Mini Versus Micro

I can remember when minicomputers extended the reach of the mainframe and eventually competed with mainframes. Then microcomputers came along and competed with both. Now, Mini v Micro is back in the news another way, software.

The x86 personal computers have long been powerful enough to run the applications of multiple users. I have often run a whole computer lab from a single PC running GNU/Linux. I used the lab’s PCs as thin clients. There is another way to use a PC for multiple users and that is to connect multiple monitors, keyboards and mice to one PC. That is what MiniFrame has been doing from M$’s OS. M$ liked that so much that they revised their EULA for Vista to include a “single-user” requirement.

MiniFrame sued M$ for anti-competitive behaviour. You can read their complaint here. The document is 66 pages long and includes these details:

  • M$ has a monopoly on the OS with ~70%+ on clients and servers,
  • before 2007 one could legally share one of M$’s PCs with multiple simultaneous users without a server,
  • MiniFrames’ SoftXpand PC-sharing software became popular and was used in 30 countries,
  • Another company, NComputing did the same,
  • In 2009, M$ made an exclusive deal with NComputing to include M$’s server OS in their PC-sharing system,
  • Vista EULA:2.b. Number of Users. Except as provided in the Device Connections (all editions), Remote Access Technologies (Home Basic and Home Premium editions) and other Access Technologies (Ultimate edition) sections below, only one user may use the software at a time.
  • M$ interfered with MiniFrame’s customers and OEMs to prevent them using MiniFrame’s software,
  • M$ refused to make a licensing deal with MiniFrame,
  • MiniFrame lost tens of $millions of business, and
  • MiniFrame demands an injunction against M$’s anti-competitive behaviour, $1 billion, costs and punitive damages.

The important legal issue in all of this is that copyright law is about making copies and MiniFrame is not suggesting or promoting the copying of anything from M$ so M$ is attempting to extend copyright law by its contracts/EULAs. If MiniFrame wins something here, it could have a weakening effect on the cursed EULA which is designed to take away the rights of people to use software and hardware that they have acquired legally. MiniFrame’s SoftXpand software is just an application that runs on a PC. M$ has no right to prevent anyone from running that software, IMHO. They do not copy M$’s OS. They do not violate M$’s copyright at all in doing so.

M$ of course, denies everything, even monopoly, and claims it can put whatever terms it wants in its contracts. Both are false. If it walks and talks like a monopoly, it is one. A contract is not a contract if it attempts to do something illegal like stifle competition. Clearly, M$ is doing that.

Whatever the outcome, the case clearly documents the high costs of using M$’s software. Both M$ and MiniFrame can be cut out of your budget by using Debian GNU/Linux. I was using thin clients with GNU/Linux back in 2003 with great success and similarly used multiseat systems in 2006 all with no licensing fees, just Free/Libre Open Source Software (FLOSS). The X Window System has allowed users of GNU/Linux to set up multiple displays or multiple users using multiple displays for years on one or many PCs.

see

GNU/Linux works for you. There’s no need to sue anyone over using it.

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. Bookmark the permalink.

