copyright infringement that BSA wants to charge you a full copy of MS Office for.

    Breach of copyright by BSA might have completely nothing todo with the software you have install. Yes the software your staff have installed and are using is critical if BSA is well and truly digging for infringement. In fact finding MS Office academic produced documents come from your company may be the very thing that brings the BSA in your door so they can have legal search order to find more.

    Fairly much the only legal way around MS Office Academic license is print the document out and retype it. There are are few not exactly legal ways to strip identifying marks. But if you just use the MS Office Academic produced file as is expect to have your ass handed to you some point in future.

    Oldfart I would be fairly sure at times you have breached the MS Office Academic viral license because you were not aware of it.

    Yes it one thing to call GPL bad you have to remember its not the only Viral class licensing out there. Some are extremely bad. Some of the worse are by Microsoft. GPL is only a moderate for risk when it comes to derivative works clause containing licenses. Having non-commercial usage stamped on everything is fairly much a day ruiner. GPL just stamps you must release source code but you can keep on using it

    Basically you have 3 basic classes of licenses.
    Non-viral. BSD,MIT,MPL….(LGPL is mostly non-viral)
    Viral requiring source code release based in derivative work clauses GPL, AGPL… Yes can be harmful to business but normally will not kill a business.
    Viral requiring horible restrictions like non-commercial usage so only possible fatal to businesses as not only due to limit your usage it can result in you being unable to use it at all. Yes this can include software your company gets under NDA or stuff like MS Office academic.

    Viral requiring horible restrictions people like to pretend don’t exist. Its very simple to ignore. Companies like Microsoft have not publicly sued over it.

    Remember in the USA your copyright Life + 70 years and this is the same for many other countries. So sometime between the time of the products release and the authors death + 70 years can the conditions of copyright be enforced.

    Copyright extreme life span means you really cannot think you out the woods because it has not been enforced yet.

  233. oldfart says:

    “I am not a blood sucking drone. If you correctly interface with open source code and make a closed source program I hold nothing against you.”

    Thats nice. Unfortunately for you, you have no standing regardless, so you get nothing.

    have a nice day!

  234. oldfart says:

    “Yes I am 100 percent sure you would screw up as you have this wrong to start off with. ”

    Nope, you made up assumptions of what I would or would not do based on your desire to win brownie points in a debate. It does not work that way.

    “MS Office all it takes is someone has done a course got a free laptop loaded with academic MS Office and does some office work on it. ”

    I now understand why Doctor Loser finds you so entertaining! Imagine – I could compromise rights to code that I developed in a programming language just because somebody typed up notes in Office Academic.

    Is this actually your contention, Fifi?

  235. oiaohm says:

    If I am going to make the investment to protect my creation, do you really think that I won’t make sure that everyone involved has the tools needed for my success?
    Yes I am 100 percent sure you would screw up as you have this wrong to start off with. Parties work from the point of view they can provide the tools it will be fine are dead men walking.

    Funny that you now are saying you will provide all the correct tools. The problem is not providing the correct tools but 100 percent enforcing that they use the correct tools.

    Two programs that look absolutely identical with different derivative work is a major nightmare when it only takes a person to stuff up once to cause a major event. Ask a person what the used to make a document or a file how often will they tell you it was an academic or pro or what ever version of Microsoft product. MS Word or Visual Studio is not enough information to know how you can legally use the file. No someone says they made the file with Libreoffice or Eclipse… You know where you stand. Why every version of that product has the same license conditions.

    MS Office all it takes is someone has done a course got a free laptop loaded with academic MS Office and does some office work on it. Issue MS Office does record inside the document what made it. Lets say someone does up your invoice template using the wrong version of MS Office. Hell to pay right.

    Now lets say your business uses Libreoffice or WPS instead of MS Office. Libreoffice LGPL contains a clear statement that your own generated derivative works are not covered by the Libreoffice license. WPS in its license also includes the same clause for all versions.

    “GPL big advantage is there is only 1 version to obey on a source code and has to be declared.”

    Yes indeed, it is a big advantage to know that all my work will go towards blood sucking drones like you and Robert Pogson!

    Oldfart there was many ways for VMWare to obey GPL and maintain provable independent work status so avoiding ending up in the current mess. Simplest was making a wrapper for freebsd or solaris drivers as well as Linux. So not creating a dependent event so limiting derivative work problem.

    I am not a blood sucking drone. If you correctly interface with open source code and make a closed source program I hold nothing against you.

    Sorry I don’t want your source code unless you are incompetent. If you don’t know how to interact correctly against derivative work clauses you cannot be trusted to audit your code base for bugs either.

    Oldfart if anyone is a idiot on this topic it you.

  236. oldfart says:

    “Oldfart is not that simple. All it takes is one member in your team to break the rules and you are tainted. ”

    If I am going to make the investment to protect my creation, do you really think that I won’t make sure that everyone involved has the tools needed for my success?

    “GPL big advantage is there is only 1 version to obey on a source code and has to be declared.”

    Yes indeed, it is a big advantage to know that all my work will go towards blood sucking drones like you and Robert Pogson!

    Idiot.

  237. oiaohm says:

    “Microsoft academic/home versions of software not so nice. .”

    So what. All I have to do is invest in the full price development software. On the other hand there is only one GPL.
    Oldfart is not that simple. All it takes is one member in your team to break the rules and you are tainted. Paying for the full price development software does not license code produced by home versions. So yes you pay full priced development software and still can land in court for copyright infringement.

    GPL big advantage is there is only 1 version to obey on a source code and has to be declared. EULA of Microsoft has no requirement that tainted status has to be applied or even recorded.

    DrLoser Oracle vs Google case is not in a Civil Court designed system. This is why you cannot compare. There are Australian copyright cases you could use. Why equal court system design. Oracle vs Google is also automated reverse engineering case. So the none of the material in use by Google is in fact a 1 to 1 copy. So question comes if Google was libable to obey terms in the case if it was held in the Civil Court systems of the world. Headers come down to fair usage arguement.

    VMWare case there are 1 to 1 file copies. So there is no question that VMWare has to obey terms. One of the terms in GPL is that all Derivative works will be under the same license. Linus clause exists on Linux because if it did not there are applications using the Linux kernel by syscall that would not be classed as independent works so liable to GPLv2. The clause VMWare carelessly removed.

    VMWare does not have Google arguement of fair usage on compatibility grounds. Take the wine project it can argue fair usage on compatibility grounds but you will notice there is not a single bit of Microsoft code inside that is against license.

    Compatibility shims have been done by many parties legally.

    Microsoft legal department called GPL viral for a reason. A lot of Microsoft Licenses are also Viral.

    Now, about those Software Conservancy claims on ESXi violation of the GPLv2, as regards SCSI?
    Subsystems of the Linux system are under GPLv2 + Linus clause. Usage without provable independent work status is a no go due to modified status.

    Lets say vmware shim system allow you to kernel source straight from kernel.org unmodified. You still have independent work status. Problem here is VMKLinux is a modified work. So is now a derivative work. VMKLinux does not provide the separation to the VMKernel.

    Yes the complete case will come down against VMWare as a derivative work agruement as I stated right off the bat DrLoser. Two independent works being used with each other don’t taint licenses of each other. Derivative works on the other hand have to respect the license they come from.

    Every fork is a Derivative work. Next Derivative work that is custom made for a task gets hazard really quickly because you don’t have independent work status.

    DrLoser think of this I can make a modification to allow XP/7/8… accept as many connections like the server form and also unlock the number of RDP users that can be logged in as well. Exactly what is the difference with what VMWare has done, not much really. Both are modify a copyrighted work without license base todo so. VMWare has added their own private interface with no real license base todo so. No Linus clause, No strong provable independent work status.

    Now the German courts might rule that VMK was enough separation. But prior cases with Libraries this has been ruled as too cute. Yes where someone has attempted to put a Library interface over a GPLv2 work without clause.

    There is a reason why you don’t see GPLv2 libraries pure. Libraries are either GPLv2+clause or LGPL. GPL derivative work clauses are down right strong without some exception clauses.

    I think there’s a big difference between using “header information” (or “unpublished APIs,” as oiaohm seems to prefer to call the stuff)
    Please stop using header inforamtion DrLoser its wrong.

    USA and German courts and other have ruled API declarations usable under fair usage. Macros and Inline code in header files have never been ruled as usable unless it passes the classic French test. Some header files in C can in fact include a int main() { stuff} inside a ifdefine. Why turns out things like gcc -DTESTING header.h will build. Unix world put very little stock in extensions.

    Macros and Inlines inside a header file might has well be in a .c file.

    DrLoser do you know why most are not answering you. Simple “header information” that is not a valid legal term when it comes to copyright. API declarations, API…. . These have more valid meaning. Part of the reason why programmers get into trouble there are a lot of myths over what is not copyrighted/protected. Not copyright protected is quite limited.

  238. DrLoser says:

    The weakest part of my argument here is Exhibit Four.

    I took the Software Conservancy at their word, and worked through their objections. (And all of those objections may well result in legal victory. That was not my point.)

    Exhibits One, Two and Three don’t seem to add up to much, as far as I can see. And apart from oiaohm, nobody here has exercised their intellect on the matters.

    It’s not enough to keep singing GPL! GPL! GPL! as though you are either a Medieval Monk who likes Gregorian Plain Chant, or otherwise a slightly etiolated soccer hooligan.

    You actually have to think for yourself, as I have done.

    Now, about those Software Conservancy claims on ESXi violation of the GPLv2, as regards SCSI?

    You’re all self-confessed blood-sucking drones, aren’t you?

    Not a single one of you is prepared to offer up an opinion.

  239. DrLoser says:

    My view on the Linux headers also remains untested on this site.

    I think there’s a big difference between using “header information” (or “unpublished APIs,” as oiaohm seems to prefer to call the stuff) above the line and below the line, as it were. It seems a judicial stretch to disallow somebody (corporate or individual) the right to offer an interface that is already there to the providers of a service at a lower level. Walking off with the ability to effectively steal all the work done above that interface, which is what I think Google may fairly be accused of doing … well, that’s slightly different.

    I realise, however, that this thesis is on legally shaky and so far untested ground. Which is why I am particularly interested in the German Court judgement on Exhibit Two.

    One further point. The Linux headers are licensed under the GPLv2, which is expressly designed not to prevent other people “making money” off them … so long as the publication and distribution terms of the GPLv2 are strictly observed.

    The Oracle vs Google case, on the other hand, is predicated on the CDDL.

    A different and, I think, thornier beast.

  240. DrLoser says:

    I think I’ve already made my thoughts on that fairly plain, oldfart.

    1. From a logical standpoint, it makes no sense whatsoever. (Provided you abide by the rules of the GPLv2.) Robert, and quite possibly oiaohm, started with the mistaken belief that this “repurposing” somehow occurs within the kernel. Which it doesn’t. I think that’s tainting their argument.
    2. From a copyright perspective … well, I still think it doesn’t. With the possible “gotcha” of the Linux headers. I’ve tried to entice either Robert or oiaohm to make a comparison here with the Oracle vs Google case, but despite the current state of that case — which would seem to favour Oracle on “headers,” and therefore Nitwit likewise — they are strangely reluctant to stick their neck out.
    3. From a commercial standpoint, it seems absurd not to allow it. Time-to-market for a hardware manufacturer wanting to interface to ESXi? About four weeks, rather than 26. Cost to market? About 10% of the alternative.

    It would seem as though Robert is only in favour of saving huge chunks of money if he’s the one doing it. If other people do it, and to quote Robert, “it’s a hanging offense.”

    German courts are interesting, though.

    a. They do, as oiaohm points out, tilt towards the Freetard side of things.
    b. They are not well-disposed to American corporations (it remains to be seen whether VMWare is big enough to catch their ire).
    c. On the other hand, Germans are a frugal lot. Good value and re-use are things that all Germans hold dear.

  241. oldfart says:

    “vmk_PCIGetDeviceName…”

    As I said Doctror, the courts will decide this one. I am however interested in your thoughts (NOT YOURS LIl’Hammie – you will just be ignored!) why a third party device driver whose code has been structured to respond to requests from a linux kernel can not be re-purposed by VMWare to respond from requests from vmkernel?

  242. DrLoser says:

    vmk_PCIGetDeviceName

    BWAHAHAHAHAHA!

  243. DrLoser says:

    But I don’t wish to leave you all without the warm fuzzies engendered by this (unacknowledged) bit of addendum. Here’s a list of worthies who have Signed Up To The Cause:

    * APRIL
    * Free Software Foundation
    * GNOME Foundation
    * The Samba Team
    * The SWIG Project
    * Dave Airlie, Linux Developer
    * Matthew Garrett, Linux Developer
    * Grant Likely, Linux Kernel Engineer
    * Michal Nazarewicz, Linux Developer
    * Luis R. Rodriguez (aka mcgrof), Linux Developer
    * Wolfram Sang, Linux Developer
    * Josh Triplett, Linux Developer
    * Rik van Riel, Linux Developer

    Boffo!

  244. DrLoser says:

    I notice that the Software Conservancy parasites have extended that cite somewhat. Apparently the diagrams weren’t enough — they’ve added a bunch of blather just beneath them. As I recall, said blather wasn’t present in the original form of the cite.

    Then again, it’s worthless gibberish. Feel free to quote from any part of it. Just remember that it was not, apparently, part of the original plaint.

  245. DrLoser says:

    I assume that 279,653 lines of code has been peer-reviewed, btw.

    An expensive, but regrettably necessary, part of Quality Assurance.

  246. DrLoser says:

    Anyway, back on topic. TheOP. Nitwit vs VMware.

    Second thoughts on Exhibits One, Two, and Three, as visible on the diagram on Robert’s FAQ cite, right there in the orange bit? (It seemed to be a reasonable strategy for organisation.)

    No? OK then, we can proceed to Exhibit Four, which is to say, SCSI drivers, sub-category Hellwig. I’ve put some preparation into this. You are welcome to do the same.

    But, while we all assemble our Gnomic Utterances, I can’t help but feel for the poor little lad:

    Christoph is one of most active developers of the Linux kernel. He has contributed 279,653 lines of code to the latest Linux 3.19 kernel, and thus ranks 20th among the 1,340 developers involved in that release.

    Yikes! How many lines of code did the top 19 contribute between 3.18 and 3.19?

    Jeez! Talk about “bloated!”

    But, I digress. Opinions on SCSI in ESXi, please. Any opinions at all. And here’s a helpful hint: the use cases appear to involve Storage Area Networks.

    There you go, Fifi. You are a (self-) acknowledged Expert in Storage Area Networks.

    Devastate us all with your Comprehensive and Irrefutable Knowledge!

  247. DrLoser says:

    DrLoser something else to remember the human body is more likely to live from a hit from bellow than a hit from above. Yes a hit on you head is way more likely to kill than the painful hit in the groin.

    That interesting observation is, again, not a part of the structural engineering canon, oiaohm. I suppose you could just about imagine it featuring in a conversation between the relative merits of ULS vs SLS … not really, I’m just kidding.

    Your honour, I submit that my client’s shoddy design intentionally maximised the possibility of causing a 500lb out-of-control weight to land in the customer’s groin at speed. In fact, it said so right there on the package. Just underneath the GPL …

  248. DrLoser says:

    Having a backfoot can kill you. If you are to a failure point that you exceed the force the frame should take and the cable snaps sending that force into the frame if frame can fall backwards that force gets dissipated. Remember no one is meant to behind a lifting frame when you are using it.

    That was a very interesting talk on how “exceeding the force the frame should take” will typically cause “the cable [to] snap,” oiaohm. A novel, indeed as yet unheard-of, approach to structural engineering.

    It went on a bit, though, didn’t it? Anyway, me, I’d have used an A-frame. Far more stable. Offers freedom of movement through 360 degrees. Four points of anchorage at all times, rather than two points when something goes wrong.

  249. DrLoser says:

    I guess there are reasons XP is still around. These folks are not pushing the envelope.

    I hardly think that being wedded to Debian, without systemd, can be considered to be “pushing the envelope,” Robert. But, sure, playing cute keyboard tricks with browsers — that’s pretty darned rad.

  250. oldfart wrote, “On the other hand there is only one GPL.”

    Exactly. One licence does it all except for libraries there’s LGPL.

  251. oldfart says:

    “Microsoft academic/home versions of software not so nice. .”

    So what. All I have to do is invest in the full price development software. On the other hand there is only one GPL.

  252. oiaohm says:

    Mine tips backwards easily with one hand but you have to go back ~20 degrees before it falls. That’s uncomfortable but not terribly dangerous/risky.
    The frames I am use to using have a tip tolerance between 30-40 degrees so 10 degrees more than yours at least I thought the legs looked a little short. That is made just by making the front feet longer and its also the length of arm. The most massive tip tolerance is 66 degrees and it still right itself but there is no way I would want to use one of those, the speed of returning front feet is foot shattering on a 66 degree version.

    The big danger is if things let loose with a great mass suspended. Then the steel which has stored enegry springs away from the breach. If the cable/chain/lifting points fail, the thing can leap into the air and fall just about anywhere at the same time as the suspended mass falls.

    The leap into air only happens if the frame cannot pivot. Frame like you build Robert since it can pivot it will fall over so redirecting that force into the ground behind it. I have seen H frame bases jump. Using one of these over a service pit to lift out a v12 engine for a semi and the gear broke. You don’t see a L C frame design jump very often if it does not very high and it will normally move backwards direction away for you due to the redirection of force and still fall over. Yes the L C frame falls over faster than gravity taking the excess force into what ever is behind it of course this is exactly where you want it. It also works to take the snapped cable/chain that is still connected to the frame away from the humans. So hopefully no one gets whipped.

    Basically if you have decided where it going to fall if it fails the danger is quite min. Its like cars before crumple zones and after. Yes the cars with designed in failures are safer. Hoist frame with design in failure point is safer than a hoist frame without a failure point. Designed failure point means you can control what happens when a failure happens.

    Safe design is includes designing in failure.

    Its normally a new personal who suggests putting a back leg on or always setting up a configurable hoist frame in L H configuration instead of L C configuration. Yes LH configuration is special cases only because it is increased danger. New personal have no understanding that you must design to fail so trying to make stuff failure proof. Failure proof means 100 percent unpredictable when the failure proof ideas fail.

    DrLoser I have heard and seen way worse suggestions than putting a back leg on. Worst I have seen is steal picking into the ground then tying by chain to the picks to the ground to connect the hoist frame so it will not fall over. Of course the complete frame ends up snapped when everything goes wrong and a person end up being hit by the ARM as it breaks free.

    If you cannot tip a free standing hoist frame over by just using reasonable force worry. Its most likely too structurally solid with no where for failure force to guided.

    DrLoser reading all this do you still want to put a back leg on a hoist frame or will you be like me and just make sure the frame cannot fall forwards.

    How does this relate to Licensing. GPL has means of addressing failures designed in. A lot of commercial licenses do not.

  253. oiaohm wrote, “Why they have built their application to run under Mono and Windows.”

    I was in my bank again today and noticed the consultant was using IE on XP still. I noticed she was using only half the width of the monitor for a site she showed. I asked her to zoom it so I could read it and she didn’t know how… I showed her F11, CTRL-SHIFT+ and CTRL-. It was new to her… 8-). I guess there are reasons XP is still around. These folks are not pushing the envelope.

  254. oiaohm wrote, “if frame can fall backwards that force gets dissipated.”

    That’s the way to look at it if it’s just a matter of stability, like leaning against it causing it to fall over. Mine tips backwards easily with one hand but you have to go back ~20 degrees before it falls. That’s uncomfortable but not terribly dangerous/risky. The big danger is if things let loose with a great mass suspended. Then the steel which has stored enegry springs away from the breach. If the cable/chain/lifting points fail, the thing can leap into the air and fall just about anywhere at the same time as the suspended mass falls. The frame could land just about anywhere and strike in just about any direction if the frame itself fails. That can be seen when a long tow rope or cable snaps. The ends fly away with lethal energy. My engine is not really such a huge mass on the scale of things and a few millimetres of deformation elastically with ~400 pounds weight on it is not that much energy. Still it needs to be respected lest it be delivered to a human being. One of the reasons I had my son around was to have a second brain/pair of eyes engaged to watch out for danger, ordinary things like not placing body parts beneath the load and positive things like wearing eye/head protection/gloves and avoiding swing in the load. We did swing it a few cm at a time but in a very gentle and controlled manner by lifting/lowering slowly. It took hours.

    If we were going into production readying these things for consumers, we would certainly have a better hoist of absolutely indestructible rigidity, like 6″ S-beams and nice wheels on the lifting point. I once worked in a tractor-factory and they had hoists that worked nicely in 3D. You could walk them in any direction safely and raise/lower with an electric winch even with ~1000lb tractor-frame on the hook. Mine was marginal by design, to minimize costs/time/space and to maximize flexibility (not the opposite of rigidity, but use where the legs could straddle a load. The hoist had to come to the load.). It will likely be more used for butchering deer from now on but it will be available for lifting heavy things anywhere in my yard.

  255. oiaohm says:

    DrLoser something else to remember the human body is more likely to live from a hit from bellow than a hit from above. Yes a hit on you head is way more likely to kill than the painful hit in the groin.

    So physics of lifting frames are that nasty. Yes you calculate for having a possible groin hit. So you 100 percent sure don’t make the front legs narrow.

  256. oiaohm says:

    DrLoser
    Let me explain to you, oiaohm, what happens when a cable under tension, suspending maybe 5000+ N, is released. It doesn’t have to be a cable. It could be anything else under tension.

    Done that one in real life. In fact you don’t understand the physics.
    Let me explain … no, wait, I don’t think I’ll bother. Let’s just say that a back foot is not an absolute necessity, but it’s quite useful when things don’t go precisely to plan.

    Its not a good idea a single back foot you have forgot the frame could go diagonal.

    The C shape of the frame from side on is critical. If you extend the front legs you also prevent it from falling over backwards. You place weight above the front legs you also prevent it from going backwards so limiting the only path as forwards.

    Having a backfoot can kill you. If you are to a failure point that you exceed the force the frame should take and the cable snaps sending that force into the frame if frame can fall backwards that force gets dissipated. Remember no one is meant to behind a lifting frame when you are using it.

    Now lets take your backleg that you liked.
    Option 1) Instead of going nicely straight backwards it can go at a angle so driving someones head into the L frame because the one of the front legs hit them driving them side ways instead of driving them straight backwards into clear space. . So you must at-least fit 2 to avoid this. Yep slide in versions into the back of the front stabilizer feet to make a H shape on the ground is the only way todo it if you do it at all.
    Option 2) the L frame buckles/snaps because it cannot move due to the legs you added now collapses forwards when the gravity pulls it back down straight on the person trying to position stuff its unlikely for the snap to be clean so its not going to be straight forwards. Yes it would have fallen over backwards if you had not added the leg.
    Option 3) the back leg snaps off and alters nothing what means it was pointless to put on in the first place and still falls over backwards.

    When ever you are using a lifting frame with back legs you have to be many times more careful.

    DrLoser choose where you want to dispose of excess force. Straight behind the lifting frame is a very good location for force disposal where you should not be standing anyhow. In fact you should mark it out as a no mans land if you are in a busy workplace.

    When things don’t go precisely to plan is why you try to avoid backlegs. You are only interesting in issues that will bring the frame forwards. Because you want it to fall backwards if anything goes wrong. Humans will be working around the front of the lifting frame doing things. Wall/machinery could be behind it. Yes Walls/Machinery are repairable.

  257. oiaohm says:

    Oldfart
    Yoy are a trip Fifi. You cite a special non commercial research license that Microsoft apparently uses for some of its offerings and compare it to GPL. As I said, entertaining bullshit.
    You did not read my second cite it over the home edition of MS Office.

    Derivative work clauses appears over everything Microsoft wants companies to pay more for. Problem here is the taint does not remove. GPL you release the source code and you can keep on using your program become you came into alignment with the License. Microsoft academic/home versions of software not so nice. Basically you should throw your work in bin and start over.

    http://www.iis.net/terms-of-use
    2. Grant of Rights

    (A) Copyright Grant – Subject to the terms of this license, including the license conditions and limitations in section 3, each contributor grants you a non-exclusive, worldwide, royalty-free copyright license to reproduce its contribution, prepare derivative works of its contribution, and distribute its contribution or any derivative works that you create.

    (B) Patent Grant – Subject to the terms of this license, including the license conditions and limitations in section 3, each contributor grants you a non-exclusive, worldwide, royalty-free license under its licensed patents to make, have made, use, sell, offer for sale, import, and/or otherwise dispose of its contribution in the software or derivative works of the contribution in the software.

    Sounds perfect.

    Until you notice this clause.
    (F) Platform Limitation – The licenses granted in sections 2(A) and 2(B) extend only to the software or derivative works that you create that run on a Microsoft Windows operating system product.

    Some get nasty. Like you make a program you port it to OS X this license does not grant any right to run on any other platform. Yes you can find yourself in court.

    Oldfart the Microsoft licenses are scattered with derivative works clauses that are way more nasty than GPL. GPL you release the source code you keep on using it. This one here if your program runs on Linux or OS X and window you are toast.

    DrLoser the commercial results from miss using any license can be catastrophic. GPL is no real exception. GPL is not as safe as BSD but its not as dangerous as some Microsoft ones. GPL worst case you release the source code and you can keep on shipping your product might give you a competition problem. Microsoft licenses and other commercial licenses that are not OSS cleared you screw up and you may never be able to ship your product again.

    The problem is like the IIS.NET example. Lot of the Microsoft licenses if you only read part of them everything looks golden but there is derivative work clause that takes everything granted back.

    The breach of the IIS.net one I find quite common these days. Why they have built their application to run under Mono and Windows. Just to be nasty that clause means making duel is break of license.

  258. oldfart wrote, ” The fantasy part comes in the notion that it is good for anyone other than those like yourself and Robert Pogson, who bask in the glow of your freebie.”

    French National Police: “The simple change from Windows and Office to Ubuntu and OpenOffice.org has saved us €2 million per year. And that’s not just down to licensing costs. The fact that we no longer need 4,500 dedicated departmental servers means that the savings just keep growing.” Not needing so many servers because GNU/Linux is flexible is not a “freebie”. It’s the right way to do IT. Why have 4500 servers that aren’t needed? What they probably refer to is M$’s limitation on what That Other OS is allowed to do with the clients. ie. a limit on sharing. Have more than X PCs? You need a server and a licence to run M$’s software on it.

    London Stock Exchange: “September 8th 2008 was one of the worst days ever for the London Stock Exchange (LSE), and high-end Windows server-based applications. That was the day that the LSE came to a crashing stop. What happened? While the LSE has never come clean on the whole story, my sources told me that the LSE’s Windows-based .NET TradElec stock exchange had crashed. What we do know is that the CEO who had brought Windows and TradElec in was fired, TradElec was dumped, and a Novell SUSE Linux-based platform was brought in to replace it. “ LSE is basking in the glow of FLOSS but it’s not about freebies. It’s about getting the job done. That Other OS doesn’t work for many people.

  259. DrLoser says:

    However, it’s a major step forward to hear you admit that VMWare “edited out the whole kernel of the thing.”

    This means, I presume, that you are admitting that all of your earlier claims to do with the interface between an ESXi Linux device driver and possible interaction with an OS kernel derived somehow from Linux were, in fact, the purest liquid form of nutritious substances squirted out the back end of a bovine with the accidental benefit of enriching the soil beneath?

    Or, in other words, Total BS?

  260. DrLoser says:

    Uh, they removed the whole kernel of the thing except for the drivers. That’s a major edit.

    Uh, no it isn’t.

    You will find many people quoting from the Old Testament, Robert. It is relatively rare that they will spew out all 39/46/51 Books.

    Sometimes you only need to acknowledge the Truth in Part.

  261. DrLoser says:

    Well, I say it sounds more realistic. You sound like the sort of chap that a Disney Princess would gravitate towards, Robert.

    A Corporate Lawyer?

    Not so much.

  262. DrLoser wrote, “no part of Linux whatsoever has been modified”.

    Uh, they removed the whole kernel of the thing except for the drivers. That’s a major edit.

  263. DrLoser says:

    1.Disney Princess says GPL is a minefield – avoid it
    2.Disney Princesssays GPL is a minefield – here is how to get through it
    3.Disney Princesssays GPL is a valid licence for software – embrace it
    4.Disney Princesssays GPL is a valid licence for software – here’s how it can work for you! Ride my Pink Unicorn!

    There you go, Robert. FTFY.

    And much more realistic it sounds, too!

  264. DrLoser says:

    This particular fantasy is worth repeating in full. It’s one of Robert’s:

    VMware could argue that each Linux driver is a “work” under copyright and they are entitled to use, examine, modify and distribute as they wish but that would be ignoring the fact that Linux is a collective work too and it includes those drivers so using a Linux driver without using Linux is an unauthorized modification of Linux.

    If Item A (a driver) is under the GPLv2 license, and Item B (a kernel) is also under the GPLv2 license, then it follows that, in order to use A in combination with B, you must follow the GPLv2 license insofar as the interactions between Item A and Item B exist. Which seem pretty tenuous to me, but that’s completely irrelevant.

    Should you wish to use and deploy Item A (the driver), whilst still respecting the GPLv2 license, and without any dependency whatsoever on Item B (the kernel), I don’t see how the license stops you doing so.

    It certainly isn’t “a modification of Linux.”

    Because no part of Linux whatsoever has been modified.

  265. DrLoser says:

    Speaking of incredibly flexible principles:

    Are Oracle correct in asserting that Google have breached copyright, as per their on-going court cases?

    How long are you all going to avoid answering this extremely simple and entirely relevant question?

  266. DrLoser says:

    There are a lot of cases involving lawyers and the GPL:

    1.Lawyer says GPL is a minefield – avoid it
    2.Lawyer says GPL is a minefield – here is how to get through it
    3.Lawyer says GPL is a valid licence for software – embrace it
    4.Lawyer says GPL is a valid licence for software – here’s how it can work for you

    I see we’re switching from our “programmers are the only important thing. Worship the Programmer!” mode to “lawyers are the only important thing. Obey the Lawyer!” mode, Robert.

    One day you will find a happy medium in that Venn intersection between programmers, lawyers, salesmen, and other corporate types. But I don’t imagine you will ever understand how a corporate lawyer works, based on that extraordinary piece of wishful thinking.

    Corporate Lawyers don’t care what you use, Robert. Corporate Lawyers care about two things, and two things only:

    1. I do not want to lose my job by signing off on any sort of contract with a hole in it.
    2. Other than that, I charge for my opinion, couched in suitably self-defensive terms. And charged at $50 per fifteen minutes.

    That’s, basically, it. And as I pointed out, Microsoft has Corporate Lawyers (very good ones). Surprise! And as I pointed out, Microsoft Corporate Lawyers have provided a wide range of opinions on internal use of software under various non-Microsoft licenses. Which shouldn’t be a surprise to you, but probably was.

    I have never heard of a Microsoft Corporate Lawyer making a single one of the completely absurd statements that you have just put into the mouth of a Generic Lawyer. I can’t imagine any other Corporate Lawyer, in any other Corporation, doing so, either. Let me explain the procedure to you, which I think is basically vanilla for all IT corps:

    1. Programmer (or team lead, or IT manager, or even CTO) wants to use Product X.
    2. Next guy up the chain consults Corporate Process. Corporate Process says, “get approval from our Lawyers.”
    3. {Several long boring email exchanges with Corporate Lawyers later} Lawyer says yes, or no. In a specific case, for a specific purpose.

    To a Corporate Lawyer, Robert, this stuff is almost literally life and blood. There are no principles involved, only money.

    Unfortunately, everybody else in the chain is either being denied the ability to use Product X (at best) or actually made to spend money, lots of it, in Lawyer Fees, to do so.

    I have one single point to make here, Robert.

    If you make it significantly more difficult to get a Corporate Lawyer to sign off on Product X, then a very large number of people are going to find an alternative to Product X.

    Whatever the legal merits of Nitwit vs VMware might be, the actual commercial results for the GPL are quite possibly catastrophic.

  267. DrLoser says:

    Of course they don’t except that Linux drivers have a “bus” concept so PCI or USB drivers share some common structures but certainly no Linux driver can control those other drivers except perhaps to load dependencies.

    It would have been simpler to admit that you were wrong, Robert. No (basic) drivers on any OS talk to the other drivers. That’s what the OS is for.

    Now, I say “basic,” and for some reason you didn’t spot that “flaw” in my argument. It isn’t actually a flaw, because the class of drivers that do talk to other drivers are composed in a driver stack, which ultimately doesn’t affect the argument.

    I’m surprised you didn’t notice it, though, because

    a. Obviously you are an expert on the architecture of modern low-level device drivers
    b. To have noticed it would have given you the opportunity to imply that ESXi somehow “steals” the Linux network stack.

    Oh well, another opportunity lost.

    VMware is running away with the ball when they claim supervisor roles for their kernel module.

    Really? How else would you run a hypervisor, Robert, without claiming such a role? Is there some way that other hypervisors or VMs can seal the deal without “claiming supervisor roles?”

    That would be a patentable idea, I feel.

    They are turning the OS inside out with a driver essentially being the kernel.

    No, they’re just letting a driver do what a driver does. And ESXi is not an Operating System. And the “kernel” of a hypervisor is a completely different beast to the “kernel” of an Operating System.

    And none of this matters in terms of the legal position vis-à-vis GPLv2, Robert. You quote the GPL quite often, but you don’t seem to bother reading it much.

    It doesn’t mention “roles” once. It doesn’t, for example, make a single mention of “an operating system” in any one of its versions.

    If the Sacred License doesn’t bother itself with these technicalities, what makes you think it worth anybody’s time to listen to you counting angels on the head of a pin?

  268. DrLoser says:

    Hoist platforms fall forwards not backwards if they are going to go any direction.

    Let me explain to you, oiaohm, what happens when a cable under tension, suspending maybe 5000+ N, is released. It doesn’t have to be a cable. It could be anything else under tension.

    Let me explain … no, wait, I don’t think I’ll bother. Let’s just say that a back foot is not an absolute necessity, but it’s quite useful when things don’t go precisely to plan.

  269. DrLoser says:

    A propos of nothing.

    I was just looking up “Charlie” quotes, what with being somewhat more irked by murderous Islamist loonies than one or two people around here, and I found an old Goons script.

    I must say, oiaohm bears an uncanny resemblance to Eccles…

    (Although obviously nowhere near as loveable.)

  270. oldfart says:

    ” you can save vast amounts of time and never have to reinstall Windows again…”

    Hmm. Let see. My last windows 7 system was running on the same image for 5 years untill I had to do a clean install of windows 7 as part of bringing it home for personal use. It replaced an 11 year old windows XP system that was still running its original image 2 years ago when I shut it down.

    Of course you will say I am lying, but then again that only shows your ignorance.

  271. oldfart says:

    “So you spend hundreds perhaps thousands and think your pompous ass is superior, totally righteous. ”

    Nope. I have specific needs that only the commercial software meets, and others where my investment in commercial software has paid and continues to pay my bills. Not everyone does the kind of work that can be done with Google apps on a chrome book.

    And as far as righteousness is concerned, I seem to recall one smug and ignorant fellow who actually suggested that pencil and paper was all that a professional composer needed to ply his craft. Tell me Dougie, did you actually explore the Finale 2014 site. (You might find the stories of who uses the software especially enlightening.) Or did you just arrogantly dismiss it as unnecessary?
    Eh?

  272. dougman says:

    Freebie you say?

    So you spend hundreds perhaps thousands and think your pompous ass is superior, totally righteous. You remind me of some of the old fools I run into, that are so thick that a mule kick in the head would not help in slightest.

    “We are doing it my way, my way!” …but, you can save vast amounts of time and never have to reinstall Windows again… “Oh, I heard of that Linux crap, its all open and hackers can infect your computer!”

    LOL.

  273. oldfart says:

    “Farting OLDMAN…”

    SO I guess you name is now DogBrain, eh?

    Fine with me.

  274. oldfart says:

    “Seems rather cut and dry to me, so that tells me the GPL has legal standing.”

    Oh its perfectly legal Dougie, where did you get the idea that I was saying anything else. The fantasy part comes in the notion that it is good for anyone other than those like yourself and Robert Pogson, who bask in the glow of your freebie.

  275. dougman says:

    Farting OLDMAN, GPL was valid enough for this case: XimpleWare Corp. v. Versata Software, Inc., 2014 WL 490940 (N.D.Cal. February 4, 2014)

    Plaintiff sued for $300M, defendant attempted to dismiss the case, Judge said “Nope, stand and face the music!”…so they settled out court for a undisclosed sum.

    Further reading, on other such cases: http://www.lexology.com/library/detail.aspx?g=9d3603a6-4e6b-49bd-bfc9-c615ec09bb0b

    Seems rather cut and dry to me, so that tells me the GPL has legal standing.

  276. oldfart says:

    “Lawyer says GPL is a valid licence for software – embrace it
    Lawyer says GPL is a valid licence for software – here’s how it can work for you”

    Pure fantasy Robert Pogson, though I am sure that you wish it so so that you can keep basking in the warm glow of your freebie, eh?

  277. oldfart says:

    make that

    “You are a trip Fifi.”

  278. oldfart says:

    “Fairly much as soon as you hear someone talking about GPL being a problem and backing Microsoft directly tells you they are incompetent and have never read the Microsoft licensing to know its laced with stacks of nasty conditions in the same class GPL contains.”

    Yoy are a trip Fifi. You cite a special non commercial research license that Microsoft apparently uses for some of its offerings and compare it to GPL. As I said, entertaining bullshit.

    But back to the topic at hand, you can babble till the cows come home, but in the end only the German court will count, and even then only if upheld all up the appeals chain.

  279. oiaohm says:

    oldfart Funny enough its not bullshit the post before one you quoted contains the cites.

    Fairly much as soon as you hear someone talking about GPL being a problem and backing Microsoft directly tells you they are incompetent and have never read the Microsoft licensing to know its laced with stacks of nasty conditions in the same class GPL contains.

  280. oldfart says:

    “Basically if companies listened to their legal departments they would not be running MS Office.”

    Such entertaining bullshit – You never cease to amaze Fifi.

  281. oiaohm says:

    Basically if companies listened to their legal departments they would not be running MS Office.

  282. oiaohm says:

    Oldfart
    No, Thats your opinion. Closer to reality is that there are a lot of business’s who listened to their lawyers instead of IMHO staying far away from the legal minefield that is the GPL completely.
    Oldfart of course is being narrow minded.

    http://research.microsoft.com/en-us/projects/pex/msr-la.txt
    You may not use or distribute this Software or any derivative works in any form for commercial purposes.
    Derivative work clauses appear in many Microsoft licenses.

    The most dangerous is in home usage restricted versions of Microsoft programs.

    https://www.microsofthup.com/hupuk/termsofuse.aspx?culture=en-GB#EEB

    Of course Oldfart and other want to make out this is just a GPL problem. You could have one of your staff using a home version of visual studio. It can also result in tainting your application under derivative work.

    Just a GPL issue I think not.

  283. oiaohm says:

    Basically if GPL is a License or a Contract or both all depends on what country you are enforcing it in. Civil law system is both.
    Civil Law
    License is the allowance to reproduce the work and the terms are Contract.

    Something like the USA. You have contract laws that rule specially over contracts. Then you have a copyright laws that rule specially over copyright licenses. Then you have patent laws that rule over patent agreements. So on and so on. This becomes a pure nightmare when you have something that is part patent, part trademark and part copyright all in 1 agreement what court do you have to hear it in hopefully not all of them. Under a Civil law system all that stuff is just contracts.

  284. oiaohm says:

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=569101

    Robert Pogson civil law countries there is no separation between License and Contract. Under civil law Contract/License is enforce by the contract law section of the civil law.

    Yes that paper I just pulled is one of the proof that GPL is enforceable in any EU country.

    Civil Law countries include Australia and all the EU including Germany….. Robert Pogson sorry I forgot Canada and USA are not civil law countries.

    Its like all the special loop holes that the USA has around copyright. Contract /License is a lot cleaner to enforce in Civil Law countries. Lot less ways to screw up writing a contact in a Civil law country. Like of someone looks of age and you let them sign in a Civil law country and the contract is fair its enforceable.

    For the VMWare case there in Germany there is no separation.

  285. oldfart says:

    “It’s the right way to do IT …”

    many people believe otherwise Robert Pogson, and thankfully we have choice.

  286. oldfart says:

    “Sorry Oldfart you are screwed here do you really want to keep on playing.”

    I’m not screwed at all by the likes of you sir. Your babble doesn’t count. The courts will decide.

  287. oiaohm wrote, “EULA and GPL are both contracts.”

    GPL is not a contract Contract law has lots of requirements to be a contract. GPL has none of those. It’s not a contract.

  288. oiaohm says:

    oldfart the think you should have worked out by now is would the courts have to decide for Xen or the Freebsd Linux driver wrapper is the kernel was contaminated or not. Answer is no you would not separation was simple simple to prove.

    Oldfart every time you say to me that the counts should have to decide something so something you should have pulling DrLoser up on for calling Christoph Hellwig a nitwit.

    Without question there is a case for VMWare to answer. It might find in VMWare favor but the odds are that it will not. Due to VMWare no doing the right things its going to be very hard to VMWare to prove that there codebase is not tainted.

    Oldfart can you say without question that the vmkernel design has not be effected by only have vmkLinux for particular drivers.

    The reality is I cannot 100 percent predict a judge/jury . No one can. Common cases like this fail what is called the too cute arguement.

    Until ESXi 5.5 particular driver types were only provided by one source. The VMK interface is in fact evidence of the vmkernel being modified to suite Linux drivers. As Linux kernel gets features vmkernel gets the same features.

    Sorry Oldfart you are screwed here do you really want to keep on playing.

    The reality here is the vmkernel has been effected by the vmkLinux. The question will come down is that fair usage and does that fall outside the GPL contract.

  289. oldfart, over-simplifying wrote, “there are a lot of business’s who listened to their lawyers instead of IMHO staying far away from the legal minefield that is the GPL completely.”

    There are a lot of cases involving lawyers and the GPL:

    1. Lawyer says GPL is a minefield – avoid it
    2. Lawyer says GPL is a minefield – here is how to get through it
    3. Lawyer says GPL is a valid licence for software – embrace it
    4. Lawyer says GPL is a valid licence for software – here’s how it can work for you
    5. There may be many more combinations as larger organizations may well have rafts of lawyers with diverse opinions and specialties. SCOG v World clarified GPL for any lawyer interested in copyright and software so I expect that large organizations would have a much better understanding than “fight or flight” reflexes all these years later. It’s lawyers, for instance, who are kicking the software-patent monster into the ditch. It’s a lawyer, for instance, who may have a family with every member using Android/Linux. There are very few corporations not using quite a bit of FLOSS these days. It’s the right way to do IT and if they had it to start over again, there’d be a lot more FLOSS in use in corporations. Many are using as much FLOSS as possible in “greenfield” developments. Many are hiring as many Linux/FLOSS people as they can find.

  290. oldfart says:

    “What do you think VMware’s customers will think of using all that beta-software? ”

    We think far better of it than we do of using the free crap that you do. VMWare is in business for real, and has a better track record producing software than the basement programmer army that you get freebies from.

  291. oldfart says:

    “Oldfart can you say without question that the vmkernel design has not be effected by only have vmkLinux for particular driver.

    Nope, but neither can you Fifi. The courts will decide.

  292. DrLoser wrote, ” If I’m VMware, I’m setting up a team of ~40 engineers in a clean room as of midnight on the day that Hellwig came up with this nitwit plaint, and I’m telling them to work on the backup plan. That’s about $5 million down the drain, but unlike you, Robert, large corporations are prepared to stump up money for their rights.”

    What do you think VMware’s customers will think of using all that beta-software? There are 2774 files in linux-3.18/drivers/*/*.c. Not all are drivers but a lot are. So, if it takes 1 man-year to develop them in a clean room and test them on all their hardware and debug them, and a man-year costs ~100K, that’s $277 million, affordable for VMware. They may not have been able to afford that years ago but now they can. They’ve also developed their own OS/kernel/whatever, so it’s costing them a bunch either way(10-K: “Our research and development expenses were approximately 21% of our total revenues during both 2014 and 2013.”). How much of that year are their customers going to wait? Can they actually hire 2774 driver programmers? They might have very high recruitment costs on top, a blow to the bottom line. Investors will notice. In the last 10-K, VMware wrote, “While no formal legal proceedings that we expect to have a material impact on our financial condition or results of operations have been commenced, there can be no assurance that actions will not be taken in the future, and final resolution of such claims could be materially different from our current estimates. Furthermore, because litigation and the outcome of regulatory proceedings are inherently unpredictable, it is possible that our business, financial condition or results of operations could be negatively affected by an unfavorable resolution of one or more of such proceedings, claims, demands or investigations.” That could change rapidly. I think a lot of investors will be disappointed to learn that VMware knew Hellwig was unhappy for the last 8 years. Surely, VMware should have expected legal action eventually, almost certainly, with such non-routine bending of the rules.

  293. oiaohm says:

    Speaking of which, did you put that back foot on the tractor hoist as I recommended, Robert? It doesn’t have to be permanent. You could weld in a nice little hinge for convenience.

    DrLoser I missed this. Hoist platforms fall forwards not backwards if they are going to go any direction. Most cases putting a back leg on does not help at that much. Make wider with front legs extending way past the hoist point so making the hoist point center of gravity and larger foot print increases stability. One of the ones we had in a workshop made out of box steal like Roberts had slide in extensions to front feet they could also be used as back feet.

    DrLoser just for future reference for these kinds of things never use a hinge particularly for items you might get tempted to put weights on . Slide in feet into the box is safe. Hinges are structurally weak compared to Mortise and tenon. Yes Mortise and tenon plus a pin is your best option for anything heavy load if you are going to have a joint. You might want to put a few sandbags on that back leg.

    Slide in extension to front legs can also be slide in back of front legs to provide back-legs to put weight on when you don’t have enough front operational space. I have used one that disassembled all bar the L frame. The stabilizers were all turnbuckles. No hinges anywhere.

  294. DrLoser begged, “I’m guessing an awful large number of major corporations are going to say no to a GPL license and yes please to an MIT license.”

    You’d love that, eh? I’m thinking an awful large number of major corporations are going to switch to Xen or KVM so they can avoid going down the drain with VMware. Unlike operating systems, virtualization is by its nature rather open, so migration away from VMware should not require too much effort compared to a court order demanding cease-and-desist using VMware. Just the hint that VMware might lose this case could cause migrations. That’s why VMware’s casual attitude to Hellwig is so surprising. If your house is on fire, the last thing I expect from you is to deny that’s so. VMware’s house is on fire. All they had to do was reveal the source code to put out the fire. The damage will be much greater by the time the court rules. I don’t think VMware can function without Linux drivers and I don’t think they can rewrite all of them, so they have to start over/shut down/release the code. The best option is to release the code but they won’t just to spite their face.

  295. DrLoser wrote, “did you put that back foot on the tractor hoist as I recommended, Robert?”

    Nope. It was sufficiently stable. The only real concerns in use were that it was darned heavy requiring a lot of effort to position and it was a bit too flexible. The steel members I used were not as stiff as I estimated. Still, there was not a creak or groan from the thing. It worked beautifully. I had to drive the tractor in at an angle to get the wheels straddling the legs and the engine was a little off-centre but it all worked out with not much difficulty. I had to reposition the engine a bit because the pulleys were not aligned well enough but that took very little time once I figured out where to apply my lever… My previous hoist was much lighter because it was made of stronger/lighter pieces. This is one of the downsides of using scrap. You get what you have. Still it cost me nothing but time and a few pounds of welding rods and I love welding. The cold was the big issue with the welding. Below about -10C, it was too painful. Even at 0C my time on task was limited. My welding gear is not designed for the cold.

    I’ve run into a couple more problems with the tractor. I’m missing 3 bolts. It’s not a big deal. I can likely buy what I need locally for a few dollars. The battery box interferes with the tool box of the thing and the air-intake filter. The throttle linkage seems to be for a different engine. I can modify it easily. It’s just a piece of stamped sheet steel. I can cut it and reweld it to fit my engine. A bigger problem is that the bolt that holds the bushing on which the linkage pivots is a few mm too short. With bushing on the bolt it can’t reach its threads. So, I need 4 bolts of 3 kinds and a bit of fabricating. It’s all good fun. I like making things.

    The wiring was interesting. The supplier gave no wiring diagram and I had to ask for images of the wiring as assembled. The puzzle was a “black box” with several wires running in with me having no idea what it was. It turns out that it’s a rectifier and the fan of the radiator also is a generator, so I won’t have to charge the battery. None of this information was in the user’s manual because I ordered the biggest unit and they sent the manual for the next smaller unit… The electrics now work and I could start the engine tomorrow. It’ll be loud because I can’t bolt on the muffler without those bolts, so I’d better take a shopping list to town.

  296. oiaohm says:

    Robert Pogson some drivers for Linux are officially independent works legally. They have wrappers for Linux, Freebsd, Solarias in the one code base.

    But if a driver is a independent work or not is if separation is provable. Yes the really simple question will this driver work without the Linux copyrighted works if yes its a independent work. Intel produces a few that ship with the Linux and freebsd kernel so are truly independent works.

    It is a minority of Linux drivers that are independent works.

  297. oiaohm says:

    Robert Pogson you don’t understand contract.
    http://en.wikipedia.org/wiki/Software_license
    Yes a majority of Software license are contract law.
    http://en.wikipedia.org/wiki/Oral_contract

    A contract does not have to be written on paper.
    One can give a total stranger a licence to use/run/examine/modify/distribute software without having a contract with them.
    This even if you don’t write it down this is still a contract. Contracts form all the time. All you have todo is prove the person knew about it so had to agree to it and its a contract.

    There is a massive number of things we do every day that fall under contract law.

    EULA and GPL are both contracts. One is written to restrict more than the other. Even BSD license is a contract.

    A contract has a bunch of legal rules that apply: legal age, sanity, meeting of the minds, consideration, etc.
    This is another lot of things people get wrong about contracts.
    A software licence does not require that.
    True software license does not require every limitation. Neither does every contract.

    http://nationalparalegal.edu/public_documents/courseware_asp_files/domesticRelations/FamilyRelationships/Contracts.asp
    Note the clause about necessaries.

    Every restriction rule that a person has to be to sign/agree to an contract there is a clause to by pass it.

    legal age, sanity, meeting of the minds, consideration, etc… Most don’t apply to contracts over software. Legal age, Sanity, meetings of the mind, consideration… is mostly bypassed by the contract not being targeted to 1 unique individual. The theory is a non targeted contract can have a class action performed if there is something super wrong with it. So the individual person agreeing only comes in to how much in damages they own.

    The anti-class action clauses in some EULA do raise some serous questions about future enforceability.

  298. DrLoser wrote, “How does a Linux driver interface to all other drivers in the Linux kernel?”

    Of course they don’t except that Linux drivers have a “bus” concept so PCI or USB drivers share some common structures but certainly no Linux driver can control those other drivers except perhaps to load dependencies. VMware is running away with the ball when they claim supervisor roles for their kernel module. They are turning the OS inside out with a driver essentially being the kernel. It’s clearly a new way of using a Linux driver far beyond what it is intended to do. Linux drivers are intended to be dependent on the Linux kernel and vice-versa.

    VMware could argue that each Linux driver is a “work” under copyright and they are entitled to use, examine, modify and distribute as they wish but that would be ignoring the fact that Linux is a collective work too and it includes those drivers so using a Linux driver without using Linux is an unauthorized modification of Linux. Take away the Linux part and you lose the licence to use Linux and its parts. A Linux driver is not a standalone work because it communicates with certain hardware/software and the Linux kernel. Having a licence to modify a chapter of a book does not give you permission to rewrite the book and then claim it is not a derivative book.