51 Responses to Mini Versus Micro

  1. Hi,
    I’ll give you an angle from someone who sold SoftXpand back when Windows XP was the Operating System. I worked for the UK distributor of SoftXpand and we were hugely successful in a very short space of time. We sold into 20% of Derbyshire Schools for example, won numerous awards including the Dell Technology Innovation of the Year award, had large UK organizations like The Co-Operative Group, and EOn ready to buy in their thousands, major banks, hundreds of SMBs, Universities with thousands of users lined up, pitched successfully to The UK Governments Parliamentary IT Committee, and then bam… Microsoft started making public comments “suggesting” that 1 Windows XP PC serving 6 users was a direct contravention of their EULA; this pretty much ended sales overnight, and for what? Revenue protection. We attempted to challenge this and were stone walled; the fact is that Microsoft never returned calls, never agreed to meet, never even acknowledged us…. All we and our end users and VAR partners ever wanted was clarification; everyone would have been happy to buy OEM versions of XP for each user because the benefits were in Green IT, the electricity savings, the reduced number of PCs to buy, support, network and dispose of. The goal was not to rob Microsoft of their desktop OS revenue. The fact remains that multiple users can share 1 copy of a Microsoft OS, just not at the same time. SoftXpand allowed this. The product was fantastic. Some users want a near to true desktop PC performance and SoftXpand was the product to achieve this; albeit in a messy way (It is a close proximity solution with users needing to be within spitting distance of each other, and great cable management is a must). I work for a manufacturer now and shared resource computing still trumps server based computing for performance and manageability. The concept was good though. My question to the board is why Microsoft brought out Windows MPS2010 which is in direct competition to SoftXpand but based on Windows Server 2008 R2..? Why they changed their EULA on Vista and 7 to have a Single User element? Why they allowed NComputing and LG and HP to be part of their alliance and not MiniFrame? Show me an NComputing deployment running on MPS2010 that actually works! All of the users I know have had to downgrade to XP to reap the benefits and justify their expenditure. Line up all of the products side by side and SoftXpand wins, but it is now redundant, and priced out of the market, plus illegal according to the Windows Vista and 7 EULAs. An extremely good product was shafted overnight, and I am convinced that MiniFrame have grounds for a legal case. I watched a company fold and have to close its doors after going from a 1 man company to 20 staff in rapid time, I saw experienced businessmen reduced to tears, I saw investors lose their shirts, and I saw the disappointment on the faces of clients when they turned round and said “we simply cannot risk buying SoftXpand and being sued by Microsoft for incorrect licensing”. MiniFrame did not copy Microsoft’s IP, it just allowed multiple concurrent users to share 1 copy of it’s OS simultaneously. They could easily have worked with MiniFrame and come up with a desktop licensing model with user CALs, but they chose not to, and that is their right ultimately. Unfortunately they exercised their right which has once again put people off taking risks in building innovative products and the world will be worse off. The argument here is whether Microsoft were anti-competetive and the answer is; YES THEY WERE.

  2. oldman says:

    Once more with feeling Pog…

  3. oldman says:

    “You most likely will have to wait a week to read this oldman because I would bet you don’t have a subscription here either.

    http://lwn.net/Articles/485263/

    Actually I read up on the original.

    http://wiki.debian.org/Multiarch

    So the Debian multiarch project is going to change the world, eh? One can never know the future but I will be that we are looking at at least 12-18 months before it goes unto unstable and quite a bit beyond that before it goes into stable. – lets be generous and say we are not ca. 2-1/2 years out. By that time windows 8 on ARM will be out for at least 1 year. So much for something getting there before windwos 8 on ARM.

    “This is the problem oldman I have access to what is going on you don’t.”

    Then having all of this knowledge at your fingertips you have run across this article

    http://www.theregister.co.uk/2011/11/01/hp_redstone_calxeda_servers/

    My favorite line of this article is, of course…

    “The big, big caveat is, of course, that you need a workload that can scale well on a modestly clocked (1.1GHz or 1.4GHz), four-core server chip that only thinks in 32-bits and only has 4GB of memory.”

    And then of course there is the ars technica article

    http://arstechnica.com/business/news/2011/11/does-hps-project-moonshot-have-enough-gas.ars

    Which contains a very interesting set of comments by Google Senior Vice President of Operations Urs Hölze

    http://static.googleusercontent.com/external_content/untrusted_dlcp/research.google.com/en/us/pubs/archive/36448.pdf

    Both of these articles pretty much mirror upon my feelings about you attempts to sell ARMageddon for microsoft on the short time frame that you are selling. HP moonshot is IMHO at least 2-3 years away, if ever, from commercial viability. The multiarch support that the debian hackers are pulling together if probably just about the same time frame, which would be well after after the even worst carse ship date from microsoft ARM version of windows 8.