  299. oiaohm wrote, “Most people fail to know that a software license is a Contract.”

    No, it’s not. One can give a total stranger a licence to use/run/examine/modify/distribute software without having a contract with them. A contract has a bunch of legal rules that apply: legal age, sanity, meeting of the minds, consideration, etc. A software licence does not require that. It’s just permission. M$’s EULA is a contract entered into for the purpose of giving a licence to use software whilst agreeing to enslavement. FSF’s GPL is a licence given. No contract is involved. That licence can be given to minors, insane people, people barely literate, etc. without mutual agreement, a signature, payment or other consideration.

  300. oiaohm says:

    The biggest mistake people make about copyright licenses is having the idea if they get caught doing it wrong they can fix it after the fact.

    Contract Law does not allow for fixing major breaches of contract after the fact. Contract you break it you are liable. What you do after you break the contract does not reduce what you are liable to pay for breaking the contract.

  301. oiaohm says:

    http://www.virtuallyghetto.com/2013/10/esxi-55-introduces-new-native-device.html

    And as you will also note, they are also transitioning to native drivers.
    Oldfart too late. Legally too late.

    If the vmkernel is ruled to be a derivative work the new native interface is also GPLv2 with extension.

    Oldfart can you say without question that the vmkernel design has not be effected by only have vmkLinux for particular drivers.

    Xen developers and many other developers working with GPL have become very aware about the provable separation.

    Over 80 percent of the drivers VMWare wanted are in BSD kernels.

    http://info.iet.unipi.it/~luigi/freebsd/linux_bsd_kld.html

    Note this is a wrapper done properly. Freebsd has its own drivers before it writes a wrapper to import Linux drivers. Provable separation include provable independent design so GPL cannot be made apply to the Freebsd kernel.

    What is really is GPL.

    Copyright allows you to attach terms under what conditions a person has to agree to so they are able to to use the work. These attached terms are Contract law not Copyright law.

    GPL the conditions include derivative work in the copyright works license conditions. Why is GPL called viral is the derivative work term means by using the source code you have agreed to protect the ideas.

    Copyright itself cannot protect the Idea in a code base. The license of the copyright can protect the idea because you have agreed to that term by using the copyrighted work.

    One is prove copyright. Next is enforce terms of License. Enforcing terms of License comes under contract law so can be used to protect anything that is legal.

    Most people fail to know that a software license is a Contract. Copyright cannot protect an idea because Copyright does not cover that.

    You want to protect an Idea. You need either a Patent or a Contract. Contract is only enforced on those who agree to it. To copy/use(ie copyright the right to copy/use) a GPL work you have to agree to the connected license/contract if you do not agree you don’t get the right to copy/use.

    GPL is part Copyright Law and part Contract Law. Its the Contract Law parts of GPL that make it viral.

    GPL is not a major threat to companies if its handled correctly. VMWare has handled GPL code very poorly.

  302. DrLoser says:

    Think of it. How does a a Linux driver interface to all other drivers in the Linux kernel for some non-Free software? That’s not what Linux drivers/modules do.

    I’ve thought about it, Robert. I trust your expert opinion on these things. How does a Linux driver interface to all other drivers in the Linux kernel?

    Buggered if I know. For the twenty-odd years that I have worked with Linux, I was under the peculiar misapprehension that it doesn’t.

    Naturally, I am an ignorant outlier on this. There are no doubt copious cases in the literature whereby one can find instances of one driver (OS not necessarily relevant) “interfacing” with another driver.

    It is a very, very, common use case. Which no doubt you will explain to us in terms that we can understand. Oh, wait a minute. I don’t wish to reserve Fridays for Fifi.

    BWAHAHAHAHAHAHAHA!

  303. DrLoser says:

    By the way, Robert.

    If you’re prepared to work up a Kickstarter campaign to get Gillian Anderson to deliver the final arguments in Nitwit vs VMWare, be assured of my fervent support.

  304. DrLoser says:

    /*
    * Radix tree node definition.
    */
    #define RADIX_TREE_MAP_SHIFT 6

    That, there, is sufficient to differentiate the Expression from the Idea.

    And don’t make me go all T.S.Eliot on your butts …

  305. DrLoser says:

    And you know what my “Exhibit One” comes down to?

    Christian Hellwig: You stole my small alterations of the Radix Tree away from me! I have hated you since 2006! I want my mommy! All 369 lines* of that were mine, apart from the license header of course, and then again there’s a bunch of other comments but those are obviously copyright, and then there’s the corfe bits of the algorithm which can probably be mapped to 100 lines, and don’t forget the rest of it, although I will, because the original bit was written by Momchil Velikov who is not a party to this case.

    I WANT MY MOMMY!

    Now, interestingly enough, and fatuous as I find this third-hand argument from an obvious Entitlement Complex to be … It probably has legs in Court.

    As oldfart says: let’s wait and see, shall we?

    * Yes, I counted them. wc is the single most significant tool that FLOSS has ever provided to the IT world.

  306. DrLoser says:

    Speaking of incredibly flexible principles:

    Are Oracle correct in asserting that Google have breached copyright, as per their on-going court cases?

    Let the peanut gallery decide!

  307. DrLoser says:

    Of course the blood sucking drones on this site should take heart, there are probably enough productive true believers to keep them basking in the glow of their freebie for the rest of their lives.

    Don’t discount the possibility that they will do a “herd switch” and accept the MIT license, oldfart. Never forget, these are people of incredibly flexible principles.

    The only two rules they live by are:

    1. It doesn’t cost me anything at all. Hey, I can even pester people with pointless and annoying and expensive bug reports! I am The Man! Other people are sheeple!
    2. Somebody else lost money or died. Well, it’s their fault for being in the wrong building at the wrong time.

    Sheep or Goat? Well, there are many possible alternatives to choose between when it comes to the Deity.

    Oh yes, and I almost forgot.

    3. There’s a coherent and fairly neutral argument about the four possible ways that VMware can defend, and possibly lose, the lawsuit in question here. I have been given ample opportunity to make a reasoned argument on each one.

    But, you know what? I believe in FLOSS. FLOSS is my Saviour. I just can’t be bothered to be rational.

    Speaking of which, did you put that back foot on the tractor hoist as I recommended, Robert? It doesn’t have to be permanent. You could weld in a nice little hinge for convenience.

  308. oldfart says:

    “I’m guessing an awful large number of major corporations are going to say no to a GPL license and yes please to an MIT license.”

    Of course the blood sucking drones on this site should take heart, there are probably enough productive true believers to keep them basking in the glow of their freebie for the rest of their lives.

  309. DrLoser says:

    The kernel module is licensed GPLv2 and so should their kernel be.

    a. Who says?
    b. It’s not an OS kernel.

    You really have no clue about any of this at all, do you, Robert?

  310. DrLoser says:

    Another interesting thing about OpenCV is that it is currently on version 2.4.11. (Yes, I’m downloading it even as we speak. Or in the case of Fifi and Dougie, stumble over the long words.)

    Now, I fully accept that FLOSS has produced a decent number of packages that pass the barrier of version 1.0. The kernel springs to mind. (Linux, not Hurd.) Other substantive things like the major libraries. All of the databases. Even GiMP.

    But, let’s just theorise here. You, as I was at the time, are sort of casting around for possible pre-packaged solutions to a slightly obscure problem.* Why, let us resort to apt-get!

    I suspect that you are not going to be impressed by the maturity of the results. Sometimes the MIT license beats the GPL license shitless.

    * Putting a big purple circle round the noses of people in a JPEG or PNG attached to a RSS news feed, since you ask. Seriously. That’s what I used it for.

  311. DrLoser says:

    I’ve accused oiaohm of straying off-subject, and I’m about to do the same. (I meant it in a friendly way, oiaohm.) One more interesting point from oldfart, then:

    One can only hope that the resulting debacle provides a wake-up call to commercial software producers to think twice before relying on GPL’d code.

    As the Resident Microsoft Troll here — ie, I have three years’ work experience with Voldemort — I think I should point out the following interesting fact about Microsoft’s (very high-powered) lawyers.

    If you push them hard enough, they will accept an Apache license. If you push a little harder, they will accept an MIT license. (I know this, because I pushed them to accept the license for OpenCV, a remarkable collection of 2,500+ AI/robotics computer visualisation algorithms issued under the MIT license. I’d recommend OpenCV to ram, who might find it useful. nobody else here would, though.)

    I’m told you can even push them to accept a GPLv2 license, although after a while trying, you probably lose the will to live.

    Now, if the result of Nitwit vs VMware goes the way you devoutly hope, Robert, what do you think is going to happen?

    I’m guessing an awful large number of major corporations are going to say no to a GPL license and yes please to an MIT license.

    Remind me again which license applies to, ahem, FreeBSD?

  312. DrLoser says:

    It’s interesting that an “operating system” called VMKlinux has a lot of drivers but there’s no mention that many of those drivers are Linux drivers, code from the Linux kernel… FSC/Hellwig should use this document as Exhibit One.

    No, that would be Exhibit Three, Robert. (“Linux device driver code.”)

    Strangely enough, one’s legal team is not required to prioritise “exhibits” by external standards of importance. You believe that “stealing Teh Codez” is the most important thing in this case. And perhaps it is.

    But from a legal perspective, and if this really is the hammer-blow you think it is, a professional group of experienced lawyers are hugely unlikely to cite it as “Exhibit One.”

    “Exhibit One” ideally creates Fear, Doubt, and Mistrust in the minds of the jury and/or judge. Something as small as the Radix Tree would do for this.

    “Exhibit the Last” ideally provides concrete evidence of undeniable malfeasance. Which in this case … I haven’t yet analysed the code … would be the SCSI drivers.

    Knocking off a confusing array of Linux device drivers — not confusing to you, Robert, because you have Read, Modified, and Distributed every last one of them, but potentially not a clarifying issue to a jury and/or judge — is something you bring up in between the corker of a start and the HBO Season Finale of the Court’s Devastating Bring-Down.

    Is that going to feature Gillian Anderson, by the way? I’m only going to watch it if it features Gillian Anderson.

    I’m a huge fan of Gillian Anderson’s … exceptional … Legal Assets.

  313. DrLoser says:

    Thank you doctor, but I will stick with with my observation that Lil’Hammie is nothing more than a proven liar and fraud.

    The only thing that would add, is that he is also a joke.

    Fifi is wasted on this site, then. It’s not over-encumbered by people with a noticeable sense of humour.

  314. DrLoser says:

    Actually, and given oldfart’s interesting observation that a victory for Hellwig et Nutters Anonymous is not likely to go down very well with other large corporations who may be skirting around the edge of the legal minefield that is the GPL …

    … Does anyone out there have any information on what the Linux Foundation’s attitude to this suit is?

    Either I’ve missed it, or Jimbo “Primo Marketroid, Never Mind The Brain-Fart, Check Out The Hair” Zemlin has been suspiciously quiet on the issue.

  315. DrLoser says:

    So, vSphere still uses Linux even if it’s only the drivers and they are GPLv2 largely, so VMware is violating the terms of the licence.

    You haven’t listened to a word I said, have you, Robert? Although thankfully we seem to have convinced you that vmklinux is not a “Linux clone.” And that it’s not even an “operating system.” Well, that’s a start.

    I hope to get on to SCSI drivers soon (it’s going to require actual research), but for now my opinion remains that — with the caveat that VMware may be found guilty of breaching copyright on the headers — there is no obvious way that they are breaching the GPLv2 by offering the ability to “repurpose” a third-party Linux driver. All they have to do is to make the license prominently available, and to provide free access to the source code.

    They seem to be doing that. Is it enough? I haven’t heard you arguing that they need to do more.

    The drivers… Can vSphere do everything without the Linux drivers? Nope. How many native drivers does vSphere have and how many are needed in the real world?

    This falls under the category of “who cares?” Tell me, Robert, if somebody had offered you a tractor for free, would you have bothered to buy one? You wouldn’t, would you?

    Equally, if a court ruling (anywhere in the world) gives VMware the choice of developing non-GPLv2 drivers, or of “opening the kimono” to the Linux Foundation (to which, weirdly, they are signed up), what do you think they are going to do?

    Do you know how long it takes to develop a basic vanilla device driver for a network card or a USB device — these being what we are talking about here? Roughly six months, that’s what. If I’m VMware, I’m setting up a team of ~40 engineers in a clean room as of midnight on the day that Hellwig came up with this nitwit plaint, and I’m telling them to work on the backup plan. That’s about $5 million down the drain, but unlike you, Robert, large corporations are prepared to stump up money for their rights.

    VMware has ~14K employees. How many of them are programmers cranking out replacements for Linux drivers that are not derivative?

    What is it with you and your obsession over the percentage of staff in any given organisation dedicated to the IT coal face? Sometimes you are like a dizzy little schoolgirl when it comes to a large quantity of sales staff and marketroids (see, Ubuntu). Sometimes you are dead against the very existence of such people (see, Microsoft).

    Is it written that a 14K company should rigidly demarcate its workforce according to monkey-like specialisation? Is there a Golden Mean? Where, O Master, can we find the Stone Tablets that will Enlighten Us?

    Nowhere, that’s where, Robert. Your crippled understanding of modern IT business means that none of this should be any concern of yours. Not with Ubuntu, not with Microsoft, not with VMWare.

  316. oldfart says:

    ” VMware has ~14K employees. How many of them are programmers cranking out replacements for Linux drivers that are not derivative?”

    The real question is how many of those drivers are even relevant to vSphere? I’ll wager that there are very few. More relevant is the fact that there are a lot of third party hardware creators who are also selling hardware into the market that uses vSphere . All those third parties have to do to maintain their investment is re-structure their drivers to respond to calls from VMWares native driver interface. While even that subset of drivers represents a large number, it isn’t insurmountable given time money and resources of VMWare and its OEM’s.

  317. oldfart says:

    “That’s nonsense. ”

    No, Thats your opinion. Closer to reality is that there are a lot of business’s who listened to their lawyers instead of IMHO staying far away from the legal minefield that is the GPL completely. They have done what they were assured was the safe way of leveraging FOSS. Now if VMWare loses their battle, all of those business may get to watch their intellectual property be forfeited. One can only hope that the resulting debacle provides a wake-up call to commercial software producers to think twice before relying on GPL’d code.

  318. oldfart wrote, “if even your cite of a linux partisan is admitting that vSphere no longer uses Linux, this would suggest that you really need to wait from the trial outcome before you come to any conclusions.”

    The drivers… Can vSphere do everything without the Linux drivers? Nope. How many native drivers does vSphere have and how many are needed in the real world? So, vSphere still uses Linux even if it’s only the drivers and they are GPLv2 largely, so VMware is violating the terms of the licence. VMware has ~14K employees. How many of them are programmers cranking out replacements for Linux drivers that are not derivative? I’ll bet many are salesmen, not programmers. Everyone finds Linux programmers are scarce. That will apply to VMware too. I anticipate that formal discovery could be quite swift in this matter because it’s a well-defined piece of code that is demanded. When the court reads the code in Linux drivers embedded in VMware’s stuff, it will be game over. Hellwig or others will be able to help the judge find matches and they will be all over the place. I expect Hellwig and others have already done some espionage to have come this far.

  319. GPLv2:“These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.”

    This is the “viral” aspect of which oldfart complains. The key is “independent”. Is VMware entirely independent of Linux drivers? No. Some parts depend on them, like their kernel module and their kernel. The kernel module is licensed GPLv2 and so should their kernel be. VMware’s kernel can’t do much in the real world without those drivers. The “native” driver thing is late to the game and the world is not rushing to create enough native VMware drivers to matter in random servers. A particular large user might be able to write a dozen or so drivers to suit his data-centre but the vast majority of users will not waste their time. So, this continues to be an issue even if VMware claims the suit is without merit. Just about every defendant’s mouthpiece says that as a layer of the defence against claims. If the court agrees nothing more needs to be done. That’s pretty rare, just frivolous suits or “lack of standing” suits come close. VMware will have to show the code to defend itself and hiding stuff from the court will be deemed ceding the point. VMware will have to defend its position and I don’t think they will succeed. Using Free Software was a huge benefit to VMware in saving them from writing a lot of code. They should follow the rules laid out in GPLv2.

  320. oldfart wrote, “It should demonstrate that all of that so called “free” code is not really free and should be avoided like the plague if one wished to benefit from ones own work.”

    That’s nonsense. The whole world is mixing FLOSS and non-Free software quite happily because it works for them. You just can’t mix it in a single work. That’s obvious to anyone who’s held two books. Both have copyrights held by the respective authour. Everyone knows one authour should not include stuff from the other’s book. There’s no “contamination” from running a non-Free application on GNU/Linux. There’s no “contamination” from linking a driver to standard Linux headers(“This copyright does *not* cover user programs that use kernel
    services by normal system calls – this is merely considered normal use
    of the kernel, and does *not* fall under the heading of “derived work”.
    Also note that the GPL below is copyrighted by the Free Software
    Foundation, but the instance of code that it refers to (the Linux
    kernel) is copyrighted by me and others who actually wrote it.
    Also note that the only valid version of the GPL as far as the kernel is concerned is _this_ particular version of the license (ie v2, not v2.2 or v3.x or whatever), unless explicitly otherwise stated.
    Linus Torvalds”
    ). These guys were trying to have their cake and eat it too. Xen did the right thing. Why didn’t VMware?

    I just checked VMware’s compatibility list. For drivers they send them to the manufacturer. For Intel’s 82540EM, say, Intel lists drivers for DOS to that other OS with lots of GNU/Linux in between and no VMware…

  321. oldfart says:

    At the end of your cite I found an interesting comment.

    “But, at the end, perhaps we’ll have a clearer idea of what constitutes a derived product of the kernel; that could be seen to be a useful service even if the enforcement effort itself fails.”

    While I am sure that you are hoping for a windfall of proprietary code that you can bask in the glow of, I am hopeful that this will become an object lesson to commercial companies who think that they can have their cake (Free code) and eat it (keep source closed). It should demonstrate that all of that so called “free” code is not really free and should be avoided like the plague if one wished to benefit from ones own work.

    I believe the old saying was:

    “He who would sup with the devil best bring a long spoon…”

  322. oldfart says:

    “Well, that’s easy for them to say but when you look at their publicly available vmklinux module,…”

    I looked at the article and found this little gem:

    “Update: in truth, newer versions of ESXi no longer need the initial Linux bootstrap; in current versions, vmkernel boots directly.”

    So if even your cite of a linux partisan is admitting that vSphere no longer uses Linux, this would suggest that you really need to wait from the trial outcome before you come to any conclusions.

    I certainly am.

    Which matches my understanding.

  323. oldfart wrote, “This article describes what they did up to vsphere 5.5”.

    Well, that’s easy for them to say but when you look at their publicly available vmklinux module, “The picture that emerges suggests that vmkernel is not just another binary-only kernel module making use of the exported interface. Instead, VMware’s developers appear to have taken a substantial amount of kernel code, adapted it heavily, and built it directly into vmkernel itself. It seems plausible that, in a situation like this, the case that vmkernel is a derived product of the Linux kernel would be relatively easy to make.” Think of it. How does a a Linux driver interface to all other drivers in the Linux kernel for some non-Free software? That’s not what Linux drivers/modules do.

    See smurf’s comment on TFA:“vmware used neither the syscall interface nor the module interface (which is usable by closed-source modules, if you buy the legal fiction that since some symbols are exported GPL-ONLY, the rest must therefore be exempt from that restriction); they simply created their own. Which makes their code a derived work.”

  324. oldfart says:

    “Did they write their own in a clean room? I don’t think so. ”

    Actually, This article describes what they did up to vsphere 5.5

    http://www.virtuallyghetto.com/2013/10/esxi-55-introduces-new-native-device.html

    And as you will also note, they are also transitioning to native drivers.

  325. oldfart says:

    “oldfart its not specialty bull crap. I have operated ESXi 5.5 servers. Maybe its gone from ESXi 6.0.”

    Sure you have – not. You are an admitted and proven liar and fraud whom no one believes much any more.

    As I said, cut the speculative bullshit and leave it for the court.

  326. oldfart wrote, “How can you tell from this document that they are from the kernel Robert Pogson?”

    In the document, ESXi Architecture, the word “driver” occurs 6 times. reading:“Functionally, ESXi is equivalent to ESX 3, offering the same
    levels of performance and scalability. However, the Linux-based service console has been removed, reducing the footprint to less than 32MB of memory. The functionally of the service console is replaced by new remote command line interfaces in conjunction with adherence to system management standards.”

    No mention of drivers at all… Where did they go? They were in Linux in ESX. TFA suggest Linux is gone but they are still driving diverse hardware all over x86-64. Did they write their own in a clean room? I don’t think so. I think they are still using the ones from Linux. It’s not impossible that VMware could write a mess of drivers but it would cost $billions and the Linux drivers are out there. It’s generally acceptable to write non-Free software drivers for Linux but I don’t know anyone who thinks it’s OK to plug a GPLv2 Linux driver into something else.

  327. oiaohm wrote, “Fairly much anyone who been using esxi you can almost bet they have committed copyright infringement some time or another most likely without even really knowing.”

    If what is being infringed are the drivers from Linux, then stuff is hitting the fan. Wealthy corporations paying big bucks to VMware don’t want to be exposed to legal liability for copyright infringement. This case could drop the curtains on VMware’s love affair with big business. A huge chunk of Linux is code for drivers all distributed under GPLv2. If VMware distributes those drivers for use with VMware’s stuff and not for Linux, they are using a derivative software and violating GPLv2 which requires source code. I only see source code for a handful of drivers in VMware’s download. If they require users to butcher Linux in house (grab Linux drivers to use for VMware), that may not be distribution in most cases but it sure is stretching copyright/fair use. We’ve had rulings that APIs can’t be copyright-protected but this is the mirror-image of that, using drivers designed for the Linux API but not in Linux. I think that makes VMware’s code a derivative of Linux. It sure looks like they are surrounding Linux code in something. How is that not a derivative? How is that different from copying and pasting extensively from Linux into VMware?

  328. oiaohm says:

    oldfart deprecation is first stage before removal.

    To be correct the old console is deprecated. Take quite a few versions before vmware does full removal of stuff that is deprecated. So at 5.5 the status of the old interface was disabled/deprecated. It might be removed in 6.0 have not checked.

  329. oiaohm says:

    oldfart its not specialty bull crap. I have operated ESXi 5.5 servers. Maybe its gone from ESXi 6.0.

  330. oldfart says:

    “Yes it possible to spin up the old ESX full blown service console on a newer ESXi.”

    Your ability to fantasize knows no bounds. Spare us the speculative bullshit and leave the work to the courts.

  331. oiaohm says:

    oldfart the complete case is talking about ESXi not just ESX. Yes there is a difference.

    https://pubs.vmware.com/vsphere-51/topic/com.vmware.vcli.getstart.doc/cli_jumpstart.3.5.html

    But what is shown is certainly NOT the the full blown service console that was a feature of ESX server until it was removed and replaced with a limited minimalist console that is only meant to be access from the vSphere physical console EXCEPT in emergencies.

    The esxcfg crap is in fact all still in there. Yes it possible to spin up the old ESX full blown service console on a newer ESXi. But you have to manually spin the bastard up and its not properly maintained so may just make matters worse not better if you use it. Oldfart have you never been in the ESXi console and seen what is really still in there. The answer is disabled not stripped life would be a lot safer if old was stripped out and not left their as interfaces for old scripts and macros and the like.

  332. oldfart says:

    “Here is a super good one. ..”

    In the end your super good one is pure speculation. We shall wait and see what the german court decides.

  333. oldfart says:

    “This prior cite I posted. Shows clearly that the console is just disabled.”

    It shows that a console process is disabled. But what is shown is certainly NOT the the full blown service console that was a feature of ESX server until it was removed and replaced with a limited minimalist console that is only meant to be access from the vSphere physical console EXCEPT in emergencies.

    I made extensive use of the service console when it was a major feature of ESX. I have also used the post Service console version of the VMWare hypervisor (ESXi) now vSphere. that console has very limited function in comparison to what it replaced.

  334. oiaohm says:

    Here is a super good one. Name 1 third party hardware vendor who provides a driver for ESXi. The answers almost none do. Then how many of those drivers use just the VMK interface the answer is none.

    Remember how VMWare attempts to say that VMK interface is for third party drivers yet no third party uses it. Is it not strange that there are no third party drivers using pure VMK interface. Kinda because it is featureless.

    http://blog.kihltech.com/2014/01/how-to-add-intel-nic-drivers-to-an-esxi-5-5-iso/

    Now this above is what is going on. Linux Drivers that are only tested against the Linux kernel are build rebuilt against the modified vmkLinux system. Note the driver e1000e that I mentioned before as being trouble. Can you now see why. Using not properly tested drivers is going to cause trouble.

    Please note the e1000e driver for Linux is GPLv2. Yet you find people providing the vib for esxi without matching source to make vib file or a offer to provide the source. This is not the only example of a vib provide for esxi missing the required source link.

    Basically around esxi has a huge stack of copyright infringement going on so people can run their hardware. Problem here it is starting at the top with vmware. Drivers using vmklinux really should have required unique packaging this would have helped to mitigate some of the problem.

    Fairly much anyone who been using esxi you can almost bet they have committed copyright infringement some time or another most likely without even really knowing.

  335. oldfart says:

    “It’s interesting that an “operating system” called VMKlinux has a lot of drivers but there’s no mention that many of those drivers are Linux drivers, code from the Linux kernel… FSC/Hellwig should use this document as Exhibit One.”

    How can you tell from this document that they are from the kernel Robert Pogson?

    Wishful thinking IMHO.

  336. oldfart wrote, “For those who can understand it.
    http://www.vmware.com/files/pdf/ESXi_architecture.pdf

    It’s interesting that an “operating system” called VMKlinux has a lot of drivers but there’s no mention that many of those drivers are Linux drivers, code from the Linux kernel… FSC/Hellwig should use this document as Exhibit One.

  337. oiaohm says:

    oldfart
    http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2004746
    This prior cite I posted. Shows clearly that the console is just disabled. The reality is you don’t read cites then will accuse me of claiming something withouts even worse go head and write incorrect information even that the cite was on this very page.