    In short , you have produced NOTHING that proves your insinuation that an ARM based x86 emulator capable of running windows applications will be out before windows 8 on ARM. In fact your have NOTHING but your own speculation to stand on. Which as far as I can see, is nothing more than bushwah.

  4. Now that I have a layered spam-filter the number of real spam getting into my queue is tiny so it is easy to find the good stuff. Still a nuisance, but a much smaller problem. Before, I was having to scroll through two long pages to fish them out. Now, it’s a few clicks and I am done. The only real problem is that I have a life and occasionally stray from the keyboard…

  5. oldman’s linked site contains:“One big problem with Microsoft’s licensing rules is the requirements that hosting hardware must be dedicated to, and for the benefit of the customer, and may not be shared by or with any other customers of that partner.”

    That is a purely imposed restriction unrelated to technology. The whole idea of the cloud is to share resources so the individual cost becomes tiny. M$ is preventing that. It is silly to have a server that can support N users running only 1. M$ clearly does not believe server-consolidation should affect them. This is the same thing M$ has been doing to OEMs for decades with their petty restrictions. The whole world hates M$. Eventually there will be many volunteers to stick the knife and twist it.

    The key thing here is that M$ wants to be paid N times for giving 1 unit of service. If they charged per CPU-second, for instance, they would only get one licence:

    M$’s “just” payment = (Total CPU-seconds used by users)/(31million seconds per year) X 1 licensing fee which comes to perhaps 10% usually and at most 100% of a licensing fee. What they want is N licensing fees. They want to be paid far more than they are worth. They don’t want service to be seen as a commodity but as a privilege they can give or take away. They don’t want Moore’s Law to reduce their revenue as it does for everyone else in IT.

  6. oldman says:

    !@#$ spam filter

    fish me out please.

  7. oldman says:

    Pog

    An interesting take on this whole issue. It seems to me amuch as non partisan take on what is going on as any.

    http://searchvirtualdesktop.techtarget.com/news/2240146634/Microsoft-clarifies-cloud-hosted-desktop-licensing-stings-OnLive?asrc=EM_NLT_16653299&track=NL-463&ad=865216USCA&

    Thoughts?

  8. oiaohm says:

    You most likely will have to wait a week to read this oldman because I would bet you don’t have a subscription here either.

    http://lwn.net/Articles/485263/

    Welcome to pandora box oldman. Now that a method to run emulation with reasonable effectiveness cross cpu types. Debian is now changing to be able to install all Arch types on the one system.

    The complete idea of needing binaries that match you arch is ending. Of course running non arch native is going to come with a price of 1/2 the number of core or less.

    But when you have arm units like the HP Moonshot half the number of cores is not a major issue. Even with half the cores is still way more processing cores than what you can put there using x86 current designs.

    The stuff I use todo special is going to become the norm oldman.

    This is the problem oldman I have access to what is going on you don’t.

  9. oiaohm says:

    http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5578325

    You asked for the paper here it is the first one in the chain oldman.

    oldman the trick is coping exactly what the x86 hyper-threading is. Translator engine runs on one core and processing engine running x86 runs on another. When processing hits a non translated part it changes threads until that sections is translated.

    This avoid the qemu issue of thump stop processing while I prep code. This is what kicks the crap out the emulation application running run stall run stall not a good thing.

    This also leads to caches in cpu’s being cleared by the translation engine starting up so when you swap back to running the converted code you have taken one hell of a speed hit with the qemu current design. Yes the stall. Don’t run translation and execution in the same cpu core if you want smooth emulation.

    There are latter papers as well. The tech is real that x86 is run-able but you are dependant on the way Linux kernel does a lot of things so it can work.

    Trade press articles is when in final production.

    So yes the technical articles are there now. There is some research into llvm for the translator with the possibilities of going to like a 120 to 130 or higher percent performance. Remember this is using 2 cores to emulate one x86 core. This is not a big problem thinking two arm cores take less silicon than 1 x86 core. Also two arm cores still use less power than 1 x86 core.