    Oldfart you might call me a lier /fraud but your completely incompetent.

    ESXi architecture document that is incorrect is replacement into many Vmware documents. I could provide a list but one was already provide you.

    Again you are asking me to present cites when the cites are already presented Oldfart.

  338. oldfart says:

    “Agian oldfart cannot get basic facts correct even his corrects are incorrect.”

    And again when Lil’Hammie makes statements without citation or backup.

    for instance:

    ” VMWare in multi documents claims own-ship for work that is not theirs. ”

    Which documents are these that you refer to? One or two will do. Oh and don’t forget to be ready to demonstrate how VMWARE is doing what you say.

    Until then I will assume that a proven and admitted liar and fraud is making things up in his latest attempt at winning an argument.

  339. oiaohm says:

    Agian oldfart cannot get basic facts correct even his corrects are incorrect.
    stripped console that had a far smaller attack surface.
    Stripped means removed the reality the console was never stripped/removed. The console is in a disabled state unless you set a particular flag to re-enable it. This can be kinda required when you are have some kind of strange event result in all VM locking up at exactly the same time.

    If you look closely at oldfart displayed pdf

    http://www.vmware.com/files/pdf/ESXi_architecture.pdf
    Where does this show that the storage and network stack are in fact taken from Linux and are in the VmkLinux bit. In fact it does not. VMWare in multi documents claims own-ship for work that is not theirs. The contents of /proc of ESXi are fairly much all Linux based.

    The architecture diagrams VMWare uses are not reality. By most VMWare documents VmkLinux does not exist.

    VMWare is in fact doing a lot wrong. Unable to prove separation and claiming own-ship for work that is not theirs.

  340. oldfart says:

    make that

    stripped console that had a far smaller attack surface.

  341. oldfart says:

    “oldfart is entitled to call you whatever he wants, Fifi. Up until you admit that you were, presumably accidentally, a dismal piece of sub-human excrescence.”

    Thank you doctor, but I will stick with with my observation that Lil’Hammie is nothing more than a proven liar and fraud.

    The only thing that would add, is that he is also a joke.

  342. oldfart says:

    IN fact, if you actually used vSphere, you would know that all of the management interfaces communicate via the documented RESTful API to talk directly to the hypervisor. This includes PowerCLI, vMA and ESXCLI as well as the (soon to be obsolete standalone viclient and the vCenter Web interface.)

  343. oldfart says:

    “Userspace that internal services and a console interface do run on.”

    Not really, the interface is shut down and not counted on as anything other than being used in an emergency. If needed it can only be enable by having physical access to the vSphere physical console, and having the necessary credentials to unlock the console.

  344. oldfart says:

    Excellent, you actually did some research! Too bad you can’t quote me correctly. WHat I said was

    “The console OS is long gone from vSphere”

    As in the ESX service console that existed before VMWare came out with the ESXi variant of their hypervisor. As you may have noticedm they didn’t even want to acknowledge that and sshd instance was available as an emergency tech support mode. From your cite:

    “The ESXi Shell (formerly known as Tech Support Mode) provides
    essential maintenance commands. It can be used in exceptional
    cases that cannot be handled through standard remote management
    or CLI tools. The ESXi Shell is primarily intended for use
    in break-fix scenarios.”

    IN fact the ESXi Shell is not only shut off by default, but you must manually declare
    the total number of minutes you wish it open. When the declared time limit is reached it will automatically shut out any attempts to connect. It is not meant to be used on a regular basis and is not counted as one of the “normal” management tools.

    connect.

    It is in short the ESXi Shell not even close to as functional as the old ESX service console – deliberately so. VMWare had apparently had enough of having to patch the Service Console instance, and finally opted to replace it with a stripped console that had a far smaller attach surface.

  345. oiaohm says:

    DrLoser the problem is Derivative work yes ESXi does have its own true Userspace that internal services and a console interface do run on.

    DrLoser remember Vmware is the process of copying have removed a clause from the license. This is a problem.

    7. My sole aim in designing the driver layer of my “hypervisor” this way is to enable the “repurposing” of a Linux device driver, thus massively reducing the cost of any hardware or firmware manufacturer when “repurposing” their product.
    That arguement works for something like Xen,

    Any clue as to how I might be able to run a random Windows OS over VMWare, given that said Windows OS has absolutely no “user space” knowledge of “Linux particular things?”

    No? Thought not.

    Thought wrong because again you are guessing without research. As I pointed out to Oldfart it still has a userspace. vCLI package on Windows or Linux running ESXi can access information on Linux particular things inside ESXi. The userspace/vm space interfaces were not designed so that vmklinux could be easily replaced as a GPL licensed work requires. This is where the wrapper arguement starts coming apart. ESXi exposed interfaces are depend on Linux parts being there. There is absolutely valid reason for this to be the case if they were just after driver compatibility.

    DrLoser the problem here is formal separation. Another one of the derivate work arguments is picture in frame and picture painted in something that looks to have a frame. Picture in frame the frame and Picture are independent works. Something picture painted that it looks to have a frame but this is part of the picture.

    The problem here is VMWare looks to have done ” Something picture painted that it looks to have a frame but this is part of the picture.”

    Reason why I said VMWare should have done like FreeBSD kernel to wrapper as well or some other OS as well was this enabled developers to make sure they have not crossed the line that turns a independent work into a derivative work. GPL is very clear if a work is officially a Derivative you have agreed to hand over the source code.

    If the Linus clause had shipped with ESXi damages are in fact limited to only the source code of vmkernel being released so the userspace tools would have been 100 percent safe. Problem that VMWare developers have got careless and removed the Linus Clause someone could argue for the user-space and vm-space utilities.

    Reality most people use VMWares hyper-visor for the utilities and the support both that should not have been at risk of VMWare had got the licensing correct.

    There have been many cases of parties just creating an API over a GPL work and losing in court because they found out they had failed to create correct separation.

    VMWare still might win this but there starting position is horribly bad. Not like it would have been that hard to include another item offering the same functionality as vmkLinux using freebsd or others. If this had been done this current case would not be that possible and the API exposed to user-space and vm space would have had a neutral nature.

    I do think VMWare needs a hit over this. If the court does not rule that VMWare has to release the vmkernel source code they should rule that VMWare has to provide a drop in replacement for vmkLinux so the provable separation status exists. Basically give VMWare a choice release vmkernel or prove that your VMK API in fact works fully to provide separation. Currently there is no true evidence that the VMK API works to provide true separation there is a massive amount of counter evidence that suggests otherwise.

    2008 VMWare did in fact promise that ESXi would operate without depending on vmkLinux yet it still depends on it today. Yes that is in the Linux kernel mailing lists. So vmware has been given 7 years yet 7 years latter they still cannot demonstrate formal separation between the GPLv2 work and their kernel.

    DrLoser reality is all that is wanted is compliance. Compliance would be achieved by being able to replace vmkLinux. Then they could keep on using vmkLinux as much as they liked with all the drivers.

    The point of view that this case is just about stealing Vmwares source code is wrong. Its more than Vmware has failed to keep word so result is court now.

  346. DrLoser says:

    Forgive me for asking you this, Robert, but you weren’t really expecting a reasoned discussion on the legal issues when you posted this, were you?

    Next time, I’d recommend actually reading your cites.

  347. DrLoser says:

    Exactly what legal right does the USA have to apply its own laws to an overseas sourced work?

    Is it in any way relevant that the work in question is sourced overseas, Fifi?

    Well, douse me with turpentine and light me up like a Christmas Candle.

    I do believe not.

    Prat.

  348. DrLoser says:

    Anyway, who here is up for Exhibit Four? (SCSI drivers.)

    I don’t wish to limit you all. You’ve been a pathetic failure on Exhibits One, Two and Three.

    There’s always the possibility that you know more about SCSI drivers than I do. In fact, I know practically nothing.

    But learning new stuff is fun, isn’t it?

  349. DrLoser says:

    Some how you think you have the right to call me a fraud.

    It’s a more generous observation than calling you a dismal piece of sub-human excrescence, Fifi. Which, given your nauseating claim that (I believe) 15 privileged little Saudi nonces and four more little privileged nonces from Kuwait and Yemen were perfectly justified to do what they did, is remarkably generous.

    oldfart is entitled to call you whatever he wants, Fifi. Up until you admit that you were, presumably accidentally, a dismal piece of sub-human excrescence.

    And you can start by admitting that your berserk “apology” for the 9/11 terrorists was, in fact, without any justification whatsoever.

  350. DrLoser says:

    Blah blah blah Oh Look! Unintentional Clerical Script!

    Exactly what legal right does the USA have to apply its own laws to an overseas sourced work. Turns out it does not have that legal right if the case would breach treaties.

    Blah blah blah blah blah.

    Stop embarrassing yourself, Fifi. Pick a single one of my eight “devil’s avocate” points, and address your noddy little microwave-enhanced mind to it, please. Or, alternatively?

    Turns out it does not have that legal right if the case would breach treaties.

    “Turns out?”

    BWAHAHAHAHAHAHA!

    Friday came early this week.

  351. oldfart says:

    “You see, the thing is, you’re looking at the driver API upside-down. ”

    Because in the end a Device driver does not talk to an API. A device driver is structured to be talked to by the Operating system. The operating system makes very specific requests of a device driver in specific prenegotiated formats to perform certain tasks on its behalf. The device driver for its part makes no response until it is told to do something. IN short, it does not speak until spoken to!

  352. oiaohm says:

    oldfart the problem here is any good lawyer is a master liar.

    About to mention something really good. DrLoser and You join on topics you don’t understand.

    Remember I said when you export the work comes under treaty not country copyright law. Same applies to when you import a copyrighted work. Christoph Hellwig work and other work in the Linux source code base is under overseas copyright to the USA.

    Exactly what legal right does the USA have to apply its own laws to an overseas sourced work. Turns out it does not have that legal right if the case would breach treaties.

    USA is allows to import overseas copyrighted works because of treaties. What is the problem at the end of World war II is the treaties that least to the UN. Part of that agreement is a nasty clause that no party can have veto right or have a seat in the UN if they don’t keep the treaties they agree to. USA really has no option bar to protect Overseas copyright by treaties not their own law if they want to keep UN power. Japan never signed into the early UN treaties.

    Sorry Oldfart attacking me really gets you absolutely nothing.

    Of course as normal Oldfart shows his incompetences.
    http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2004746

    Notice he said console OS access to ESXi is gone. Reality is only disabled by default any properly trained vSphere operator would know this. Why you have to access the shell to configure particular performance options. Yes Linux driver tweaking.

    Its funny Oldfart its you who claim to be an expert yet you get stuff like this wrong all the time. Some how you think you have the right to call me a fraud. Sorry Oldfart you are a Fraud yourself you are some form of management staff over IT front line staff not front line staff yourself. If you are front line staff you are highly incompetent.

  353. DrLoser says:

    Right, I’m going to go out on a limb here. I doubt anybody will believe me, but this is purely a theoretical position. It may or may not appear in the case before the German Court. And if it does, I have no opinion on its validity either way. But here’s how I would argue Exhibit Three, were I a lawyer for VMWare or an amicus curiae:

    1. I accept that simply copying a Linux device driver is a breach of GPLv2 copyright.
    2. I am not proposing to do so.
    3. I am offering the ability to “repurpose” a Linux device driver, so that it may also be used in the VMWare hypervisor.
    4. This opportunity is available to any author of a GPLv2 Linux device driver.
    5. The resultant, modified, Linux device driver code is, and will always be, freely available on the VMWare site. And it will be advertised as such. And it will have the relevant GPLv2 license attached.
    6. Consequently, anybody will be able to exercise the “Four Freedoms” on the modified Linux device driver code. Home users. Commercial competitors. Anybody.
    7. My sole aim in designing the driver layer of my “hypervisor” this way is to enable the “repurposing” of a Linux device driver, thus massively reducing the cost of any hardware or firmware manufacturer when “repurposing” their product.
    8. The case rests on Exhibit Three, M’Lud.

    Now, are they going to argue that way? Maybe. Maybe not. But I see it as a viable argument.

    You see, the thing is, you’re looking at the driver API upside-down. This isn’t like Oracle vs Google, where Google are basically trying to pass off a nasty little software heist in order to steal somebody else’s work from below.

    This is fairly obviously a case where the “little guys,” which I admit is a weird way of describing Chinese OEMs, can gain huge benefit by simply conforming to what is a de facto standard for consumers of a device driver.

    It’s kinda hard to argue against saving millions of dollars that way. Whilst still providing a free service to the ultimate consumer. And still publishing the code, freely. And still offering a “hypervisor” platform that is completely Linux-friendly.

    The deranged nitwit Hellweg might very well have a problem with this. But up until now, I’d thought better of you guys.

  354. DrLoser says:

    VMK API passes up a lot of Linux particular things that end up exposed to user-space. A wrapper is about hiding the existence of what is inside.

    I see, oiaohm. Very informative.

    Any clue as to how I might be able to run a random Windows OS over VMWare, given that said Windows OS has absolutely no “user space” knowledge of “Linux particular things?”

    No? Thought not.

  355. DrLoser says:

    Look, I picked a paragraph of gibberish at random. I’m open to the accusation that I rolled the dice wrong. But, really:

    This is where VMWare gets the idea of derivative work just because you can run other drivers under VMK does not mean it contains enough to be classed as a complete operating system.

    1. None of us has any idea what view VMWare or their lawyers have on “derivative work.” I chose not to speculate on that subject. For some reason, oiaohm, you have prematurely concretised your interpretation. Why would you do that?
    2. vmklinux is not a complete OS.
    3. Nor do VMWare assert that it is any such thing, as far as I am aware.
    4. vmklinux appears to be a layer within the VMWare “hypervisor.”
    5. As far as I am aware, that is the only claim that VMWare would make about it.
    6. This is quite obviously the view taken by the OS Conservancy (see diagram).
    7. Only Fifi knows otherwise. Fifi is wise. Fifi is all-knowing. Fifi is evidently a direct lineal descendant of Humpty Dumpty.

    Hardly a point at issue, Fifi, but it was fun watching you making an ass out of yourself.

  356. DrLoser says:

    DrLoser I wish I could remember the french.

    Never mind, Fifi, you’ve pretty much demonstrated that you can’t cope with either German or Latin. I assume, and your sole cite of Scène à faire suggests that you can’t really cope with French colloquialisms unless Wikipedia (incorrectly) spells the meaning out for you.

    Your understanding of Clerical Script was a bit dubious, too. Not to mention your ability to identify “Ancient Greek.” Or to understand medieval hyphenated scripts. I seem to recall that your ability to misquote Beowulf is quite impressive, so I suppose that’s something. In some sort of way.

    Never mind, little chap. You’re doing awfully well in English.

    For a fourth-grader.

  357. oldfart says:

    ” The possibly that someone might write something some day to allow the ESXI vmkernel to run without vmklinux kernel will not be considered by the courts the question is if it possible why has VMWare not done it.”

    What makes you so sure they haven’t. The console OS is long gone from vSphere, and along with it all of the major visible Linux components. The Loader sequence seems to have been completely re-written as well. Of course this could all be cosmetic.

    But in the end its up to the courts to decide, not some admitted and proven liar and fraud who babbles on as if he is an accredited expert.

  358. oiaohm says:

    What people are not aware of is that derivative work idea is very old. Old master paintings use to be done in studios this was where the master painter would do most of the base drawing then commission out filling in painting.

    The key test if it your own independent work or a derivative work is if the work done by the other party is work you could have operated without.

    Derivative work cases appear that infrequently that most people forget the rules of them.

  359. oiaohm says:

    DrLoser I wish I could remember the french. The test of removing the work that is accused of infringing and seeing if the work stands on its two feet the translation is something like book of poems test. The example was a book of peoms where one poem the printer had not got the copyright to. Since you could remove the peom from the book and the book still be readable only a percentage of the value of the book had to be paid. Its a historic and a very simple test to work out if you have a derivative work or a independent work.

  360. oiaohm says:

    Here. Let me hand you a lump of code. This lump of code is a Linux device driver. More precisely, it is a Linux device driver that conforms to the API for said Linux device driver, but which has been modified in order to work with ESXi.

    Enumerate the possible copyright violations, please.
    1) Licensed under the wrong License. Yes Linux Kernel is GPLv2 but the copying include an extra clause. So the extra clause better appear.
    2) You need to answer the question is the device driver an independent work or will the result be derivative work.

    In the case of the networking and scsi subsystems that VMWare pulls in the its not like VXworks usage of network or scsi subsystems. VXWorks has their own networking and scsi subsystems. So a Linux subsystem on VXWorks is duplication to what is already has. ESXI on the other hand does not have its own networking or scsi subsystems. Like porting a Linux kernel driver to Windows has very good chances of remaining independent work even if you pulled in a large section of the Linux kernel. Why Windows can stand on its own two feet without it.

    This is where VMWare gets the idea of derivative work just because you can run other drivers under VMK does not mean it contains enough to be classed as a complete operating system.

    The problem here is due to VMWare alteration neither can operate without the other ESXI vmkernel or the vmklinux. The possibly that someone might write something some day to allow the ESXI vmkernel to run without vmklinux kernel will not be considered by the courts the question is if it possible why has VMWare not done it.

    Please also note freebsd and linux use two completely different networking stack and scsi stack designs. So an OS that can use freebsd and Linux drivers has to create their own scsi and networking interfaces for bridging between the two to hide the differences from downstream.

    3) VMK API passes up a lot of Linux particular things that end up exposed to user-space. A wrapper is about hiding the existence of what is inside.

    Lets look at Xen it had to solve the same problem

    http://wiki.xen.org/wiki/Dom0_Kernels_for_Xen

    Note Xen will operate with or without the Linux parts. A dom0 for Xen could be Linux, BSD or Solarias. So Xen maintains the independent work test. So Xen is 100 percent not a derivative work even that its making heavy usage of Linux subsystems.

    The thing to remember here if Xen Dom0 were in kernel mode Xen would still have a very good case that the Xen kernel is truly a independent work due to the fact the infringing part is removal-able without causing any issues.

    Xen over the years has worked on modifying Xen Dom0, DomU images to be able to boot with Xen missing making the independent work status clearer.

    So an open source project that has done it correctly exists. Now Xen could get higher performance if they copy what VMWare has done.

    Its going to be hard for VMware to win a not derivative work case and call their kernel a independent work. It is going to fail the most basic test that is quite simply run without the parts you have been accused of infringing on. You are able to run without the parts you are accused of infringing you have just proven independent work so restricting damages.

  361. DrLoser says:

    “idea-vs-representation.” DrLoser was used in the Orcale vs Google case …

    No it wasn’t. It was a typo by yours truly. As pointed out by you, of all people.

    Please stop pissing into the wind, and do what little you can to concentrate on the matter at hand.

    The source code of Linux device drivers, Fifi. The source code of Linux device drivers.

  362. DrLoser says:

    Linux due to being 1 a monolithic kernel design 2 going after performance lots and lots of places Linux kernel does not keep proper separation between drivers and core kernel. This make emulation hard.

    Emulation is not an issue here, oiaohm. Neither is a monolithic kernel. And whether or not the Linux kernel is so totally incompetent that it cannot maintain “proper separation between drivers and core kernel” is purely a matter between the totally incompetent kernel and the device drivers concerned.

    Here. Let me hand you a lump of code. This lump of code is a Linux device driver. More precisely, it is a Linux device driver that conforms to the API for said Linux device driver, but which has been modified in order to work with ESXi.

    Enumerate the possible copyright violations, please.

  363. oiaohm says:

    “idea-vs-representation.” DrLoser was used in the Orcale vs Google case arguing for copyright cancel that has repeatably in USA courts been over turned. This is why I don’t really want to mix non USA cases with USA ones. There are just too many horible things that happen in USA cases that don’t happen in any other copyright case. So I don’t exactly believe it was just a typo it was your memory still thinking USA.

  364. oiaohm says:

    Can you repurpose the code of a Linux device driver, without breaking copyright?

    Is it impossible? If not, what are the limitations?

    Its not impossible but its not exactly constant between Linux drivers.

    http://en.wikipedia.org/wiki/Direct_Rendering_Infrastructure

    DRI for example is GPL/MIT in the Linux kernel. These drivers are in fact independent works added to Linux. The source is used between multi different kernels. So reusing something like DRI is quite simple.

    Some Linux drivers are independent works. If the driver is an independant work converting it is not hard.

    Problem here Scsi and networking is very Linux only. Note DRI has a documented API for driver interface that they use. This documented API is OS neutral and is basically identical on freebsd solaris….

    USB not so much other than Linux having the most USB drivers out of any of the open source OS systems.

    Investment to unbind the likes of the scsi stack drivers on Linux would be massive. Not impossible but costly.

    Derivative work to be 100 percent provable requires proper separation. Linux due to being 1 a monolithic kernel design 2 going after performance lots and lots of places Linux kernel does not keep proper separation between drivers and core kernel. This make emulation hard.

    The reality trying to embed sections of the Linux kernel by VMWare is avoiding having to find about the 1.2 billion todo it properly.

  365. DrLoser says:

    Representation is not a word to use. Expression/Implementation/Work fine.

    And I would have edited it, given the chance. This site, as with so many others — and I do not regard it as an important defect — does not give me the chance.

    Naturally, it gives an imbecile the chance to jump on a completely irrelevant side issue and claim some sort of weirdo “victory.”

    As I say, I aim to please, Fifi. If that makes you happy, then by all means focus on your Happy Place.

    Now. Concerning the use of Linux device drivers in ESXi.

    Anything relevant to add to the discussion?

  366. oiaohm says:

    DrLoser the USA court is like a rabbit hole of prior case rulings. Some quite insane.

    “idea-vs-representation.” That is a bad bad word you have used there.

    Representation is not a word to use. Expression/Implementation/Work fine.

    Representation is another word for algorithm. Remember idea can be patented algorithm cannot be. This is why I strictly don’t want to mix them. It comes very simple to get lost in that stupid idea that representation of an algorithm could not be patented or copyrighted in anyway that a long dead foolish judge allowed into USA law. So now current day USA judges have to over rule that.

    On the issue of “idea versus expression,” however, I think you will find a similarity.

    Since this case is in Germany we don’t have idea vs expression in a copyright case at all. So no similarity at all.

    Its the French version of fair usage this does cover derivative works is all that is at play in the German case.

    Basically there is nothing of value Oracle vs Google case to the VMWare Lawsuit. Ok the next district court hearing over fair usage might have some relivence. Basically everything about the Oracle vs Google case to date is completley not useful at all when talking about the VMWare case. All the arguments in the Oracle vs Google case so far have no grounds in a german court.

  367. DrLoser says:

    DrLoser forks can be incompatible and compatible both happen. iceweasel fork of firefox is a compatible fork.

    I have done my best to humour you, oiaohm, but I think you have now tipped over into whimsy and hand-waving. Time to return to Exhibit Three:

    1. Is it allowable (through GPLv2) to take a body of code intended as a Linux device driver, to rework it for a different environment, and to repurpose it?
    2. If so, under what circumstances is this allowable?

    Why is it time to return to this? Because it is on-topic.

    Given the large assumption I make behind this (and I invite you all to question that assumption), this is basically what those two little boxes on the SFConservancy site were all about.

    Can you repurpose the code of a Linux device driver, without breaking copyright?

    Is it impossible? If not, what are the limitations?

  368. DrLoser says:

    Read that again.

    I did. It reads the same both times.
    1. Solve the issue at the cost of a possible load problem.
    2. Upgrade your ESXi host.

    I think either would work. Waiting for some putative judgement against VMware in a German Court, resulting in God-knows-what possible technical solution, would be …

    Pretty !!!!!!!!$# Stupid, if you ask me.

  369. oiaohm says:

    DrLoser forks can be incompatible and compatible both happen. iceweasel fork of firefox is a compatible fork.

    Compatible fork can in fact share patches without requiring major modifications most of the time.

    All most all the code base of vmkLinux is fairly much taken from the Linux tree.

    https://github.com/libos-nuse/net-next-nuse This is a fork as well.

    There are forks off the Linux kernel like pure libos-nuse that is just the networking stacks. Just because its only part does not make it not a fork.

  370. DrLoser says:

    Sorry DrLoser. Oracle vs Google is different legal framework so we don’t compare that much.

    That framework being your fantasised “German vs USA” framework, I assume?

    Because both cases have echoes of “idea-vs-representation.” In fact, on very similar bases.

    Should you wish, you can retreat to cheese-paring differences between international legal systems. I have no idea why that would make you happy, but do feel free.

    On the issue of “idea versus expression,” however, I think you will find a similarity.

    And it’s actually in your favour, you lunatic. If Oracle can assert the primacy of copyright over nine lines of stupid Java code, then I’m pretty sure that Hellwig can do the same.

  371. DrLoser says:

    ESXi is in no way a fork is kinda wrong. vmkLinux is a fork of the Linux kernel code base in an incompatible way.

    It isn’t obvious to me that vmklinux is a fork of the Linux kernel code base at all. And if you remember, I asked about this. Tumbleweed and silence, as always.

    It isn’t even obvious to me that the fork, should it be a fork, is incompatible. In fact, for FLOSS, I don’t even see what “incompatible” means.

    But, you know, evidence really isn’t all that important. Futile ignorant bigotry, that’s the way that FLOSS can be advanced.

  372. oiaohm says:

    Unlike, say, the rest of you, with your resounding silence when it comes to comparing this case against Oracle vs Google.
    Sorry DrLoser. Oracle vs Google is different legal framework so we don’t compare that much. Comparing between cases happening in two different legal frameworks is how to get yourself supper confused and lost.

  373. oiaohm says:

    DrLoser
    The quick fix for this if you can’t update your ESXi host is to change the NIC adapter from e1000 (or e1000e) to the vmxnet3 driver. The default for Windows guests is the latter, but causes problems under load.
    Read that again.