    90 percent from a silicon point of view. Less silicon cheaper chip. So an arm quad core equal a x86 duel core of the same speed.

    There is also a few looking when no code translated events that do happen inside all x86 processors since the age of pentuim.

    Due to arm design not using decanted hardware for translation and execution like x86 chips do this allows options of putting all 4 cores into translating code when none is ready and putting all 4 cores into running code when there is nothing to translate. Of course this has to be assessed for when this is wise. Yes nuking the cache in cpu to run something else hurts.

    There is a lot more research yet oldman. We are at a tip of iceberg. There is a good chance silicon vs silicon arm is going to hand current design x86 processors there head for not being flexible enough to make good use of the silicon.

    So yes the argument that arm does not run x86 by current research results looks like a joke oldman. In time arm looks like its going to kick x86 in every way unless something exceptional turns up in a x86 chip.

    Basically on a 4 core arm. Arm code rums 4 core and the x86 code runs equal to duel core or better. With both types of applications able to run on the arm cpu at the same time. Less silicon Less licensing costs so cheaper devices. Also more application support.

    In raw performance terms from 1/6 to 1/2. 1/2 is more than enough with how many times smaller a arm cpu is.

    90% of a duel core is really without optimise the process so the cores running the translation idle when there is nothing to translate and the cores running the converted code also idle when waiting on translation of code. So yes its going to get faster.

    Also qemu skips out on doing a lot of optimisations that the x86 chips real core translator does due to the single thread turn around issue so could not spend the time doing them since it would cause more lag. Duel core translator can spend more time checking for optimisations.

    Translators from arm to x86 have not been that successful.

    I think in silicon oldman since this equal cost that is where my 90 percent comes from.

  10. oldman says:

    “A Tegra 3 quad-core ARM beat a single-core x86 chip.”

    Unfortunately we don’t know by how much Pog, as this article does not mention numbers. We do know from the article that the difference between the dual core arm and the single core Atom as only 10%. That is not all that much to crow about IMHO.

    The bigger question however is,of course, what does this citation have to do with the assertion that a some magical person will be shipping an x86 emulator that will allow a Quad core ARM chip to run x86 code at 90% of the real processor speed?

  11. oldman wrote, a system that could emulate x86_64 cpu’s at 90% of speed on ARM would be big news.

    It was, last year. see Intel Medfield phone benchmarked: Bests Nexus but trails Tegra 3

    A Tegra 3 quad-core ARM beat a single-core x86 chip. That’s Android Dalvik interpreted code. That’s a good indication that interpreted x86 code would have the same result. Compare price/performance of Wintel and */Linux on ARM and ARM wins handsomely.

  12. oldman says:

    “Sorry oldman you are simple out of date on what is being developed. The game is changing at very quick pace at moment.”

    Nope. I deal in reality, and reality states that ANY such developments would have been announced in the market already. Furthermore to meet your stated timeline of being out before windows 8 such a product would have to have been in beta at least. And a system that could emulate x86_64 cpu’s at 90% of speed on ARM would be big news.

    I say again. do you have proof beyond general assertions, URL’s to technical papers? trade press articles?

  13. oiaohm says:

    oldman
    “Really? So now you are effective claiming that QEMU or some magical stuff will go to 16% of x86 speed to 90% of x86 speed in real stable shipping code before 1Q2013 when windows 8 comes out.”

    There is nothing magical about it. Its a simple design split. You are able to use the tested QEMU x86 to arm translator. So its not make a new x86 translator. It is how you use that translator.

    The simple fact is oldman you have not been watching the experiments to see the minor changes they did to get massive speed boosts.

    The bigger hits are coming from not recycling translated code as a x86 processor does. Copy the x86 processor design in implementation closer and the performance goes straight up. Yes adding a converted cache what is basically replicating what happens in l1 cache inside and x86.