    Both are bust. e1000 or e1000e or vmxnet3 driver. vmxnet3 driver was having trouble under load. Turns out all of it is a buggy network stack patch pulled in to the Linux side causing it to be under responsive so making the fault appear in the e1000/e1000e driver yes problem disappeared as soon as up updated the vmkLinux side.

  374. DrLoser says:

    So they are not matters to take lightly at all.

    I have four, count them, four, major exhibits, modelled on Robert’s cite. You, oiaohm, are practically the only person here willing to address those exhibits.

    You think I’m taking this lightly? Bugger me. I haven’t even been able to put in the effort on that rotten subsidiary SCSI thing yet. This stuff is bloody exhausting.

    No, Fifi, I am not taking this lightly.

    Unlike, say, the rest of you, with your resounding silence when it comes to comparing this case against Oracle vs Google.

  375. DrLoser says:

    Does bring its own problems.

    http://serverfault.com/questions/585488/why-is-vmware-esxi-5-5-crashing

    Or, to quote an answer on that link:

    The quick fix for this if you can’t update your ESXi host is to change the NIC adapter from e1000 (or e1000e) to the vmxnet3 driver. The default for Windows guests is the latter, but causes problems under load.

    I wonder who wrote the vmxnet3 driver?

  376. oiaohm says:

    DrLoser you have to remember in the past company have end up bank rupt due to major copyright infringement or other IP stealing. So they are not matters to take lightly at all.

  377. oiaohm says:

    DrLoser
    And not, obviously, equal to manslaughter.

    And not, obviously, equal to aggravated bodily harm
    Both of these presume no premeditation or profit from the action.

    Sorry My alignment is perfect from a company entity point of view.

  378. oiaohm says:

    DrLoser Note I said selfish reason. xemacs, Libreoffice, eglibc. Have not exactly been out of greed. glibc and Libreoffice forks were both last restorts because it could not code upstream.

    xemacs http://www.xemacs.org/About/XEmacsVsGNUemacs.html history again last resort when feature could not be merged up stream customers required.

    MariaDB again mostly exists due to upstream miss management.

    ESXi is in no way a fork is kinda wrong. vmkLinux is a fork of the Linux kernel code base in an incompatible way.

  379. DrLoser says:

    Copyright stealing is the company entity equal to attempted murder.

    But not, obviously, equal to murder.

    And not, obviously, equal to manslaughter.

    And not, obviously, equal to aggravated bodily harm.

    And not, obviously, equal to purple monkey dishwasher.

    In fact, not really equivalent to anything, Fifi, is it? I sentence you, etc … by the way, how’s that chlamydia infection going? Try rubbing rhubarb jam over the swollen parts!

    As always, I endeavour to be of assistance. Although you’ll have to find somebody else to apply the rhubarb jam. I have an allergy to rhubarb.

  380. oiaohm says:

    http://tyrannyoftradition.com/2014/05/07/danzig-to-sue-everyone-on-earth/

    DrLoser I do agree that Robert is kinda over stating since there have been some mad people who have really attempted to sue everyone.

  381. DrLoser says:

    (I shouldn’t have to point this out. But the reason that Fifi’s assertion re. forks has no immediate relevance is that ESXi is in no way a fork.)

  382. DrLoser says:

    Forking a open source project for some selfish reason normally end up hurting your customers.

    A comment that has no immediate relevance. However, whilst we’re at it, is there any evidence for this?

    Interestingly, there are very few FOSS (I omit the L for generality and historical reasons) forks available for consideration. I can give you four, I think:

    1. emacs forked to xemacs.
    2. libc forked to eglibc.
    3. MySQL forked to MariaDB.
    4. OpenOffice forked to LibreOffice.

    It isn’t entirely clear to me that even a single one of those (quite major) forks supports your contention here, Fifi.

  383. oiaohm says:

    DrLoser remember a lot of other companies like Redhat Novell… supplied money to code and develop Linux. VMware is only a silver member who really from a money point of view has not put in that much either.

    There is more than 1 commercial entity here. 1 commercial entity stealing from another commercial entity is not on.

  384. oiaohm says:

    DrLoser really you want to skip the point. Paying money to attempt to get a person not to punish you for doing wrong is criminal. Intellectual properly is a life blood of a lot of companies. Stealing it is like killing.

    Companies are a entity. Sorry hanging judge would have got it. Copyright stealing is the company entity equal to attempted murder.

  385. DrLoser says:

    Yes, there was. SCOG sued IBM, AutoZone, Chrysler … etc

    I believe I covered that, Robert. I covered it in the very post you are quoting. I pointed out that this does not constitute a suit against “the World.”

    It’s not even as though it was a single unitary suit. Which I also covered.

    Bad day for reading today, isn’t it, Robert? Never mind. Decrepitude is merely a state of mind.

  386. DrLoser says:

    The pay the dues argument is exactly like that.

    Hanging Judge: “Fifi la Ténia, you are accused of not paying to kill Dr Loser.”
    la Ténia: “What? I never said that! Besides which, he deserves to be killed! There’s a moral equivalency here! And I never said that, and you can’t connect me with the corpse. Not that there is a corpse, of course. Well, there might be, but I don’t know about that — I was back home parboiling frogs on a Pogson recipe.”
    Hanging Judge: “So, you admit knowledge of the time that the non-existent crime occurred, Mademoiselle la Ténia?”
    Fifi la Ténia: “Haha! I’ve got you there on a technicality! It’s Berne, in Latin! No, wait, it isn’t! No, your Honour, I spend most of my waking hours parboiling frogs. On the balance of probabilities, I was parboiling a frog whilst Dr Loser wasn’t being murdered. And even if he was, which he wasn’t, I wouldn’t be able to pay for the murder because I’m behind on collecting semen from Prize Bulls.”
    Hanging Judge: “You may have misheard the accusation, la Ténia. You are accused of the Heinous Crime of Not Paying Your Linux Dues.
    Fifi la Ténia: “Thank Christ. I thought it was that prostitution thing. Well, obviously I don’t pay my dues. Nobody does. It’s Linux! Well, apart from people like VMware, who do pay their dues. But you can discount them, because they were obviously at the scene of the murderous crime. Even though it didn’t happen, and there was no scene. And nobody paid for the hit in the first place. Can I go home now? My chlamydia is playing up something rotten.”
    Hanging Judge: “Regrettably not. I sentence you, Fifi la Ténia, to be taken to the place from whence you came, and from there be taken to a place of execution. You shall be hung by the neck until the body be dead.

    “And if God can locate your shrivelled little soul, then he’s even more Omnipotent than I thought he was in the first place.”

  387. DrLoser wrote, “There was never any such case as “SCOG v World.” Not even metaphorically.”

    Yes, there was. SCOG sued IBM, AutoZone, Chrysler, wrote hundreds of letters demanding licensing fees for every installation of Linux without evening owning the copyright. They sued Novell when IBM called them on it. IBM gave a long list of defences, including SCOG not owning the software in question (A UNIX OS, not Linux.). See GROKLAW

  388. oiaohm says:

    Forking a open source project for some selfish reason normally end up hurting your customers.

  389. oiaohm says:

    DrLoser majority of the Linux kernel drivers are written in the main project. Not as independent code to add on. Updates and other items are reviewed against the mainline not against items like ESXi. Result is ESXi ends up with older versions of the code. Does bring its own problems.

    http://serverfault.com/questions/585488/why-is-vmware-esxi-5-5-crashing

    Yep Vmware sometimes failes to backport patches correctly and customers were it. Sorting out this licensing mess one way or the other might allow the ESXi Linux usage to end up mainline so reducing the disasters.

  390. oiaohm says:

    And I assume that 100% of VMWare’s customers could live with them, too.

    Not that they have to. But in the event that VMware is forced to swap to Microsoft driver APIs in, say, two years’ time …

    … It isn’t really going to be an issue, is it?

    In fact that is the issue. VMWares Hypervisor customers are after performance the Windows only high end cards are poor. Think you need to be able to move running virtual machines between hosts. Slower the network card slower that can happen. Yes time is money.

  391. DrLoser says:

    No third parties don’t rewrite drivers to support ESXi VMware gets stuck doing that. VMWare has been very shocked at how few would.

    No cites as usual.

    But, you know what? I don’t care if you have to petition the Pope to get the driver rewritten.

    The pertinent point here is that — ESXi uses device drivers written against the Linux API, and it doesn’t go through the Linux kernel to do so, and those drivers need to be rewritten.

    Who does so? What possible difference can it make?

  392. oiaohm says:

    DrLoser so how much to pay to kill you and get away with it.

    The pay the dues arguement is exactly like that. Murder is never right neither is miss using copyrighted works.

  393. DrLoser says:

    You want 40 and 100G Network cards right. Decent percentage of these are Linux only.

    I think I can live with the residuals, Fifi.

    And I assume that 100% of VMWare’s customers could live with them, too.

    Not that they have to. But in the event that VMware is forced to swap to Microsoft driver APIs in, say, two years’ time …

    … It isn’t really going to be an issue, is it?

  394. DrLoser says:

    No comments on the surprising overlap with Oracle vs Google, I see …

    How totally not a surprise.

  395. oiaohm says:

    The entire Windows-using universe seems to get along quite nicely without these SuperPrickBoutique(C) items …
    Really inside mega companies you don’t get along at all.
    http://www.hitechglobal.com/IPCores/40-100Gig_Ethernet_MAC.htm

    You want 40 and 100G Network cards right. Decent percentage of these are Linux only.

    The purpose of the “Linux device driver API” in ESXi, oiaohm, is simply to allow third-party Linux device drivers to modify their code a little and thereby to be able to interoperate with ESXi.
    No third parties don’t rewrite drivers to support ESXi VMware gets stuck doing that. VMWare has been very shocked at how few would.

    ” The reality is most of the Linux mainline drivers are not ported to user-space so are bound so strictly to the main kernel its not funny.”

    –And none of this matters. In fact, it shows up the Linux Kernel more than it shows up ESXi.–

    Ok other than the fact some of the functions ceasing to exist are security fixes. You will notice ESXi has some Linux 3.3 files. Sorry ESXi is having to meld stuff they would not need to if they were more mainline.

  396. oiaohm says:

    http://support.citrix.com/article/CTX116603

    Let’s see. I am a German Mega-Corp who has chosen to spend money on a professional, semi-closed-source, Virtualisation Solution. This solution is no longer available. What do I do?
    The reality its simpler to Migrate VMWare operational images to Xen or KVM than it is to hyper-v. Performance is only slightly worse and you remain 100 percent hardware compatible with Xen or KVM why you are still using Linux drivers. Hyper-v missing Linux drivers means migrating to it is problematic.

    http://www.embedded.com/design/operating-systems/4401769/Device-drivers-in-user-space

    Insane as it sounds Embedded Linux uses a user-space network drivers in particular cases because they can in fact be faster.

    Mind you there is https://github.com/libos-nuse/net-next-nuse This is Linux kernel network stack cut out and put in userspace. Please note the coping and applications not GPL interface with it by syscall and it interfaces with the outside world by syscall. Problem here is speed hit.

    Linux USB device drivers that are kernel only include lots of things like odd model keyboards and mice.

  397. DrLoser says:

    For what it’s worth, VMware are a “Silver Member” of the Linux Foundation.

    They might not do much, but at least, unlike you lot, they pay their dues.

  398. DrLoser says:

    Old saying no such thing as a free lunch. VMWare has attempted to take a free lunch by the looks of it.

    Actually, they appear to have invested considerably more time, effort, and money into supporting Linux than, say …

    Fifi, Robert, Dougie, oe, Lpbear, Luvr, the list goes on and on. Blood-sucking drones, the lot of you.

    (I exempt ram from this list, because he actually appears to have very employable abilities in the area.)

  399. DrLoser says:

    Just to be a super prick there are many network and scsi drivers only for Linux and Freebsd. They are never ported to Windows.

    I’d recommend you reserve your super prick observations for comments on 9/11, Fifi.

    The entire Windows-using universe seems to get along quite nicely without these SuperPrickBoutique(C) items …

  400. DrLoser says:

    Well, at least oiaohm brings us back to Exhibit Three (Linux device drivers).

    The reality is most of the Linux mainline drivers are not ported to user-space so are bound so strictly to the main kernel its not funny.

    And none of this matters. In fact, it shows up the Linux Kernel more than it shows up ESXi.

    The purpose of the “Linux device driver API” in ESXi, oiaohm, is simply to allow third-party Linux device drivers to modify their code a little and thereby to be able to interoperate with ESXi.

    By design, they do not need to interoperate with the Linux Kernel.

    It therefore matters not a smidgeon how many pointless API tergiversations the Linux Kernel goes through.

    All that ESXi needs is a relatively stable, occasionally updated, interface to the Linux device drivers necessary for operation of VMWare.

    The rest is on VMWare’s schedule, not on the schedule of a bunch of klutzes like Greg K-H.

  401. oiaohm says:

    DrLoser yes vmware could have decide to use Microsoft driver interfaces.

    http://en.wikipedia.org/wiki/NDISwrapper
    http://en.wikipedia.org/wiki/Windows_Driver_Frameworks
    Other than the fact Microsoft are even worse of a mess than Linux. Binary Windows drivers from third parties make presumes that memory layout will match Windows this was found particularly with NDISwrapper.

    Network drivers is particularly stupid by the way to think Microsoft projects a solution here. Just to be a super prick there are many network and scsi drivers only for Linux and Freebsd. They are never ported to Windows.

    You are thinking we would be upset of VMWare went Microsoft. If it means they are license conforming who cares. Yes VMWare can find themselfs in major trouble using Windows drivers as well. Patent licensing and other horible things.

    Yes FreeBSD and Linux both have the require patent grants for any usage. Not all drivers for Windows has this.

    Old saying no such thing as a free lunch. VMWare has attempted to take a free lunch by the looks of it.

  402. DrLoser says:

    DrLoser history disagrees. German Mega Corps next highest VM usage is Xen followed by KVM. Microsoft is almost not existent.

    A happy, yet entirely uncited, state of affairs that you seem to be quite willing to sacrifice on the altar of … well, it’s all very confusing. Robert’s misapprehension that ESXi is basically the Linux Kernel with knobs on? Hellwig’s Hissy Fit in 2006? Some trivial implementation of the Radix Tree, which I may remind you was originally written by somebody else entirely?

    Let’s see. I am a German Mega-Corp who has chosen to spend money on a professional, semi-closed-source, Virtualisation Solution. This solution is no longer available. What do I do?

    I could be wrong, oiaohm, but I think I would at least consider the Microsoft alternative.

    Germans are not partial to amateurish botch-jobs.

  403. DrLoser says:

    The reality is most of the Linux mainline drivers are not ported to user-space so are bound so strictly to the main kernel its not funny.

    The humorous inadequacy of the Linux kernel model aside, oiaohm, this is scarcely relevant.

    A USB stack is not, really, a huge performance issue. ESXi includes libusb. For all either of us knows, that’s all they need to support Linux USB device drivers. (It’s a very popular interface.)

    Anyway, on to network Linux device drivers, if you will. And after that, SCSI drivers. (Exhibit Four, although I will actually have to put some work into that by “examining the code.”)

  404. oiaohm says:

    If you’re seriously interested in driving German Mega-Corps into the willing arms of Microsoft, then be my guest.
    DrLoser history disagrees. German Mega Corps next highest VM usage is Xen followed by KVM. Microsoft is almost not existent.

    3. This has no bearing on its legal position on sales within the United States.
    DrLoser prior busybox case over routers disagrees. DCMA was effective at blocking import/export and movement using a German ruling.

  405. DrLoser says:

    Answer is something horible optimizing for speed. MainLine Linux kernel drivers call undocumented internal functions for performance reasons. Besides Microsoft has done this in the past its not a abnormal behavior.

    Well, that’s one answer. It’s a cretinous answer, but it’s an answer. And bringing up Microsoft in this context is a complete irrelevance. Microsoft device drivers (and/or interfaces) are not the issue here.

    Although, that’s an interesting point. Let’s fantasise about a potential future where Hellwig wins the case and the results propagate worldwide.

    Now, if I were VMware and I had a desperate need to move from “copyright tainted” device driver APIs to non-tainted APIs, what would I do?

    Chug, chug, chug … Oh, I know!

    I’d partner Microsoft and use their driver APIs instead!

    What a win for Freedom that would be!

  406. oiaohm says:

    DrLoser libusb on Linux interfaces by syscall declared functions so protected by Linus clause. It is also “GNU Lesser General Public License version 2.1” Note the Lesser it contains something more permitting than the Linus clause.

    And so is a Linux device driver header, oiaohm. The possibility that it will change on a whim merely suggests that the implementer exercises caveat emptor. Which in this case is what the implementer is doing in the first place.
    Answer is something horible optimizing for speed. MainLine Linux kernel drivers call undocumented internal functions for performance reasons. Besides Microsoft has done this in the past its not a abnormal behavior.

    Semantic patching with Coccinelle is a good read http://lwn.net/Articles/315686/ Yes Semantic patching make deleting complete functions or data structures fairly painless as long as you have all the source code.

    Using undocumented and frequently changing functions is normally problem but when your OS is fully open source its not a problem.

    In other words, in front of a German Court, that’s a rubbish argument. But let’s take a single example of a relevant API. I mean, we really only have three areas to choose from: network drivers, USB drivers, and SCSI drivers.

    scsi, usb and network drivers for Linux can be done in userspace.

    http://stgt.sourceforge.net/ Yes the stgt interfaces are stable.

    The reality is most of the Linux mainline drivers are not ported to user-space so are bound so strictly to the main kernel its not funny. So 95percent+ of Linux SCSI, USB and network drivers are bound to the Linux kernel of course there is the remaining that the developer of them has decided to have user-space version to make debugging simpler. Yes not all the drivers shipped with the Linux kernel are kernel mode only some are Kernel/Userspace.

    Did VMWare invest the money to add the if defines to allow the Linux drivers to operate by syscall so making them independent works fully. No they did not. Just because code is under a open source license does not mean it has to be in a form you can legally use in your project.

  407. DrLoser says:

    By the way, I think a very large number of German Mega-Corps are going to be hugely pissed-off in the unlikely event that this case prevents them from using VMWare.

    If you’re seriously interested in driving German Mega-Corps into the willing arms of Microsoft, then be my guest.

  408. DrLoser says:

    Yep that is what the German case would prove if VMWare loses that the VMWare product contains a fragment that is protected by the Berne Convention that VMWare does not have the right to ship.

    1. No it wouldn’t. This is a legal document covering the international rules for “take down” notices, oiaohm. Nothing more. Nothing less.
    2. If VMWare loses in Germany (and as oldfart observes, this would require an ultimate judgement in the German Federal Court), and does not make redress, then of course it won’t be able to sell into Germany, imported or otherwise.
    3. This has no bearing on its legal position on sales within the United States.

    Pointless drivel, Fifi.

  409. DrLoser says:

    The reason why the declared system call interface functions are usable is because they are documented API interface.

    And so is a Linux device driver header, oiaohm. The possibility that it will change on a whim merely suggests that the implementer exercises caveat emptor. Which in this case is what the implementer is doing in the first place … I won’t elaborate, because that would spoil everybody’s fun when trying to figure out the Big Assumption that, I mentioned below, applies to Exhibit Three.

    In other words, in front of a German Court, that’s a rubbish argument. But let’s take a single example of a relevant API. I mean, we really only have three areas to choose from: network drivers, USB drivers, and SCSI drivers.

    Are you going to argue that libusb is somehow an unstable API? It appears on the VMWare list of “Open Source Licenses,” btw.

  410. oiaohm says:

    A work qualifies for international copyright protection under the Berne Convention for the Protection of Literary and Artistic Works (Berne Convention) when it becomes attached.
    Yep that is what the German case would prove if VMWare loses that the VMWare product contains a fragment that is protected by the Berne Convention that VMWare does not have the right to ship.

    The USA DCMA law does not say proof that a work is Berne Convention protected has to be done in the USA. Something contains a fragment under Berne Convention or later copyright treaties the USA has signed that the author has legal right todo enforcement as long as its proven in a count somewhere(yes this includes Germany). The party under DCMA attack has to take out the counter case. This is basically a loop hole in USA law bigger enough to fit a super tanker.

    By treaties headers files are always copyright covered. There are a few cases in USA court history that suggest loophole for files headers. Yes these turned the Orcale vs google case into a mess. Really all cases that header files are not covered by copyright should be strut from USA copyright system. Instead replace it with proper fair usage clauses.

  411. DrLoser says:

    The core of Exhibit Three is very deliberately disingenuous, by the way. Naturally, I wouldn’t admit this in a Court Environment, but we are all amongst friends here … also, I can’t really face talking to a collection of mental midgets on this topic for the next twelve months, because unlike lawyers I am not on a retainer.

    1. Is it allowable (through GPLv2) to take a body of code intended as a Linux device driver, to rework it for a different environment, and to repurpose it?
    2. If so, under what circumstances is this allowable?

    There is, of course, a big assumption buried behind these questions. To whit … well, I’ll leave you all the fun of working out what it might be.

  412. oiaohm says:

    DrLoser the syscall allowance is specially granted by Linus. Its not GPLv2.

    It even has some merit in practical and judicial terms, although from an academic point of view I can’t really see what differentiates a documented syscall from a documented API interface.

    Name the documented Linux API interface. Yep the only documented Linux API interface is the syscall interface and supporting functions. Internal Linux Kernel API by legal standards are officially undocumented. Yes all the complaints about it changing all the time is because its not a documented API. If the function is used by syscall its documented and stable if its not used by syscall is fairly much undocumented and subject to change at a moments notice by Linux kernel developers.

    The reason why the declared system call interface functions are usable is because they are documented API interface.

    To have legal separation you have to have something documented as a interface you could make a emulator of. Note adding your own to a GPLv2 work is not really workable. GPLv2 was not design to be friendly to be mixed with other works.

    “unfair competition” comes into damages to be award over copyright infringement. Since Citrix(Xen) and Redhat(KVM) have restricted themselves to legal methods if this is proven to be copyright infringement by VMWare the result can be VMWare owning them damages as well. Yes the list of people wanting payments could get quite long.

  413. DrLoser says:

    Sorry the idea that VMWare is safe inside the USA and that a USA case will have to happen. VMWare will have to bring a counter case but this counter case will also be restricted to treaty limitations.

    Oh dear.

    You were doing so well, Fifi.

    Your cite has no relevance to this case. I don’t need to explain why. I simply need to quote from near the top to make it blindingly obvious:

    A rights owner cannot consider utilizing takedown procedures abroad without first establishing that their works qualify for international copyright protection. A work qualifies for international copyright protection under the Berne Convention for the Protection of Literary and Artistic Works (Berne Convention) when it becomes attached.

    I’ve added emphasis for the hard-of-thinking.

  414. DrLoser says:

    Also just to be horible some of the copyright treaties in the USA were agreed to on a state by state base.

    Not since the dawn of the twentieth century, I believe, oiaohm. Perhaps you also have documented proof that said interstate treaties were written in Latin?

    You’re not much cop at being horrible, although I’ll admit you try extremely hard at it.

  415. DrLoser says:

    The difference between xen and ESXi, by the way, is that ESXi doesn’t need syscalls to talk to a Linux device driver (network, USB, and presumably SCSI). Why? Because it talks to them directly.

    Which is why it uses the Linux device driver APIs, as potentially copyrightable via the headers.

    Oh, what the heck, I’ll throw you all another bone. I know how much you love and cherish the excellent Florian Mueller. Here he is on the question of whether APIs can be copyrighted.

    His (expert) opinion is not only that they can be, but that they always have been.

    It’s an opinion applied to Oracle vs Google, which might make it an awkward defense hereabouts, but I would suggest it applies quite broadly.

  416. oiaohm says:

    DrLoser remember export is the restriction. This includes electronic means.

    Also just to be horible some of the copyright treaties in the USA were agreed to on a state by state base.

    If VMware loses the case in Germany, it will still be able to sell to the USA. Why? Because until tested in a US court, it hasn’t breached US copyright law.
    Sorry the some of the Treaties effect interstate trade in the USA. Yes even with a US court case it can turn horible for VMWare.

    http://theipexporter.com/2013/03/25/enforcing-online-copyright-protections-abroad-understanding-foreign-takedown-notice-requirements/

    DrLoser once you have a ruling based on Treaties the usage of the USA DCMA system without any more court orders come into play.

    Sorry the idea that VMWare is safe inside the USA and that a USA case will have to happen. VMWare will have to bring a counter case but this counter case will also be restricted to treaty limitations.

    Yes get ruling in Germany then do enforcement by DMCA then the DMCA counter cases have to be Treaty based so USA bullshit copyright extensions don’t apply.

    Enforcement inside the USA would require DMCA operations anyhow.

  417. DrLoser says:

    I do wish to express my appreciation that oioahm stayed on-topic, for however brief a period. (And to be fair to him, his various diversions are still of general relevance.)

    Nobody else seems inclined to do so.

    1. Is it allowable (through GPLv2) to take a body of code intended as a Linux device driver, to rework it for a different environment, and to repurpose it?
    2. If so, under what circumstances is this allowable?
    To answer this you have to look no more than vxworks, Xen and a few others.

    They use the syscall boundary to separate themselves from the Linux kernel.

    This is an interesting theory. It even has some merit in practical and judicial terms, although from an academic point of view I can’t really see what differentiates a documented syscall from a documented API interface. Nonetheless, I agree that a German Court is very unlikely to take my academic point of view into consideration.

    Score +1 for oiaohm.

    I don’t see how other VM technologies are particularly relevant to the case of copyright infringement, however. The thing is either infringed or it isn’t. The actual context of infringement is of relevance only when considering “fair use,” I think. And it seems unlikely that a German Court in a copyright case is going to concern itself with possible “unfair competition” across a range of virtual machines and hypervisors.

    But, who knows? It’s still an interesting point. I’ll be generous.

    Score +2 to oiaohm.

    And nothing to anybody else, because nobody else has seen fit to address the issue in Exhibit Two.

    Which I do believe has a lot in common with Oracle vs Google, so don’t be afraid to bring that subject back up again.

  418. DrLoser says:

    SCOG v Novell was the fallout of SCOG v World. The copyright case failed in so many ways, and ultimately SCOG could not even claim copyright and stand.

    1. There was never any such case as “SCOG v World.” Not even metaphorically.
    2. The SCOG vs Novell case was about copyright ownership. Not about copyright violation.
    3. The parallel cases (which I assume you are lazily subsuming under “world”) against IBM, Red Hat, Daimler et al did not concern themselves with copyright violation either, as far as I can see.
    4. If indeed they did concern themselves with copyright violation, they would make a pretty poor argument for the Hellweg side. Because SCO in essence lost every single one of them.
    5. No mention of Oracle vs Google, I see. No surprise there, then.

  419. DrLoser says:

    Treaty enforcement is not arbitrary external interference. A Treaty enforcement does not stop VMWare selling inside the USA. So does not interfere with a countries local laws. Breaching treaties ends up with countries facing embargoes. Trade is based on treaties.

    This is just about the only point of note in that whole screed, oiaohm. If VMware loses the case in Germany, it will still be able to sell to the USA. Why? Because until tested in a US court, it hasn’t breached US copyright law.

    Now, I’m happy to accept that under Directive 2001/29/EC of the European Parliament and of the Council of 22 it should be relatively straightforward to replicate a German Court judgement on copyright law across all EU member states.

    This is not a small thing … but then again, it’s not the thing you were talking about at all, was it?

  420. oiaohm says:

    This is going to come as some surprise to the rest of the world, oiaohm, each separate part of which has its own implementation of copyright law. And each separate part of which jealously guards its own legal system from arbitrary external interference.
    The reality is that each counties implementation is trumped by treaties as soon as you export.

    Its like Modchips are legal here in Australia even to be made here in Australia. This is our law they are allowed. But you cannot export them to the USA or any other country that has signed particular treaties or you will be exported yourself to face trial. Yes treaty violations are this bad.

    Treaty enforcement is not arbitrary external interference. A Treaty enforcement does not stop VMWare selling inside the USA. So does not interfere with a countries local laws. Breaching treaties ends up with countries facing embargoes. Trade is based on treaties.

    Particular courts around the world are agreed as acceptable to hold treaty cases. German court is one of those IP.

    For criminal charges the treaty based court is “International Criminal Court” yes war crimes. Note it does not matter if the country you are in calls it legal but if the country you are in has signed the treaty or not. If they have signed the treaty they have agreed to uphold it even if it counter to the countries normal laws. Yes Treaties are part of your own countries laws just you normally don’t need to know because you don’t export.

    Locations that hold treaty cases should always cause worry. Yes they have the means to make a global ruling preventing something being traded between countries.

    Your actual references demonstrate nothing more than the fact that a company in (say) the USA who loses a lawsuit in (say) Germany cannot refuse to make reparations required by (say) the German court on the basis that they are (say) a US company.
    The Microsoft vs Samba case end up with treaty violation on the table after Microsoft just said they would just keep on paying fine. At that point Microsoft decided it was a very good time to start cooperating. Yes the legal ruling was done that the EU could 100 percent request it and the USA would enforce it.

    Note if this case was in most other counts in the EU it would not be able to do Treaty enforcement without first being heard in another court to get treaty enforcement.

    http://en.wikipedia.org/wiki/Treaty_of_Vienna_%281738%29
    DrLoser this is the last Peace Treaty written in Latin 1738. But there are other treaties country to country written in Latin leading up to the Berne Convention. Basically it takes about 200 years for Latin to disappear from being used in country to country treaties. Mostly replaced by French or English.

    The latin treaties before Paris Convention for the Protection of Industrial Property and Berne Convention for the Protection of Literary and Artistic Works are quite interesting read for the warped ways they though rights or protection should work.

    Basically before we have nice formal copyright we have printing presses with very extreme differences in point of view of what you were allowed to just copy.

    Gutenberg Bible Notice not written in German but Latin.

    The first formal copyrights are French for artworks predating the printing press. This is why the history is French-Latin-French/English for copyrights. Early French rules are fairly sane it even includes Scène à faire yes someone painting the same looking painting because they painted from the same location or the same subject(yes this is 300 years before modern day copyright). The Latin time frame is nuts like not the author of the work has the copyright but the owner of the printing press owns the work as long as he made its own typeset if he used someone else’s typeset they own the work. Yes 15 century to the 18 century copyright rules written in latin on a country by country deal base. Yet fragments of some of these latin agreements manage to get in modern day treaties.

    DrLoser the Google/Orcale case would be solved really quickly if it was Treaty based case. Because all the arguments over is this copyright or is this not copyrighted die instantly. Because the treaties don’t give such thing as not copyrighted.

    http://en.wikipedia.org/wiki/Sc%C3%A8nes_%C3%A0_faire

    Copyright based on Treaties you have to fight over this. So USA court has fairly much going around in circles slowly coming back to doing exactly what a copyright case based treaty case would have solved in the first place. Arguing over if something is copyright work or not is just wastes courts time. Its either fair and valid usage or its not.

    Google could have still won in a treaty based court. Due to less arguement options 4 hearing and it would have been all over.

    http://en.wikipedia.org/wiki/Oracle_America,_Inc._v._Google,_Inc.

    At the rate this is going it not going to be resolved until 2016 at the earliest. Oracle would have been so much better off bring the case in Germany or England… Yes one of the countries that would hear the case as treaty based.

  421. DrLoser wrote, ” I have no idea why you’re bring SCOG vs Novell into it, since that was a case built on a completely different issue (ownership of copyrights, and asset transfer thereof).”

    SCOG v Novell was the fallout of SCOG v World. The copyright case failed in so many ways, and ultimately SCOG could not even claim copyright and stand.

  422. DrLoser says:

    Basically the German court can give an almost global ruling.

    This is going to come as some surprise to the rest of the world, oiaohm, each separate part of which has its own implementation of copyright law. And each separate part of which jealously guards its own legal system from arbitrary external interference.

    Your actual references demonstrate nothing more than the fact that a company in (say) the USA who loses a lawsuit in (say) Germany cannot refuse to make reparations required by (say) the German court on the basis that they are (say) a US company.

    Which really isn’t all that interesting a point to make, although it’s certainly gratifying that there’s a level playing field in this respect.

  423. DrLoser says:

    This is why I said in copyright cases it pays to know the language of the treaties.

    And since you based your original argument on the absurd theory that “the treaties” were written in Latin (!!!), it’s safe to say that everything else you’ve said since can reasonably be ignored.

    Bit of a lucky break for you, really, oiaohm, because I don’t imagine that your grasp of Latin is any better than your grasp of German.

    (Mine is fairly decent, btw.)

  424. DrLoser says:

    No what were we talking about in 2012. That right Oracle vs Google and SCOG vs Novell both USA cases. Everything you have been saying is correct if this was a USA court case.

    It’s precisely what you were claiming in 2012. I have no idea why you’re bring SCOG vs Novell into it, since that was a case built on a completely different issue (ownership of copyrights, and asset transfer thereof). But in the case of Oracle vs Google, you were all enthusiastically insisting that Google had the right of expression of an Oracle idea.

    As of May 2014, that argument of yours isn’t faring too well in court, by the way. Which might explain why there has been no mention of it on this site recently … I assume Robert’s earlier enthusiasm to take the case to the Supreme Court (at no cost to himself, naturally) has basically evaporated.

    A possibly pertinent part of the 2014 judgement:

    “We are mindful that the application of copyright law in the computer context is often a difficult task,” the opinion reads, putting it mildly. “On this record, however, we find that the district court failed to distinguish between the threshold question of what is copyrightable – which presents a low bar – and the scope of conduct that constitutes infringing activity.”

  425. oldfart says:

    “As such I predict this case will lead to some kind of action in the plantiff’s favor!”

    Perhaps, We’ll See…

  426. LinuxGentlesir says:

    oiaohm,

    Understood! I mention these cases to show that GPL enforcement is not a novel thing. As such I predict this case will lead to some kind of action in the plantiff’s favor!

  427. oiaohm says:

    LinuxGentlesir those two were before the unified IP court in the USA.

    http://en.wikipedia.org/wiki/Board_of_Patent_Appeals_and_Interferences

    Basically since 2012 its been impractical to run a copyright or patent case in the USA. Thank goodness copyright can be conducted in a count obeying internationally agreed treaty conditions(as Germany does) then forced back on countries who courts no longer work.

    The USA count for copyright has turned into a pack of academic nit pickers with a backlog that you might die before the case is resolved.

    The academic nit pickers have forgotten that copyright is define by treaty to allow commercial trade so a USA only copyright condition only applies if the product never leaves the USA. As soon as the product is exported it has to conform to the treaties.

    http://en.wikipedia.org/wiki/Patent_court

    Surprising to most people is there are only 6 global courts operating and in future possibly 7 to address copyright, trademark, trade secrets and patent cases. The thing is a ruling any one of them as long as it is based on treaty applies to them all.

    This is why I said in copyright cases it pays to know the language of the treaties. Treaties supersede local laws when it comes to products being traded across international boarders. This was confirmed as true in the Microsoft vs Samba case.

  428. LinuxGentlesir says:

    Linus Torvalds: “Imagine ten people putting in 1 hour each every day on the project. They put in one hour of work, but because they share the end results they get nine hours of “other peoples work” for free. It sounds unfair: get nine hours of work for doing one hour. But it obviously is not.”

    This is a great quote Mr. Pogson. The magicial thing about software is it can be copied and shared with no limits!! And if I give you a copy, it does not diminish my own! This property makes it especially conductive to sharing in a way physicial goods can not be shared. A key reason why FOSS is so successful!!

    Properitary software in a world of abundance, has left us in want!! In a where we can share with no limits properitary software limits!! Properitary software development fundamentally puts greed over human progress and lovekindness!

  429. oldfart says:

    ” If its not solid and you just want to be a jackass you would conduct it somewhere like the USA.”

    Whatever… We’ll See.

  430. oiaohm says:

    oldfart the Mircosoft vs Samba case was also copyright. So this is not exactly we will see we have precedence. And the prior case means the EU court can go directly to debt collectors in the USA to get fines does not even need to have a USA court hearing to collect the fines. This kind of case if VMWare loses enforcement to the USA company is more than possible. Yes enforcement to VMWare in India and other places is also possible.

    Basically VMWare cannot afford to lose to the point the German court wishes to fine them to death.

    oldfart also do remember double jeopardy laws do apply to this. So if this case is lost in Germany it cannot be brought in any of the covered countries. Only if it won can it be decided to be brought in the covered countries for extra fines under those countries laws. So we don’t have to worry about the case being revived country after country.

    If you have a truly solid copyright infringement case you will be conducting it in Germany to reduce down the number of times you have count cases. If its not solid and you just want to be a jackass you would conduct it somewhere like the USA.

  431. oldfart says:

    “Yes this is a hang over from World War II.”

    We’ll see.

  432. oiaohm says:

    Please note to use the World War II treaty to send the ruling almost completely global you must conduct the case in Germany.

  433. oiaohm says:

    oldfart
    Of course as our Australian know it all has already admitted, US copyright laws will prevent this suit from going anywhere fast in the USA and probably other companies to boot.

    Other than the fact you have completely missed the Treaty between the EU and the USA allowing EU count to apply fines to USA companies. This was proven in the Microsoft vs Samba case. Yes Microsoft attempted to retreat to the USA to claim the case had to be brought under USA law only to find out the EU can in fact enforce its fines in the USA so the case never need to be brought to the USA. Yes this is a hang over from World War II.

    Basically the German court can give an almost global ruling.

    Due to the treaties the German court ruling covers USA, China, India and all of the EU including Russia. Only major power not covered is Japan because they never signed in.

  434. oldfart says:

    “And you brainless dingbats choose to believe that they should be prevented from doing so.

    ARE YOU ALL MAD?”

    Nope. They are mooching FOSStard losers exercising their freedom of choice to impose their choices on others. I’m sure that those of the who understand what the VMware hypervisor represents are salivating at the possibility that VMWare will be the next contributor of commercial code to the cause.

    Of course we have to get there first – The german court has to rule against VMWare, and VMWare has to lose all appeals up to the federal level. And then all other countries but Germany have to then accept this ruling and act accordingly. Of course as our Australian know it all has already admitted, US copyright laws will prevent this suit from going anywhere fast in the USA and probably other companies to boot.

    So this is all moot.

  435. oiaohm says:

    1. Is it allowable (through GPLv2) to take a body of code intended as a Linux device driver, to rework it for a different environment, and to repurpose it?
    2. If so, under what circumstances is this allowable?

    To answer this you have to look no more than vxworks, Xen and a few others.

    They use the syscall boundary to separate themselves from the Linux kernel.

    http://en.wikipedia.org/wiki/Ship_of_Theseus
    One the question we have here is this problem Ship of Theseus. Has VMWare added enough of their own work to rename the work. Remember everyone else to run Linux to run drivers has called there alteration of Linux a modified Linux. Something VMWare has tried to say its just a compatibility layer.

    Freebsd has a Linux compadiblity layer. How much code of Linux does it contain Zero.

    The real reason that VMware appeals to corporate customers (or in the case of oldfart, academic customers) is that they can spin up multiple copies of any Linux server they want.

    And you brainless dingbats choose to believe that they should be prevented from doing so.
    DrLoser remember Hyper-v, Xen and KVM exists todo the same roles as VMWare. If VMWare is cheating the commercial vendors behind these products have unfair competition that need to be removed from the market.

    Removing VMWare from market does not really prevent Oldfart form doing the same things he use todo using different software. But due to oldfart being so suck on using particular software he would most likely fail big time.

    In fact, every FLOSSie here adverted the precisely opposite opinion on “idea versus expression,” just because they felt like it.
    No what were we talking about in 2012. That right Oracle vs Google and SCOG vs Novell both USA cases. Everything you have been saying is correct if this was a USA court case.

    Idea and Expression is a confused mess in USA law. The French law that the international treats come from is very clear cut.

    Its also why I goofed up badly in a few places in that case. Because Australian copyright and USA copyright really don’t match. German and Australian kinda do. Australia is not a EU. EU countries use the french. Australia uses a rewrite to local laws that is very close.

    http://en.wikipedia.org/wiki/Sc%C3%A8nes_%C3%A0_faire
    Scène à faire (French for “scene to be made” or “scene that must be done”; plural: scènes à faire)
    This is the copyright enforcement test under german law and Australian law.

  436. LinuxGentlesir says:

    The real reason that VMware appeals to corporate customers (or in the case of oldfart, academic customers) is that they can spin up multiple copies of any Linux server they want.

    … that’s no excuse!! Perservation of Freedom is more important then promotion. If they do not value Freedom, they aren’t promoting anything we stand for! Perserving the Free Software/Copyleft nature of GNU/Linux takes precendence over all other concerns, because Freedom is the soul of GNU/Linux!! VMWare must abide by the terms of the GPL.

    No excuses!!!!

  437. DrLoser says:

    Anyway. Back to the Big Question. Consider Exhibit Three, to whit the repurposing of a limited sub-set of Linux device drivers.

    Any particular reason to see this as Evil?

    No, I thought not. Oh well, on to Exhibit Four, then.

  438. DrLoser says:

    Speaking of Virgil:

    Ambo florentes aetatibus, Arcades ambo,
    Et cantare pares, et repondere parati.

    Nice little Iambic Pentameter there.

  439. DrLoser says:

    Want to know the truly weird thing about all this?

    I’m just guessing. There is value in VMWare for software developers who want to spin up VM variants of M$ (say, different cultures) or whatever. But, you know what?

    Again, it’s a pure guess. (Backed up by oldfart’s experience.)

    The real reason that VMware appeals to corporate customers (or in the case of oldfart, academic customers) is that they can spin up multiple copies of any Linux server they want.

    And you brainless dingbats choose to believe that they should be prevented from doing so.

    ARE YOU ALL MAD?

  440. DrLoser says:

    We haven’t quite plumbed your infinite depths concerning ignorance in Latin, though, Fifi. Obviously you’ve tried your hand on the subject.

    A relevant epithet, perhaps? Juvenal? Pliny, either the Younger or the Elder? Catullus?

    Gawd knows, even Goethe (German, just a hint) or even Virgil (Latin, just a hint) might give your gibberings a little authorative weight.

    Not that I’m holding my breath, Fifi. You are very possibly the most ignorant person I have ever come across.

  441. DrLoser says:

    No further comment, M’Lud. The floor is open to the Plaintiff’s expert staff.

    I am not one to limit other people’s freedoms, unlike the blood-sucking drone Pogson (“Thou shalt not compose music unless such music is composed via broken crap unusable FLOSS software! That there is the Mark of the Slave!”).

    Go ahead, Fifi. Make of your nitwit disturbing ignorant psychopathic terrorist-apologist self something of worth, for once.

    Be a prominent part of the Expert Staff!

    (I wouldn’t try citations in German, btw. You’re a pathetic failure at that.)

  442. DrLoser says:

    Not that I object to the possibility of discussing “idea versus expression.”

    On the contrary, I’m all for it. As was everybody else here in 2012, as I recall. In fact, every FLOSSie here adverted the precisely opposite opinion on “idea versus expression,” just because they felt like it. You want me to bring that topic back into focus, oiaohm or Robert?

    You really don’t./i> Trust me on this one.

    You’re all flailing around in an intellectually incoherent swamp when it comes to Exhibits One, Two, and Three. Let’s pretend 2012 and subsequent appeals (oops, giving the game away there) never happened.

    You’ve all got enough on your plate trying to apologise for your present smorgasbord of incoherence, haven’t you?

  443. DrLoser says:

    This is a train-wreck happening in slow motion.

    True, Robert. Sometimes the FLOSS process isn’t quite as fast as you would like. And it’s very odd that FLOSS champions such as, say, IBM, Red Hat, Google … just to name a few … have lingered these long ten years, isn’t it? I’m sure they’d have gotten around to it eventually.

    It’s taken years after Hellwig first complained…

    As oldfart pointed out, Robert, all that the nitwit did in 2006 was to go into a hissy fit on an LKML thread. Which thread was otherwise populated by rather less borderline-weirdo types. Your point here?

    …and they still haven’t released the source code.

    What, apart from that ISO that you linked to, Robert?

    Let’s cut to the chase here.

    What source code are you missing from VMware?

    What’s with that? Are these movers and shakers of virtualization that stupid?

    Depends on your definition of “stupid,” Robert. Are they Manitobans?

  444. DrLoser says:

    Just out of interest, my little psychopathic friend, what constitutes a “major memory error?”

    The one small moment in your life that you watched two planes crash into the Twin Towers on 9/11 and kill three thousand people for no good reason at all?

    I still have that memory. I classify it as “major.”

    For you, it was just a minor irritant in being forced to move from one tin-pot Outback job to another, wasn’t it?

    You careless slob.

  445. DrLoser says:

    So my comment about Latin was a minor memory error. I was thinking the second copyright law was Berne Convention.

    You seem to be more and more subject to these “minor memory errors,” Fifi: because more and more you are being exposed as the ignorant buffoon that you are. Have you tried re-tuning the microwave rheostat in your brain? (I’m not the one that claims magical microwave powers for your brain, as you know.)

    Your German is sadly non-existent, isn’t it?

    DrLoser http://en.wikipedia.org/wiki/Berne_Convention
    1887 Urheberrechtsgesetz is legally deprecated.

    And written in Latin. And one of the parties was the German State. And the original convention was signed in 1710, in Latin, by a German State that didn’t even exist.

    Stop fooling yourself, Fifi. In this one particular instance, you are completely and utterly ignorant.

  446. DrLoser says:

    How does an ESXi device driver access the Linux device driver?

    The simple answer (I apologise for the elision of the sub-clause) here, Robert is — that’s not what ESXi does.

    It doesn’t “access the Linux driver.” As a matter of fact, it doesn’t “access” anything at all inside the Linux kernel.

    However, there is an argument to be made that ESXi infringes copyright on any given Linux device driver, because, whaddya know, they actually use the code base of the driver itself.

    Criminal!

    Or not. It isn’t possible to simply plug in a Linux device driver to ESXi without modification, Robert. This is admitted by VMware themselves. Instead, what you need to do (as the provider of a device driver, which is an important part of this discussion) is to modify said driver in order to make it work with ESXi.

    It isn’t a drop-in replacement for a driver within the Linux kernel. It doesn’t even talk to the Linux kernel. There is no interface to Linux. There is only “modified code.”

    That’s the basic position on Exhibit Three, then.

    1. Is it allowable (through GPLv2) to take a body of code intended as a Linux device driver, to rework it for a different environment, and to repurpose it?
    2. If so, under what circumstances is this allowable?

    No further comment, M’Lud. The floor is open to the Plaintiff’s expert staff.

  447. DrLoser says:

    And now we move on to Exhibit Three.

    How does an ESXi device driver access the Linux device driver if not through the Linux headers?

    A pardonable confusion, Robert, and I’ve been guilty of similar confusion myself. We’re not talking about the ESXi driver accessing anything at all. We’re talking about vmklinux accessing the Linux device driver. And yes, it specifically does so through the Linux headers, as far as I can see. So let’s, for argument’s sake, take it as a given that it does so. (That was an Exhibit Two point of debate.)

    I’m going to leave SCSI drivers to one side as Exhibit Four here, based on the original cite. Let’s consider all the other Linux drivers out there, which seem to be coloured in orange and are limited (other than SCSI) to USB and Network drivers.

    Now, there is definitely a Network stack inside the Linux kernel. I know this because I have worked on the BSD equivalent. And there is almost certainly a USB stack inside the Linux kernel. (I’m guessing.) But, underneath both stacks, there are “Linux device drivers.”

    Now, Linux device drivers come in all shapes and sizes. Some of them are binary blobs. Some of them are provided by NVIDIA or, well, practically any other third party you care to name. And some of them are part of the repository tree of the Linux kernel, although very few of these are worth spit, because they’re generally a haphazard mixture of vanilla “it sorta works” and random guesswork.

    So, here’s my question. Is there an obvious way that VMware can be in breach of the GPLv2 by interfacing ESXi to “Linux device drivers?”

    Well, no. There may be a way. But it isn’t obvious.

  448. DrLoser says:

    Give your head a shake. Their code runs on Linux 2.4-whatever.

    Does it? Does it really? Does ESXi “run on Linux 2.4-whatever?” Do you have any evidence of this? Are you prepared to submit a link to that effect? Do you have a single clue what you are talking about, Robert?

    RH7.3 is 2.4-something.

    And unlike you, Robert, I have built software on top of RHEL. Apparently VMWare have done the same. It’s quite a common practise in the real world, you know. There is no copyright infringement involved in building software on any given platform whatsoever.

    Proof of a breach is lack of source code modified by VMware.

    I’m sure there’s a kernel of a good idea in there, Robert. But as it stands, it’s self-contradictory nonsense, isn’t it?

    Where is it? VMware claims their “kernel” is a wrapper around Linux.

    No they don’t. Where do they claim this? What are you smoking, Robert? What sort of lunatic corporation would make this claim in the face of a suit brought to a German Court? Are you imagining stuff?

    How is that possible without copying or modifying Linux code?

    Well, technically, it was a silly and completely irrelevant question. But since you ask, adding a wrapper around any code whatsoever is trivially simple without bothering to copy or modify the code. It’s sort of how Linux applications work, for example.

    How does an ESXi device driver access the Linux device driver if not through the Linux headers?

    Right. Here we get down to the real discussion.

  449. DrLoser wrote, “A requirement for RHEL in order to “build” any software package whatsoever is hardly a breach of the GPLv2 copyright, is it?”

    Give your head a shake. Their code runs on Linux 2.4-whatever. RH7.3 is 2.4-something. Proof of a breach is lack of source code modified by VMware. Where is it? VMware claims their “kernel” is a wrapper around Linux. How is that possible without copying or modifying Linux code? How does an ESXi device driver access the Linux device driver if not through the Linux headers? How does an ESXi device driver have a different interface to Linux without modifying Linux code? RH7.3 has different driver interfaces than ESXi.

    “If we take a look back at the early days of ESX, VMware made a decision to use Linux derived drivers to provide the widest variety of support for storage, network and other hardware devices for ESX. Since ESX and specifically the VMkernel is NOT Linux, to accomplish this we built a translation (shim) layer module called vmklinux which sits in between the VMkernel and drivers. This vmklinux module is what enables ESX to function with the linux derived drivers and provides an API which can speak directly to the VMkernel.” See ESXi 5.5 introduces a new Native Device Driver Architecture

    That’s not how Linux drivers work but that’s how ESXi uses a Linux loadable kernel module. How’d they do that without modifying Linux? Where’s the source code of vmklinux?

  450. DrLoser wrote,

    • “there is as yet no concrete evidence that VMWare has been legally naughty in respect of Linux and the GPLv2.” Sure, there is. Hellwig called them on it back in 2006 and they had nothing to say… Some juries would take that as prima facie evidence of guilt like a shooter running away.
    • “Add the small bits of support that FreeBSD is missing.” That’s not true at all. Linux has had ten times as many developers as FreeBSD contributing drivers for ages. FreeBSD does not have sufficient bodies, nor sufficient access to hardware to create them. VMware could have helped out FreeBSD; in fact they have a release for FreeBSD. They needed a release for Linux because that’s what most customers use. Folks claim Linux is starved of talent. FreeBSD is skin and bones by comparison.
    • “I think the other two options would actually be far easier, let alone more fun.” With the resources they have available, VMware could do whatever they want as long as it’s legal. That’s why this is an entirely preventable train-wreck. The problem was pointed out to them in 2006 and they increased speed down this track. This is what makes this a tragedy, that human frailty causes people to choose the wrong path and stick to it until the bitter end, when the stage is littered with corpses. Greed seems to overrule all common sense sometimes.
  451. DrLoser wrote, “What, precisely, is your big beef with FreeBSD?”