    “Better still name the platform that it will run on.” Funded by ARM chip production companies themselves so x86 maker question is very stupid.

    Really the difference is single threaded design emulating a core that is multi threaded vs proper muli-threaded replicating a multi threaded system. Multi core cpus in volume on arm and mips have been a rarity. The result of going multi threaded is a massive speed boost.

    Sorry oldman you are simple out of date on what is being developed. The game is changing at very quick pace at moment.

  14. That only works for a while. Eventually, the OS in the VM will not be updated… Not a problem if the system is locked up tightly.

  15. oldman says:

    Pog

    Please fish me out of the spam bin…

  16. oldman says:

    “CodeWeavers is making a living selling Wine services for GNU/Linux and MacOS. Clearly, some businesses do pay to use Wine.”

    I am well aware that there are some individuals who are idealogically committed to running a pure linux shop but who are faced with needing to run some windows based application. For those of these people who have legacy applications that they are not updating that happen to run under the codeweavers environment, this product might be a good investment.

    However for anyone who needs or wants to keep their windows application up to date W(h)ine is a minefield.

    I have personally converted several individuals who had given up on W(h)ine to VMWare workstation on Linux. It was amazing how quickly they dumped the codeweavers kludge when they saw the VM working, and realized that they also has something that was supportable by the vendor to boot!

  17. oldman says:

    “Again did not answer my question in a case of Wine or Nothing? What are you going to choose.”

    We have played this game before and I have told you that you do not get to dictate my choices Mr. Microsoft VAR.

    I choose windows apps in a VM on Red Hat Enterprise Linux x86_64 if I have to.

    “Current qemu tech is 1/6 of the cpu performance in the emulated environment. The problem is this is not where emulation will be by the time Windows 8 releases.”

    Really? So now you are effective claiming that QEMU or some magical stuff will go to 16% of x86 speed to 90% of x86 speed in real stable shipping code before 1Q2013 when windows 8 comes out.

    I’ll tell you what…
    Name the commercial vendor who will be distributing this wonder.

    Better still name the platform that it will run on.

    Even better still name the x86 vendors who are going to bet the farm on this wonder.

    “Its about time you catch up with what is happening oldman.”

    Actually its time you come back down to reality Mr. Microsoft VAR.

    Hows that Linux on ARM business startup going BTW?

  18. CodeWeavers is making a living selling Wine services for GNU/Linux and MacOS. Clearly, some businesses do pay to use Wine.

  19. oiaohm says:

    oldman when does Linux have to face Windows 8 on arm. Not for another 12 months at least. More than enough time to go from the current prototype to fully functional emulation.

    This is the problem.

    oldman
    “Wine is a fact, as is the reality that it does not and can not deal provide what any business person would consider a full and supportable environment to anyone running linux. It os far easier for an individual to run an application in a virtual machine under linux even with the lesser freebie hypervisors that you like, than to play in the minefield that is wine. And I speak from experience on THAT sir!”

    Again did not answer my question in a case of Wine or Nothing? What are you going to choose.

    Current qemu tech is 1/6 of the cpu performance in the emulated environment. The problem is this is not where emulation will be by the time Windows 8 releases.

    The alteration in design splitting the x86 byte code translator from the native code generated so they can run on different cores even different designed cores alters the complete performance shape of x86 emulation on arm and every other platform.

    Of course you did not consider for one min this will allow full x86 virtual machines as well to run at almost native speeds on arm. So you complaint about wine is a no-op its you lack of knowledge about what is going on caused you to respond wrong oldman.

    There is research that is ground breaking. Biggest change in virtualisation in almost 12 years. This has completely missed your radar oldman.

    Its about time you catch up with what is happening oldman.

  20. oldman says:

    “So oldman you are person who magically believes solutions come out of mid air. They don’t.”

    Nope. I know all about prototyping, though it has been a long time. I also know that 99.9% of prototypes go nowhere. The fact that the chinese did an expetiment is meaningless. Show me the working product in commercial form – put up or shut up sir.

    “Linux means to run windows applications on any CPU is a tested fact with performance hits.”

    Irrelevant. They are not supported by the vendors wo make the code. What you or some other resusenik geeks up does not count in the business world.

    As usual sir, you are speculating on what hope might be based on arcana.

    “Of course oldman wants that no one is using wine so you can keep you head in the sand. Geek patrol excuse is how you avoid facts oldman. You can not forecast where something will go without facts.”

    Wine is a fact, as is the reality that it does not and can not deal provide what any business person would consider a full and supportable environment to anyone running linux. It os far easier for an individual to run an application in a virtual machine under linux even with the lesser freebie hypervisors that you like, than to play in the minefield that is wine. And I speak from experience on THAT sir!

    You facts as you call them may be there, it is your linking of them together to a specific goal that I question. You show all the symptoms of the typical techno geek, all speculation about the possibilities of some arcana with zero understanding of the reality of the steps that MUST occur to make your speculation into reality.

    Come to think of it, if you are so sure that this is the future, When are you abandoning your Microsoft VAR status and going into business selling pure linux solutions on ARM eh?

  21. oiaohm says:

    oldman

    “And near zero use beyond the geek patrol sir. Wine itself is a kludge that few if any business use, especially when it is easier to run the application in a virtual machine.”

    Notice something talk reality it has to be the geek patrol. Does not change the fact Linux world has been doing this for years. You call it a kludge I could call MS 32 bit support on 64 bit systems a kludge as well. Reality is Reality.

    Of course oldman wants that no one is using wine so you can keep you head in the sand. Geek patrol excuse is how you avoid facts oldman. You can not forecast where something will go without facts.

    Really this is your problem. If you had read my posts there was a experiment in china that proved the 90 percent by using all 4 cores in a big little system.

    http://www.youtube.com/watch?v=bVYlrv0kbdA People are trying may methods. Most following what the china experiment proved.

    oldman
    “As far as you panda board is concerned, another of the myriad of one off prototype boards that knock about. It has zero relevance to the mainstream of ARM computing.”
    Where do you think new tech starts off. Lot of ARM devices are development on the panda board.

    Panda board is more expensive that the final products in most cases. If you want to test it out yourself if you have a Panda board I might as well point you to the source code.

    So oldman you are person who magically believes solutions come out of mid air. They don’t.

    Linux means to run windows applications on any CPU is a tested fact with performance hits.

    Performance hits are being reduced as emulators are coping how a x86 really works. That the 1 x86 core is really 2 cores one processing and converting instructions to the processing cpu native language.

    This is seeing lot of overhead reduction and more x86 like performance on non x86 systems. There is a high possibility of even faster. The glitch in emulation has been the fact the translation and execution and been treated as same basic thread process.

    From 1/6 to 90+ percent by one change in emulation design. With 90+ running windows programs on Linux is more than practical. Ok limited to what wine supports but hey its better than nothing that the windows on arm is offering.

    Sorry oldman wine or nothing what wins.

  22. oldman says:

    “Linux has a long history of running on x86 on non x86 processes. ”

    And near zero use beyond the geek patrol sir. Wine itself is a kludge that few if any business use, especially when it is easier to run the application in a virtual machine.

    As far as you panda board is concerned, another of the myriad of one off prototype boards that knock about. It has zero relevance to the mainstream of ARM computing.

    “Dec-ate those two older design cores to instruction conversion shock horror you can get inside 90 percent perform of a x86 duel core the same speed. ”

    Proof please not geek speculation Mr. Microsoft VAR.

  23. oiaohm says:

    Clarence Moon “Linux based netbooks early on” did have other issues other than just being a odd ball.

    From the server room Linux had been power hungry. So Windows based netbook would run longer than a Linux Based one on the same hardware.