    *BSD has a long history but at the time Linus and friends did the Linux thing, *BSD was locked up tightly until 1991, about the same time as the first release of Linux. BSD was probably ahead of Linux then. Now, FreeBSD is FLOSS in that it complies with the definition of Free Software but it does permit someone modifying the code, distributing the modification but not the modified source code. That’s not proper sharing which I think is important. Most developers seem to feel GPLv2 is better because it requires sharing modified code and they contribute vigorously to Linux for that reason. What’s the point of contributing to FreeBSD if your code gets distributed by someone else in binary-only form after modification and you can’t get the benefit of those modifications to use in your own coding? Linus has stated that he thinks choosing GPLv2 was a great decision. I think he’s right. GPLv2 is a better licence for most components of an operating system. That way the whole world can contribute and share the result. With FreeBSD, the licence permits interfering with that by hiding code. That makes no sense to me, and apparently many others. FreeBSD has a few hundred committers. Linux has many thousands. I’m not claiming quantity is quality, but it certainly covers more hardware which is very important. There’s certainly no lack of resources for Linux. *BSD does have a great rep for servers but the more diverse hardware of clients is a problem. Enough written.

  452. DrLoser wrote, “Linux device driver headers”.

    This case is not about such things, if they exist. In Linux, device drivers are either embedded or loadable kernel modules. The source code of the drivers is linked to Linux through Linux kernel headers. This case is about modifications of Linux at other levels which VMware distributes but does not reveal the source code as required by GPLv2. Hint: VMware does not use a release from kernel.org. They distribute a modified kernel but not its source code. Xen does. Why doesn’t VMware? They won’t answer so they will be forced to answer in court.

    This is a train-wreck happening in slow motion. It’s taken years after Hellwig first complained and they still haven’t released the source code. What’s with that? Are these movers and shakers of virtualization that stupid? Hellwig and SFC can get a court order to reveal the code, build it, show that it is or is not what VMware distributes and VMware is toast. If the build matches, Hellwig has proof of copyright violation without compliance with GPLv2. If the build does not match, Hellwig has proof of contempt of court. All they have to do is publish the damned code to apply the brakes but they won’t. VMware is gambling with VMware’s existence. If they go all the way, VMware could be shut down. What are they thinking?

    Car analogy: everyone from the father to the driving instructor has told the teenager that going around the curve too fast is deadly, but the kid does it anyway… I had a kid do that on my corner lot once, taking out 20 ft. of fencing. He told his Dad the tires “blew”. Yes, they were ripped off going sideways. It’s a wonder the thing didn’t roll and kill people. The kid managed only 45 degrees of a 90 degree turn. VMware is that kid.

  453. LinuxGentlesir wrote, “Companies will have to learn that violating the GPL is not something that will go unpunished.”

    It’s not about punishment. It’s about doing the right thing. This lawsuit was the last thing they wanted. I don’t think VMware can win this thing. They have to comply ASAP or it’s all downhill for them. All those big corporations using VMware will go elsewhere once the court rules or the evidence of violation is public. It’s in VMware’s best interest to settle and that requires showing the code.

  454. oiaohm says:

    By the way EU cross country trade laws language usage go in
    French-Latin-[French/English]
    That is in the order of writing from oldest to newest.

    So my comment about Latin was a minor memory error. I was thinking the second copyright law was Berne Convention. The mistake I was thinking of the ruling not to specal tax books was copyright(this is in Latin) predates the Berne Convention by 4 years. Yes the Berne Convention is the start of the usage of French/English in EU cross country laws.

  455. oiaohm says:

    DrLoser yes there are newer ruling at

    No, Fifi, you are wrong. For some very bizarre reason that need not trouble us at the moment, the German State tends to write German Law in German.
    Expect for copyright law and other laws related to trade as those are written in English and French. Those laws are ratified by German State and translated into German.

    German has rulings that effect interperationation of the Law that are written in German. Rulings are law are to different things DrLoser.

    Laws effecting trade across EU have by tradition been written in French. To be correct a difference between English and French version the French version wins.

    Yes the EU is a very strange place at times. Fighting copyright in any country in the EU you need a person who can read French to look for law loopholes.

  456. DrLoser says:

    I take it, by the way, that nobody has a coherent objection to my claim that a possible defence to using Linux device driver headers is that:

    To all intents and purposes, they are equivalent to the definition of an Open Standard for device drivers?

    Register any and all objections now, please. And I will now continue to Exhibit Three of four in the case of nitwit vs VMWare.

  457. DrLoser says:

    French and English is what the common copyright law is written in of EU. So if you are reading a document in German its either obsolete or a translation.

    Well, there goes the proud boast of Prussian Ancestry, Fifi. Presumably you signed up for that, too, on the back of a bubble-gum wrapper?

    Your entire miserable existence seems to be a random walk in high heels through a series of discarded bubble-gum wrappers. That, and a confused yet obsessive collection of irrelevant Internet links, is all you have, isn’t it?

    No, Fifi, you are wrong. For some very bizarre reason that need not trouble us at the moment, the German State tends to write German Law in German.

    Me, I think it’s a mistake. But possibly they haven’t completed the German to Klingon dictionary quite yet. There is still hope for a rational future.

    Dimwit.

  458. DrLoser says:

    DrLoser http://en.wikipedia.org/wiki/Berne_Convention
    1887 Urheberrechtsgesetz is legally deprecated.

    I think not, you nauseous psychopathic little prat.

  459. DrLoser says:

    Of course I was joking around with the latin comment. Again you completely miss my mistake because you don’t know the topic.

    Of course Fifi caught out trying funny latin comment. Here’s unbelievable funny thing, Fifi come up with pointless drivel. Again. Better. Better. Better. Dr Loser completely miss Fifi-stake.

    That “mistake” being, Fifi?

    The topic I don’t know being, Fifi?

    Any single topic in the whole history of topics that you “know,” Fifi?

    BWAHAHAHAHAHA

  460. oiaohm says:

    DrLoser http://en.wikipedia.org/wiki/Berne_Convention
    1887 Urheberrechtsgesetz is legally deprecated.

    Of course I was joking around with the latin comment. Again you completely miss my mistake because you don’t know the topic.

    French and English is what the common copyright law is written in of EU. So if you are reading a document in German its either obsolete or a translation. Breaking the common copyright law breaks means to trade in the EU.

    So as soon as you started claiming you had read a German document you were wrong DrLoser. Stop trying to google you way out of this one.

    <b.There is, indeed, a single element of that header that qualifies as a non-trivial inline
    DrLoser I do understand the idea of inline code. Issue here is copyright law does not give exceptions in germany/EU for header contents. Copyright law treats a inline like copy pasting into your own work.

    Again in Germany/EU the court ruling to declare stuff trivial does not exist in Germany/EU. Fair usage is also extremely limited.

    DrLoser again you have fall back on a USA idea.

  461. LinuxGentlesir says:

    The SFC has good record of productive outcomes in these legal matters. I look forward to a favorable legal outcome against VMWare. Companies will have to learn that violating the GPL is not something that will go unpunished.

  462. DrLoser says:

    Interesting that you should mention the possibility that VMWare had a choice of software to “rip off,” however, Robert.

    Here is the FreeBSD equivalent to the Radix Tree implementation.

    Well, that seems to have cleared up any lingering issues with Exhibit One.

    Can we get back to Exhibit Two, which is to say Linux device driver headers, please?

    Submissions welcome. Otherwise I shall proceed, as promised, to Exhibit Three.

  463. DrLoser says:

    Incidentally, you bring up an interesting question, Robert.

    What, precisely, is your big beef with FreeBSD?

    I assume that it’s because the FreeBSD license doesn’t allow you to contribute back to the Community. But that’s just an assumption. I welcome clarification on this issue.

  464. DrLoser says:

    They have to show the source code to distribute Linux or they have to use some other kernel. They could do that with *BSD for instance but they won’t because *BSD doesn’t have the hardware-support of Linux.

    What is this, Robert? Plan Nine From Outer Space?

    First of all, there is as yet no concrete evidence that VMWare has been legally naughty in respect of Linux and the GPLv2. But to short-circuit the discussion, let’s assume that they are.

    And let’s assume that they had a need to rip off an entire OS kernel, for some as yet unspecified reason.

    And let’s assume that the kernel of alternate choice would have been FreeBSD. Not a bad choice IMHO.

    And let’s assume that FreeBSD doesn’t support the necessary range of hardware for VMWare. That actually sounds very unlikely, but it’s a valid assumption.

    Now, Robert. Put yourself in the position of the ESXi development team. What do you do?

    1. Add the small bits of support that FreeBSD is missing.
    2. Add a Hardware Abstraction Layer so that the entire problem disappears.
    3. Rip off the Linux kernel.

    I can’t see a dedicated team of professionals working on the hardware/OS boundary choosing number three, Robert. I think the other two options would actually be far easier, let alone more fun.

    Let’s be honest here. Your whacko theory is completely without any sort of substantiation outside the echoes in your own head, isn’t it?

  465. DrLoser says:

    I’m still not denying the possibility that VMware are in breach of copyright. I even asserted (earlier) that they might be found in breach of copyright on both “exhibits” I have been allowed to bring to your attention earlier, and quite possibly the next two.

    I have no skin in this game whatsoever. It is a purely intellectual exercise. And even so, it might be trumped (on either side) by some legalistic nonsense. We all know that.

    Now, put your thinking hat on for once, Robert, and leave the evangelism for a separate thread.

  466. DrLoser says:

    No, I don’t. The copyright violation is in copying Linux without following the rules of GPL v2. VMware’s download is not their current stuff. RH7.3 is not what they are currently using yet that source code clearly states that one needs RH7.3 to build it.

    Yes you do.

    You have provided no evidence that ESXi “copies Linux without following the rules of GPLv2.” I’ve spent a week or more examining this possibility in a painstaking way, with very little help from you, although oiaohm is contributing more than usual sense.

    A requirement for RHEL in order to “build” any software package whatsoever is hardly a breach of the GPLv2 copyright, is it?

  467. DrLoser says:

    Just to sux nuts the original you need to read is written in latin as part of a treaty between England, Germany and other countries of that area over copyright.

    Let me get this one straight, Fifi. You are claiming, evidently with a straight face, that the original text of the Urheberrechtsgesetz, a German legal apparatus laid down originally in, I think, 1965, was written in Latin as “part of a treaty between” England, Germany and “other countries?”

    REALLY???

    Your source presumably cites 1710. Or, pick a year between then and, well, 1965. Now, here is an interesting yet entirely irrelevant highlight of your particularly fatuous observation, Fifi: Germany didn’t even exist until 1870.

    You have seriously outdone yourself on completely cretinous babbling here, Fifi. I am so proud of you, Little Man! You never let me down! Just in time for the traditional Friday Salutation!

    BWAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!

  468. DrLoser wrote, “You need to read enough to be able to spot at least one possible copyright violation.”

    No, I don’t. The copyright violation is in copying Linux without following the rules of GPL v2. VMware’s download is not their current stuff. RH7.3 is not what they are currently using yet that source code clearly states that one needs RH7.3 to build it. If that’s not true and they don’t maintain their software well enough to fix such discrepancies, there’s no reason at all to use their software. Their download does not include Linux, just some drivers/modules. What Hellwig is claiming is that they’ve modified Linux and distributed the modification without revealing the source code. That download is consistent with that claim. VMware’s “counter” is not supported by that download. Show us the real code. VMware’s whine that they are OK because folks can link to the version of Linux that VMware distributes is irrelevant to the issue. They have to show the source code to distribute Linux or they have to use some other kernel. They could do that with *BSD for instance but they won’t because *BSD doesn’t have the hardware-support of Linux. VMware has been caught out.

  469. DrLoser says:

    I’ve read enough of it.

    I don’t think that’s a good enough explanation, Robert. Which bits of this rather large download of freely available (though not necessarily FLOSS) code did you read?

    The comments?
    The layout of the directory structure?
    The headers?
    The code?

    You’ve read enough to convince yourself … but that’s pretty worthless. You need to read enough to be able to spot at least one possible copyright violation.

    I very much doubt that you have done so.

  470. DrLoser says:

    Each time you redistribute the Program (or any work based on the
    Program), the recipient automatically receives a license from the
    original licensor to copy, distribute or modify the Program subject to
    these terms and conditions.

    Section 6 of GPLv2 makes removing the clause a breach. Why removing the Linus declare clause restricts the receiving users rights.

    Which particular syllable of the word “automatically” do you not understand, Fifi?

  471. DrLoser says:

    Mind you the radix-tree.h as a header file appears. Wonder if their has been a lot of in-lining.

    Having examined the header in your cite, oiaohm, I can confidently assert that there has been no inlining whatsoever, with the trivial exceptions of redirection to:

    1. static inline int radix_tree_deref_retry(void *arg);
    2. static inline int radix_tree_exceptional_entry(void *arg);
    3. static inline int radix_tree_exception(void *arg);
    4. static inline void radix_tree_replace_slot(void **pslot, void *item);
    5. static inline void radix_tree_preload_end(void);
    6. static __always_inline void ** radix_tree_iter_init(struct radix_tree_iter *iter, unsigned long start);
    7. static __always_inline unsigned radix_tree_chunk_size(struct radix_tree_iter *iter);

    Since you are evidently incompetent to read your own cite, when it refers to a header in C, I thought I’d help out there. Oh, and one more favour. There is, indeed, a single element of that header that qualifies as a non-trivial inline:

    static __always_inline unsigned radix_tree_chunk_size(struct radix_tree_iter *iter)

    The code for which is:

    if (flags & RADIX_TREE_ITER_TAGGED) {
    iter->tags >>= 1;
    if (likely(iter->tags & 1ul)) {
    iter->index++;
    return slot + 1;
    }
    if (!(flags & RADIX_TREE_ITER_CONTIG) && likely(iter->tags)) {
    unsigned offset = __ffs(iter->tags);

    iter->tags >>= offset;
    iter->index += offset + 1;
    return slot + offset + 1;
    }
    } else {
    unsigned size = radix_tree_chunk_size(iter) - 1;

    while (size--) {
    slot++;
    iter->index++;
    if (likely(*slot))
    return slot;
    if (flags & RADIX_TREE_ITER_CONTIG) {
    /* forbid switching to the next chunk */
    iter->next_index = 0;
    break;
    }
    }
    }
    return NULL;

    A not remotely interesting lump of bit-twiddling, unless of course you also use the meat of the thing, as contained in radix_tree.c.

    You don’t even understand the concept of “inlined code,” do you, Fifi?

  472. DrLoser says:

    DrLoser also the scsi subsystem core files are also not in that list either. So its not just radix_tree.c that is missing from the list.

    Did I say it was, Fifi? I didn’t say it was. Stop being paranoid. I’m offering up a neutral analysis, as far as the “incorporated parts” go. And that neutral analysis clearly requires the recognition of the “incorporated parts” that are missing from the list.

    What is your problem here?

  473. oiaohm says:

    DrLoser also the scsi subsystem core files are also not in that list either. So its not just radix_tree.c that is missing from the list.

    include/linux/radix-tree.h-v3.3-rc1

    Mind you the radix-tree.h as a header file appears. Wonder if their has been a lot of in-lining.

    http://lxr.free-electrons.com/source/include/linux/radix-tree.h

    Yes the header file is not just function declares and kinda requires radix-tree.c to function completely.

    DrLoser the more you look the more this looks iffy for vmware.

  474. oiaohm says:

    2.6.18 is a very old Linux kernel.

    DrLoser
    SECTION 2: GNU General Public License, V2.0

    Nice clear license delineation.
    Other than the fact its wrong.
    https://www.kernel.org/pub/linux/kernel/COPYING

    The Linux kernel code by Linus is under GPL V2.0 + Clause.

    6. Each time you redistribute the Program (or any work based on the
    Program), the recipient automatically receives a license from the
    original licensor to copy, distribute or modify the Program subject to
    these terms and conditions. You may not impose any further
    restrictions on the recipients’ exercise of the rights granted herein.
    You are not responsible for enforcing compliance by third parties to
    this License.

    Section 6 of GPLv2 makes removing the clause a breach. Why removing the Linus declare clause restricts the receiving users rights.

    I have actually read the damn thing, in the original German.
    Just to sux nuts the original you need to read is written in latin as part of a treaty between England, Germany and other countries of that area over copyright. The German version that you read is in fact a translation that is not the treaty. Just to be fun many of the translations have typos the treaty define defeats all. Yes Australian copyright is effected by the same treaty.

  475. DrLoser wrote, “slightly premature until you have read through the freely-available download”.

    I’ve read enough of it. It references RH 7.3, stuff I used around 2003 or so. They seem to be giving a snapshot of their old stuff and calling it the current release. Is there a time-warp going on that I don’t know about? What the lawsuit is about is the current release, I presume, and stuff that Hellwig complained about in 2006. That download tells us nothing of their current situation. So, it seems to be true that they are not divulging the source code as required. Surely, they are not still using RH 7.3! What about all the bugs/vulnerabilities they’ve fixed since those old days? Where’s the source code for that? Now, it could be that current customers are given other more current source code. That would technically comply with GPLv2, but you’d think somewhere in the world, someone would have leaked a GPLed copy, legally. Where is it? Perhaps Hellwig has it and it tells the story the court will see. There’s just something not right about VMware’s position. It looks a lot like lawyers have told them what to do as token compliance with GPL v 2 in hoping that none of the thousands of Linux developers would object. That’s pretty foolish. What’s with offering to settle under NDA? What value is an NDA in a case of FLOSS? There’s something broken here and I don’t think it’s Hellwig’s case. I think VMware made a mistake of some kind way back then and now have painted themselves into a corner. SCOG did that and got a prominent legal firm to sue the world over nothing at all. Legal advice about software seems to be unreliable. Courts and governments have to correct these kinds of blunders. It will be interesting to see the result and we should not have to wait long because German courts tend not to tolerate delaying tactics like SCOG used. Instead of 10 years like SCOG, I think we shall see this matter resolved one way or another this year.

  476. DrLoser says:

    It would be completely unfair of me to discount the internationally renowned knowledge of German Jurisprudence acquired partly by Prussian Ancestry and partly by a correspondence course spotted on the back of a bubble-gum packet drifting away in the wind on a particularly dispiriting “client free” day under a lamp-post in northern New South Wales.

    I do not wish to leave aside a possibly useful avenue of exploration. After all, the following assertion is positively magisterial:

    This is USA define
    “Copyright only protects the original expression”
    There is a word missing in the German define.
    “Copyright only protects the expression”

    How about it, Fifi? What does the Urheberrechtsgesetz have to say on that particular omission of an adjective?

    I have actually read the damn thing, in the original German.

    You haven’t, have you, Fifi?

  477. DrLoser says:

    It isn’t Friday, but I see no reason not to enjoy the following Fifi-ism:

    Derived work case would go a lot better for Vmware ESXi if they had done the same with freebsd as well.

    BWAHAHAHAHAHA!

    Oh, and incidentally:

    SECTION 1: BSD-STYLE, MIT-STYLE, OR SIMILAR STYLE LICENSES

    BWAHAHAHAHAHAHAHAHAHAHAHAHA FreeBSD

  478. DrLoser says:

    Where is the source code they think is “vmklinux”?

    An uneducated guess, Robert, although my uneducated guess is presumably as valid as yours, until we are both better informed.

    Might I suggest that ‘this source code they think is “vmklinux”‘ can be largely characterised by your excellent cite?

    Exercise the Second Freedom, Robert. Exercise the Second Freedom.

    I know I will, thanks to your diligence.

  479. DrLoser says:

    However, it either is or isn’t a derivative work of Linux.

    Staying with Exhibit Two (the Linux device driver API headers) for a moment, Robert, would you consider these to be either:

    a. Impossible to use without creating a derivative work of Linux or
    b. Essentially the same as an Open Firmware Standard

    I’m happy either way. It would be useful if you’d quit sitting on the fence, however.

  480. DrLoser says:

    Code analysis can wait, I feel. The License is the thing! OK, let’s divert ourselves momentarily with The License:

    open_source_license.txt … VMware ESXi 5.5 GA

    The following copyright statements and licenses apply to various open
    source software components (or portions thereof) that are distributed with
    this VMware software products.

    Good start.

    SECTION 2: GNU General Public License, V2.0

    Nice clear license delineation.


    >>> sata_nv-3.x
    >>> sata_promise-2.x
    >>> sata_sil-2.x
    >>> sata_svw-2.x
    >>> scsi/mpt/mptcommon/csmi/csmisas.c-4.23.01.00-8vmw
    >>> scsi/mpt/mptcommon/mptbase.c-4.23.01.00-8vmw
    >>> scsi/mpt/mptcommon/mptctl.c-4.23.01.00-8vmw
    >>> scsi/mpt/mptcommon/mptctl.h-4.23.01.00-8vmw
    >>> scsi/mpt/mptcommon/mptscsih.c-4.23.01.00-8vmw
    >>> scsi/mpt/mptcommon/mptscsih.h-4.23.01.00-8vmw
    >>> scsi/scsi_proc.c-2.6.18

    Interestingly, nothing about radix_tree.c.

    Lots and lots of headers, plus this SCSI stuff.

    Extracted purely in the hopes that it advances the discussion.

  481. DrLoser says:

    The real question is what is in their other code and whether or not it links to something derived from Linux. Methinks VMware’s lawyers are pulling a fast one, thinking one level of indirection is enough to hide the workings of their software. I would bet they will fight tooth and nail to prevent the court from seeing all their code. Where is the source code they think is “vmklinux”?

    Good to see you come back with a concrete argument, Robert. I’m assuming we are still discussing device drivers (Exhibits Two and possibly Three) here. You can return to algorithms (Exhibit One) should you so wish.

    And I want to congratulate you, sincerely, for taking the time (unlike oiaohm) to drag up a very relevant link: a download site for Open Source VMware ESXi.

    It’s quite large. It will take me a while to read through it and consider its impact on the current case. But, and thank you, Robert, it is actual relevant information for once.

    Of course, your conclusion that “VMWare’s lawyers are pulling a fast one” might be described as slightly premature until you have read through the freely-available download. What with that being Freedom 2: The freedom to redistribute copies so you can help your neighbour and all.

  482. DrLoser says:

    This clause of Linus is magical.

    Actually, I’m not even sure it applies to device drivers, which appeared to be your thesis.

    But what the heck, Fifi. White mice, pumpkins at midnight, whatever magic your little heart desires.

  483. DrLoser says:

    DrLoser aparently you cannot use google.

    http://www.twobirds.com/en/news/articles/2007/review-german-gnu-general-public-license

    If you had googled that broken link you would have found the working link.

    Fifi apparently you cannot supply a working link.

    It is not my business to follow the circus elephant around with a suitable capacious dung bucket.

    Just get it right next time. The onus is on you, not on your readers.

  484. oiaohm provided a link to VMware’s argument…
    “ESXi’s vmkernel has a stable, general-purpose API called “VMK API” that enables device drivers and other loadable modules to perform specialized functions.
    Third parties can write drivers and modules that interact directly with the vmkernel utilizing the VMK API. And while these drivers do not need to be Linux drivers, when they are, we offer a compatibility alternative through a loadable kernel module called “vmklinux”, which in association with any Linux drivers, is loaded by the vmkernel and interfaces with the vmkernel through VMK API.
    VMware offers vmklinux to third parties under the GPL and makes this source code available. For the reasons we’ve outlined above we are confident that our operating system is not a derivative work of Linux code and that we comply with our obligations under the GPL.

    If they provide their stuff in source code under the GPL, it’s not even relevant that “our operating system is not a derivative work of Linux code”. However, it either is or isn’t a derivative work of Linux. That should be easy to determine with code-comparisons etc. The real question is what is in their other code and whether or not it links to something derived from Linux. Methinks VMware’s lawyers are pulling a fast one, thinking one level of indirection is enough to hide the workings of their software. I would bet they will fight tooth and nail to prevent the court from seeing all their code. Where is the source code they think is “vmklinux”? ewwww! I looked at this link, which is their “Open Source” stuff for ESXi 5.5, and it seems to do a lot of binary replacements for RH 7.3… That’s depending on RedHat from 2002 for a release from 2013! What are they trying to hide?

    I don’t think black holes and time-warps will play well in court. As Linus likes to say, “show us the code”!

  485. oiaohm says:

    NOTE! This copyright does *not* cover user programs that use kernel
    services by normal system calls – this is merely considered normal use
    of the kernel, and does *not* fall under the heading of “derived work”.
    Also note that the GPL below is copyrighted by the Free Software
    Foundation, but the instance of code that it refers to (the linux
    kernel) is copyrighted by me and others who actually wrote it.

    Linus Torvalds
    This clause of Linus is magical. So if you write your driver in userspace before making kernel space version using normal system calls you are not limited GPLv2. Now if that work is now made a kernel mode driver sticking to the syscalls again it still not limited by GPLv2.

    Lets say you now skip the syscall wrapper in kernelspace. So your code contains a userspace and kernel space version using the same functions. Guess what . Still not a “derived work”. Why the driver operates in and outside the kernel space that is declared by license not be a derived work. This is what you call working inside license.

    The one court test to prove not derived is still being operational with the part removed.

    This clause of Linus is the only thing that really lets closed source drivers without having to be multi platform binary. Something horible about VMware kernel is the missing Linus clause.

    Vmware ESXi runs into few particular problems.

    VMK API is not implemented by any other kernel.
    vmklinux only works on Vmware ESXi vmkernel .

    VMware has taken the point of view that since VMK API can be used to implement other drivers that vmkernel is ok. Problem is the Derived work requirement is part of GPL ie vmklinux not the Vmware ESX vmkernel is that the GPL code will work without the closed source parts so its not a Derived work.

    Derived work case would go a lot better for Vmware ESXi if they had done the same with freebsd as well. So making the Linux part truly optional of course the court could still have banned Vmware from using the Linux parts.

    I would say Vmware suffers from lack of due care here. Questionable section of law should have seen due care in having a fall back replacement in case of legal trouble(BSD version).

    The problem for VMware is it would cost billions to make the network and scsi drivers.

    If you want a competitor example look no more than Xen domO. Yes it uses the Linux kernel for drivers but it can also use Freebsd and other OS’s for drivers. So Xen can prove independences.

  486. oiaohm says:

    DrLoser aparently you cannot use google.

    http://www.twobirds.com/en/news/articles/2007/review-german-gnu-general-public-license

    If you had googled that broken link you would have found the working link.

    Funny you didn’t mention it in the first place, oioahm. Nothing like insisting that people take this stuff on trust.

    Well, I don’t take it on trust. Where’s your evidence that the software in your cite was used for this purpose, oiaohm?

    It’s completely non-existent as usual, isn’t it?
    I find no reason to fill you in on a topic that TMR should have fully researched. Drloser you should have already know this stuff.

    The reality is everything in yellow in the diagram is questionable.

    http://www.vmware.com/company/news/vmware-update-to-mr-hellwigs-legal-proceedings

    Reading over this more closely this is going to come down to not derivative work agruement. VMWare is now saying that will give up the required GPL source code to go that path. There was a issue with VMware developers claiming the vmLinux was gone in the most recent version yet it still shipping with it and no source code.

    DrLoser here is something nasty. Most binary drivers with wrappers that have issues breaking between Linux releases turn out to be copyright infringement more often than not. They are using functions that there is no license exception to use and those functions the Linux kernel does not promise stability with.

  487. DrLoser says:

    I take it that we are all agreed that the Radix Tree implementation is of questionable use in the Nitwit vs VMware suit, then.

    Jolly good. Now, about those Linux driver headers. oioahm has expressed a very forceful opinion, and I admire him for that.

    Nobody else has.

  488. DrLoser says:

    Page not found.

    You really do squirrel away all this tosh like a brainless magpie, don’t you, oiaohm?

  489. DrLoser says:

    Already tested in German courts DrLoser.
    http://www.twobirds.com/en/news/articles/2007/review-german-
    2004 case was also brought by a non primary GPL author but by a author that had less than 4 lines of code and no header declaration.

    Is that the most exact comparison you can come up with, oiaohm? Four lines?

    Not relevant to the current case. Particularly considering that the “four lines” would be Hellwig’s, and the original work would be Momchil Velikov’s.

    A little more thought before you post further irrelevant blather, please.

  490. DrLoser says:

    DrLoser if you pull up the prior cases the diagram that has just been provided is always generated from the data from the binary search tool.

    Funny you didn’t mention it in the first place, oioahm. Nothing like insisting that people take this stuff on trust.

    Well, I don’t take it on trust. Where’s your evidence that the software in your cite was used for this purpose, oiaohm?

    It’s completely non-existent as usual, isn’t it?

  491. oiaohm says:

    oiaohm completely evaded this question and just repeated his original statement in slightly longer form. My conclusion therefore is that he has no evidence, nor even a suspicion, that it has been done in this case.

    In which case there is no reason whatsoever to bring up the possibility.

    DrLoser if you pull up the prior cases the diagram that has just been provided is always generated from the data from the binary search tool. So the evidence that a binary search has been done is the diagram. The statement that it contains modified Linux kernel source is what the binary search tool detects. It tells the difference between a binary build from stock sources or modified sources. Interesting that there is zero green most prior cases against hardware vendors some of the Linux kernel code in their kernel was displayed in green.

    There are standard ways to present http://www.binaryanalysis.org results in German counts due to how many times it been done. Sorry DrLoser this is not guess work you would have known it if you had looked over the prior evidence from prior cases.

    Portions Copyright (C) 2001 Christoph Hellwig
    This in the header is critical.

    And I tend to think that slapping another copyright notice on top (in this case, the GPLv2) is a prima facie evidence of such alienation.
    Already tested in German courts DrLoser.
    http://www.twobirds.com/en/news/articles/2007/review-german-
    2004 case was also brought by a non primary GPL author but by a author that had less than 4 lines of code and no header declaration. Ruling since all the Authors are licensing under the same license it is in fact acceptable by the courts as long as the prior notice is not removed and they can prove they submitted by a git log. Here is DrLoser making stuff up without proper research again. It is worse. If I had submitted code and it had been removed by authors following me due to the German law clause you just pointed out I could still do license enforcement because the future work could have been based off my work. So as long as you can prove by the git log that at some time you had a line of code in the file you can claim ownship for enforcement in a German count. So there are a lot more files that Christoph Hellwig could claim.

    Portions Copyright (C) 2001 Christoph Hellwig if you follow the git you will find Hellwig is in fact a joint author this is closer to a primary author than the first GPL case in Germany. So without question enforceable.

    And I tend to think that slapping another copyright notice on top (in this case, the GPLv2) is a prima facie evidence of such alienation.
    That is what VMware has done. There is no notice that their hyper-visor is using GPL code or obeying GPL rules. Notice the skype case in the link that Skype was libable for the infringement because it was profiting from the device. This VMWare case has nasty legs due to the Skype case. Yes hosting providers using VMWare could be in risk of legal problems.

    I don’t want to get ahead of myself, but this is an interesting statement of copyright intent, in re usage of Linux device driver code.
    Yes I know what stupidity you are thinking. Already argued by a hardware vendor. Yes there is a reason why in the digram the drivers are displayed in their own box. Problem is here all drivers in the Linux kernel has there own individual License declare so in fact are not part of the Main Linux kernel license. Are effected by the main linux kernel when using particular code that pulls in in-lines and the like from header files that are covered by GPLv2.
    http://www.tldp.org/LDP/lkmpg/2.6/html/x279.html
    The fact that each driver has its own individual license to the main kernel code is why some drivers can be shared with FreeBSD.

    Also most people are not aware that the Linux kernel has a “”Proprietary” [Non free products]”
    mode for drivers.

    People forget a long time ago the Linux kernel shipped in two halves 1 file kernel and 1 file all the modules.

    So a binary blob with a wrapper is perfectly legal to use with Linux as long as you only use the kernel features exposed for Proprietary drivers. Yes just happen to be all the features exported to user space.

    DrLoser the main copying file in the Linux kernel source is not the License of the drivers.

  492. DrLoser says:

    I’d like to examine Exhibit Two (Linux device driver headers) further, but whilst we’re waiting for the slower members of the Community sitting at the back and refusing to commit to an opinion on copyright, it would be remiss of me not to point out that oiaohm has just offered up a whacking great hostage to fortune on Exhibit Three … which we will presumably get to soon:

    NOTE! This copyright does *not* cover user programs that use kernel services by normal system calls – this is merely considered normal use of the kernel, and does *not* fall under the heading of “derived work”.

    I don’t want to get ahead of myself, but this is an interesting statement of copyright intent, in re usage of Linux device driver code.

  493. DrLoser says:

    On a similar point, Robert, have you yet decided whether algorithms are copyrightable or not?

    You’ve been silent on this one for days now. I wonder why?

  494. DrLoser says:

    That’s just a stupid conclusion. Oiaohm is not likely a party to the case so does not know the details yet it is widely known that complaints of violating copyright in busybox have been widespread because it is such useful software and folks think they can get away with it.

    It’s a perfectly reasonable conclusion.

    My question was, has this been done in this particular case?

    oiaohm completely evaded this question and just repeated his original statement in slightly longer form. My conclusion therefore is that he has no evidence, nor even a suspicion, that it has been done in this case.

    In which case there is no reason whatsoever to bring up the possibility.

    It doesn’t even qualify as “hearsay.”

  495. DrLoser says:

    This is USA define
    “Copyright only protects the original expression”
    There is a word missing in the German define.
    “Copyright only protects the expression”

    Oh, goody. I was beginning to believe that you had run out of things to be completely ignorant about, Fifi; apparently your store of such things, unlike your red leather miniskirt, is bottomless.

    The “German define” isn’t quite as simple as you would have it. In fact, in some respects, it’s actually far worse for your purposes than the “USA define:”

    Das Urheberrecht kann in Erfüllung einer Verfügung von Todes wegen oder an Miterben im Wege der Erbauseinandersetzung übertragen werden. Im übrigen ist es nicht übertragbar.

    My emphasis. You could make a case, with German copyright law, that any copyrightable artefact produced by a German and in consequence under the auspices of that law cannot have its copyright alienated. And I tend to think that slapping another copyright notice on top (in this case, the GPLv2) is a prima facie evidence of such alienation.

    Which is interesting.

    1. Under the Urheberrechtsgesetz, Herr Hellwig doesn’t seem to have any copyright authority at all over the radix_tree.c implementation. The author of this implementation, after all, is Momchil Velikov (unless traced further back, which is possible).
    2. On a larger scale, it sounds like any German contributor to a GPLv2 project — any project — is entitled to assert authorship and to reclaim the copyright thereof.

    Now, I don’t think that would stand up if, say, the author revoked the privileges of Linux to use that code. I assume this comes under “established usage,” or some such.

    But I do suspect that said author could, in principle, and according to the Urheberrechtsgesetz, allow redistribution of his/her code even when supposedly limited by the GPLv2.

    It doesn’t apply in this case, but I see no reason why it shouldn’t apply in future.

  496. DrLoser, being a pompous ass, wrote, “I take it that you’re admitting that, in this case, it hasn’t been done. Which was the actual question at issue.”

    That’s just a stupid conclusion. Oiaohm is not likely a party to the case so does not know the details yet it is widely known that complaints of violating copyright in busybox have been widespread because it is such useful software and folks think they can get away with it. Even the GNU “strings” command reveals, “BusyBox is copyrighted by many authors between 1998-2012.
    Licensed under GPLv2. See source distribution for detailed copyright notices.”
    An EULA may prohibit doing that but a contract made for an illegal purpose is void and won’t stand in court.

  497. DrLoser says:

    Lots of things can be done. Has this been done?
    http://sfconservancy.org always using the binary scan tool. They have brought many cases against device makers already over busybox and Linux kernel usage.

    I take it that you’re admitting that, in this case, it hasn’t been done. Which was the actual question at issue.

    Why don’t you just come straight out with an honest admission every now and again, oiaohm? Are you allergic to honesty?

  498. DrLoser says:

    But the K&R implementation was copyrighted by German law.

    Indeed? In that case you will have no problem providing us with evidence of this little detail, oiaohm.

    My understanding is that absolutely every last bit of K&R “C Programming Language” is copyright Bell Labs (and presumably their successor organisation). But it’s copyrighted in the normal way that a book is copyrighted, which allows for extensive “fair use.”

    Considering quite how often the K&R implementation has been quoted, used, and embedded in other people’s systems, I find it hard to imagine even the most swivel-eyed German Court in existence enforcing the copyright. In which case, it’s a pretty worthless copyright as far as the actual code goes.

    The only difference I see between qsort and the Hellwig Radix Tree is that somebody slapped a GPLv2 license on the latter at some point. But slapping a copyright license on something, and enforcing that copyright in court, are two entirely different matters.

    After all, as Robert is fond of telling us, you can’t copyright an algorithm. And if the Court agrees with that basic principle, then it’s hard to see how an isolated use of the Hellwig Radix Tree implementation cannot be asserted as either “fair use” or “de minimis.”

  499. oiaohm says:

    Lots of things can be done. Has this been done?
    http://sfconservancy.org always using the binary scan tool. They have brought many cases against device makers already over busybox and Linux kernel usage.

    Nobody has commented on K&R quicksort, so I believe we can all agree that K&R quicksort was never copyrightable.

    Nobody has seen fit to draw a distinction between K&R quicksort and “Hellwig Radix Trees,” so I will tentatively assume that we all agree that the “Hellwig Radix Tree” is, similarly, not copyrightable.
    DrLoser this is so far wrong its not funny. K&R quicksort algorthim was not copyrighted correct. But the K&R implementation was copyrighted by German law.

    Basically you cannot use copyright to prevent someone implementing there own version of an algorithm. But they cannot just copy your work either.

    With one caveat*, I hereby assert that the APIs exposed by a Linux header file are not, in fact, copyrightable under GPLv2. In fact, should they be globally agreed to be copyrightable, then the entire position of firmware vendors who bundle Linux device driver blobs would be undermined, because obviously they do so according to the exposed API. And would therefore be open to being sued for copyright infringement.

    Since this hasn’t happened, I’m making the pro-tem assumption that there is no reason for the German Court to find otherwise in the case of Nitwit vs VMWare.
    You need to check the German Court record sfconservancy has already sued hardware vendors and won over these points. You assumption is completely wrong. There is a clause on the Linux kernel that allows some binary drivers.

    NOTE! This copyright does *not* cover user programs that use kernel
    services by normal system calls – this is merely considered normal use
    of the kernel, and does *not* fall under the heading of “derived work”.
    Also note that the GPL below is copyrighted by the Free Software
    Foundation, but the instance of code that it refers to (the linux
    kernel) is copyrighted by me and others who actually wrote it.

    Linus Torvalds

    Nvidia and AMD closed source drivers in kernel space in fact interface with the kernel by syscall functions so are outside GPL. Also there are some header files in the Linux kernel taged with include exceptions.

    DrLoser you are making the most common mistake this case is in Germany. Not copyright-able does not exist in Germany. You can copyright everything. The thing becomes enforcement.

    This is USA define
    “Copyright only protects the original expression”
    There is a word missing in the German define.
    “Copyright only protects the expression”

    This limitation in the USA comes from the 1879 case of Baker v. Selden that never happened in Germany. In fact it never happened in the UK so does not apply to Australia either.

    Copyright is a legal right created by the law of a country, that grants the creator of an original work exclusive rights to its use and distribution
    German define of original work is anything you create. Does not matter if it based on an algorithm or not.

    DrLoser this is a German court the copyright law there is different.

    Australia, UK and German in copyright right has the terms implementation/work and idea/algorithm. Yes you can reimplement an idea/algorithm without legal trouble. But you cannot straight up copy the work.

    DrLoser the exception to copyright for header files has never been proven in the German or UK court system either. Most of VMWare possible defenses depend on a USA count system.

  500. DrLoser says:

    Anybody here ever written a Radix Tree (as I have done. Twice)? No? I thought not.

    Anybody here ever considered the possible algorithmic niceties of Radix Trees, including specific adjustments for, say, SCSI drivers in the Linux kernel? No? I thought not.

    Anybody here got a valid opinion on the twin possible extensions of black/white nodes and adaptive node sizes in a Radix Tree? (I have.) No? I thought not.

    Never mind. What with being actually qualified in the field, and what with actually having experience in the use of algorithms — copyrighted or otherwise — I probably have what most people here would consider to be an unfair advantage which, as luvr observes (irrelevantly and inaccurately) is liable to “derail the topic.”

    I’m OK with having my unfair advantage on Exhibit One being completely ignored. It’s pretty much what I expected.

    Care to try out my unfair advantage on Exhibit Two — the headers for Linux Device Drivers?

    If it makes you feel any better, you can give up after twelve months’ of futile attempts to discover the truth.

    According to both oiaohm and Robert, this is precisely what German Courts do, all the time.

    And who could deny the obvious fact that oiaohm and Robert both have extensive experience of the mechanisms through which German Courts work?

    Not I. I would never be so presumptuous.

  501. DrLoser says:

    In SCOG v World we saw SCOG claim the sun and then delay for years afterwards.

    It is the mark of a gentleman to concede minor and otherwise inconsequential points in a debate, Robert. It is the mark of an obsessive to ignore such points.

    I’ve already pointed out that there was no such case as SCOG vs World. I’ve already pointed out that this is not even a supportable analogy, given that the actual case was SCOG vs Novell.

    And we all know how popular Novell is in FLOSS circles. Why, there’s even a Boycott Novell Wiki!

    I advise you to listen to your inner instincts as a gentleman, Robert, and thereby cease and desist this foolish calumny.

  502. DrLoser says:

    One last retort to oiaohm on Radix Trees:

    the usage of radix_tree.c is provable. Why turns out Linux scsi drivers don’t work without it.

    There are three blatantly obvious objections to this supposition (which has no cite behind it).

    1. You seriously need some sort of evidence that the precise Hellwig &al implementation of Radix Trees is a sine qua non for using a Linux scsi driver in a Linux environment.
    2. You seriously need equivalent evidence that the precise etc implementation is a sine qua non for using a Linux scsi driver in any other environment. Surprise! ESXi is not Linux!
    3. Point three is so bleeding obvious that I’m just going to leave oiaohm flailing around for his standard six days or so before he finally picks up on the copious clues I strew along the way.

  503. DrLoser says:

    As an overall observation, I encourage everybody who comments to remember the de minimis rule on copyright. You remember the de minimis rule? It featured quite heavily in the Oracle vs Google court battles over Java.

    In the particular case of Exhibit Two (Linux device driver APIs), it would be fairly easy to argue de minimis, I think. Most particularly since this information is essentially equivalent to an Open Standards Specification for Device Drivers.

    And Open Standards are always desirable things, aren’t they? Nobody here has yet made the insane claim that Open Standards should be constrained by any form of Gnu Public License whatsoever

    (Over to you, oiaohm.)

  504. DrLoser says:

    I’m happy to leave the topic of algorithms as soon as the relevant people state their position on Radix Trees (implementation thereof), but while we’re here, how about.

    I’m equally happy to leave the topic of algorithms, having given the relevant people ample opportunity to state their case on Exhibit One (algorithms).

    Nobody has commented on K&R quicksort, so I believe we can all agree that K&R quicksort was never copyrightable.

    Nobody has seen fit to draw a distinction between K&R quicksort and “Hellwig Radix Trees,” so I will tentatively assume that we all agree that the “Hellwig Radix Tree” is, similarly, not copyrightable.

    Robert has not denied my statement of his position as “you cannot copyright an algorithm,” so I will tentatively assume that he agrees on this point: VMware is not liable for copyright infringement as far as the “Hellwig Radix Tree” goes.

    I say all this to be fair to oiaohm and to advance the discussion to Exhibit Two: device drivers. Actually I’m going to cheat a little and make “device drivers” per se Exhibit Three. Before making use of a device driver originally written against the Linux device driver API, you have to implement the Linux device driver API. Which appears to be what VMware did with ESXi, so let’s go with that as Exhibit Two.

    Now, the way you do this (and I believe the way VMware does this) is to … well, OK, it’s obvious. You examine the device driver API. Which is, straightforwardly, available in Linux header files (written in C).

    With one caveat*, I hereby assert that the APIs exposed by a Linux header file are not, in fact, copyrightable under GPLv2. In fact, should they be globally agreed to be copyrightable, then the entire position of firmware vendors who bundle Linux device driver blobs would be undermined, because obviously they do so according to the exposed API. And would therefore be open to being sued for copyright infringement.

    Since this hasn’t happened, I’m making the pro-tem assumption that there is no reason for the German Court to find otherwise in the case of Nitwit vs VMWare.

    Whether or not this is true will be important for both Exhibits Three (actual use of Linux drivers) and Exhibit Four (to be revealed). In the mean time, I throw the issue out for discussion. And I mean discussion. Don’t assume I hold an unshakeable belief in this; don’t assume prejudice on my part. I’m just trying to figure out what this court case is all about.

    As part of Exhibit Two, btw, I’d be interested to know exactly what vmklinux contains, and which parts of it potentially infringe GPLv2 copyright. The reason that I’m interested here is that, as far as I can see, all it needs to contain is a small wrapper (entirely home-grown) around the Linux device driver APIs. There may be more to it; I don’t know.

    * The caveat is obviously that C header files can contain actual code. For simplicity’s sake I’m assuming that none of this actual code has been carried across to ESXi. Perhaps it has. That would be a different issue for a different Exhibit.

  505. DrLoser says:

    The reality is Ribben has eaten up more of your screen space than the Sidebar/menu combination Libreoffice uses.

    The twin realities are:

    1. This was not the point at issue
    2. Nobody cares anyway.

    I’d never seen the sidebar before (other than on opening Libre in the first place, where it offers you the slightly redundant choice of a new Writer file, a new Calc file, etc), so I tried it. It’s a distracting waste of time, isn’t it? Who on earth wants to switch focus between the top of the screen and the side of the screen all the time?

    The default options (text and paragraph formatting) occupy a huge amount of space for no appreciable value. This is 4.3, so maybe (just maybe) it’s possible to customise the thing a la Microsoft Ribbon, but somehow I don’t see this being the rip-snorting feature that you suggest, oiaohm.

    It isn’t a question of making more screen real-estate available. It’s about using it intelligently. The Side Bar would appear to have had absolutely no intelligence applied to it whatsoever.

    You know what? Forget slavishly imitating Microsoft (and getting it horribly wrong). Why not slavishly imitate Adobe (and get it horribly wrong).

    Adobe actually uses their version of the side-bar for something not otherwise visually representable, to whit, thumbnails for pages in a long document.

    That would have been useful. This excrescence, not so much.

  506. DrLoser says:

    Yes the open source tools to slice and dice ESXi binaries and find the GPL stuff exist.

    So what?

    Lots of things can be done. Has this been done?

  507. oiaohm says:

    DrLoser closed source is not that closed.
    http://www.binaryanalysis.org/en/home

    Yes the open source tools to slice and dice ESXi binaries and find the GPL stuff exist.

  508. oiaohm wrote, “Germany has a 12 month rule”.

    They have other rules like not allowing lawyers to delay anything. In SCOG v World we saw SCOG claim the sun and then delay for years afterwards. That just doesn’t fly in Germany.

  509. oiaohm says:

    Deaf Spy
    Except that it doesn’t because it makes it harder to display two pages side-by-side without zooming out. You’ve never actually done any serious writing, haven’t you?
    You have never used the sidebar. The side bar in libreoffice fold out of way.

    16/9 works out slightly wider than 2 A4 pages when sidebar is folded 2 pages fit perfectly. Remove the toolbars and keep the menu and sidebar you have the most screen space to display the 2 A4 pages.

    However, on a full hd monitor (1920×1080), having two A4 pages in full screen, with Ribbon on top, is just natural.
    The reality is Ribben has eaten up more of your screen space than the Sidebar/menu combination Libreoffice uses.

  510. oiaohm says:

    DrLoser
    http://www.opensource.apple.com/source/gcc/gcc-1765/libbanshee/libcompat/radix-tree.c
    This here is possible violation. It playing in a really unknown gray area. Everything in libcompat on different platforms libbanshee can be built without. So libbanshee is depending that it classed as a not derivative work because if its not a derivative work the rest of the library does not need to be under GPL. But unlike vmware the source is still published as required by GPL so they can attempt the not derivative work arguement.

    And if you come over here.
    https://github.com/gcc-mirror/gcc/tree/gcc-4_9_2-release
    You will notice no libbanshee it has been rejected for include in gcc due to possible licensing issues. Fairly much that libcompat directory is a problem that apple might get sued over. I would recommend finding a non GPL version of radix_tree.c

    Do you regard the use of radix_tree.c in ESXi as a violation of copyright or not? It isn’t proven that the code is even present in ESXi. But, if it is, does it breach copyright?
    ESXi is different to the libbanshee problem.
    1 the usage of radix_tree.c is provable. Why turns out Linux scsi drivers don’t work without it.
    2 Vmware unlike libbanshee has not published the source code ESXi uses. To use the Derivative Work defense you must either publish the source code or make a offer to provide the source code this does not have to include your code but has to include all the GPL parts. So everything in yellow should have been published.

    The problem here is in Germany FOSS developers can choose to take payment in source code. So Payment for not publishing the GPL parts could be all the source code to ESXi completely avoiding not derivative work defense.

    Fair use and fair dealing provisions in German counts don’t apply to Public Domain or Open Source Licenses. Yes using a Public Domain work or Open Source work without declaring is classed as stealing from the Public.

    Plagiarism in Germany is an offense. You see people forced out of government in Germany for doing it. Yes you can get a criminal record in Germany by doing Plagiarism and getting caught.

    The nasty reality here is Vmware might enter court judge rules Plagiarism and its over with Vmware forced to pay in source code never getting to the GPL license bits.

    Germany in a Copyright case is not defendant friendly. Vmware need todo a lot of things before the court date comes up.

    Germany is one of the few countries where you can choose to take source code instead of cash. This would also explain why they are raising money to cover the count case.

    There are very big reasons as a international company never to infringe on Open Source copyrights. One of those is German counts.

    DrLoser you example you have not stated who the author is. Hopefully its code you wrote otherwises it technically infringement by German rules. If it from an open source or public domain work and you are in Germany you are screwed. If its from book or website with all rights reserved copyright you could argue fair usage.

    DrLoser I guess both of us are not in Germany and the site is USA/CA so we don’t have to play by these strict rules. FOSS enforcement is so likely to happen in germany its not funny.

    Germany has a 12 month rule as well if a copyright case drags out for more than 12 months injunction relief must be granted. So this should be over in 12 months. Settlement, Court ruling… something inside 12 months.

  511. DrLoser says:

    I’m happy to leave the topic of algorithms as soon as the relevant people state their position on Radix Trees (implementation thereof), but while we’re here, how about:


    /* sort v[left]...v[right] into increasing order */
    void qsort(int v[], int left, int right)
    {
    int i, last;
    void swap(int v[], int i, int j);
    if (left >= right) /* do nothing if array contains */
    return; /* fewer than two elements */
    swap(v, left, (left + right)/2); /* move partition elem */
    last = left; /* to v[0] */
    for (i = left + 1; i <= right; i++) /* partition */
    if (v[i] < v[left])
    swap(v, ++last, i);
    swap(v, left, last); /* restore partition elem */
    qsort(v, left, last-1);
    qsort(v, last+1, right);
    }

    Is that copyrightable? And if not, why not?

  512. DrLoser says:

    I should point out that, technically, I believe that VMware are very probably in breach of copyright in all four areas I think we should discuss. The Radix Tree is just the first of the four.

    I’m not arguing for the defense here. I’m just trying to apply what little forensic analysis is available to us all. And it doesn’t really help if Robert won’t even cough up a position on Exhibit One.

    I mean, oiaohm and me haven’t even yet got to Exhibit Two.

    What’s the problem here, Robert? It was your chosen topic of discussion in the first place, wasn’t it?

    Did I frighten you off by reminding you of your rather questionable line on algorithms, the copyrightability of such?

    Did I? Well, you’re going to love it when we get to drivers. You’ll have even less to say on the matter.

    And then there’s the other two topics that are patently trailered. Do you have anything to say about vmklinux, for example?

    Don’t let oiaohm beat you to it! Be your own man, Robert!