    That the Linux based netbook sold at all showed interest they were technically beaten by windows in so many ways.

    Linux has a long history of running on x86 on non x86 processes. Wine on powerpc for OS X is an example of this. So yes you point is valid why is a person going to buy Windows on Arm when Linux on Arm will run more of there old Windows applications.

    You have a very good sales pitch for Linux on Arm. Arm processors are cheaper than x86.

    Phenom
    “Pogson, you disappoint me. Running a VM on a single ARM chip, even dual-core? That would be a sad, unfortunate experiment.”
    On a big little system arm its not bad. This is a quad core system arm. But two of the cores are older versions of arm. Dec-ate those two older design cores to instruction conversion shock horror you can get inside 90 percent perform of a x86 duel core the same speed. Yes a quad core big little arm takes less silicon than a duel core x86 chip. So cost of production a quad core big little is cheaper than a duel core x86.

    Panda board that is a 100 dollar board contains a big little chip.

  24. Clarence Moon says:

    You must not be running on ARM…

    There are no Windows legacy apps (yet) that would run on ARM. There are some apps to be available, I understand, such as MS Office, that will work on ARM processors in some way akin to how they work on x86, but you would have to buy them separately.

    The compatibility issue is mostly about a consumer being able to re-use existing, previously acquired, applications on the new OS. If that were not possible, due to some incompatibility with APIs or such, the consumer would be put off and unwilling to buy the new machine with the new OS or to update an old machine to the new OS. Once the consumers find out that an ARM PC will not run anything that they currently own, it is likely that the ARM PC will go back to the store for a refund or else will never be purchased to begin with if the consumer finds out about the incompatibility before the sale.

    That is what happened with the Linux based netbooks early on. Who wants an odd ball like that?

  25. Phenom says:

    Pogson, you disappoint me. Running a VM on a single ARM chip, even dual-core? That would be a sad, unfortunate experiment.

  26. Ah! You must not be running on ARM…

  27. oldman says:

    “How are your legacy apps on “8″ and have you joined your domain?”

    So far every application that worked on windows 7 has been working just fine. This includes Office 2010, Firefox, THunderbird, PowerGUI, PowerCLI extensions for Powershell from VMWare. Installing these applications has allowed me to run several custom powershell scripts without issue.

    As far as joining our campus domain is concerned, I haven’t done this yet, but I would be surprised if I had problems.

  28. oldman wrote, “As far as windows 8 is concerned I am running it on my desktop in a VM right now and it runs quite well thank you.”

    How are your legacy apps on “8” and have you joined your domain?

  29. oiaohm says:

    Clarence Moon
    “you are not actually doing anything yourself that is at all effective in making that happen”

    This is also false. I am working on a few things.

    Like looking into one of the worse bugs of POSIX.
    http://austingroupbugs.net/view.php?id=545

    Also do op in winehq on freenode to keep trouble makers out.

    There are a few other projects were I am doing things.

    Clarence Moon what are you really doing?????

  30. oiaohm says:

    Clarence Moon
    “That is why you are not running a big, successful business like Steve Ballmer.”
    Ok what has happened under Steve Ballmer rain.

    Loss of the foot hold in Mobile phones Microsoft had. Reducing in market share in all segments.

    Sorry Clarence Moon Steve Ballmer is no Bill Gates. Bill Gates Microsoft grew. What they call Microsoft 2.0 has been in decline the complete time hidden by the fact new markets have been opened up. This will stop.

    We are not looking at a 50 year window here it is way shorter Clarence Moon.

    Clarence Moon
    “Why should you even care?”

    This is simple working doing support you need to know what is coming. So you don’t get caught on the back foot.

  31. oldman says:

    “Well, well, oldman. I was beginning to think you had died. I haven’t seen many comments from you in a while…”

    Nope, I am still here, just very busy. Being part of the team that supports a mixed windows and linux environment of 700+ servers and counting is a full time job.

    did you miss me ;-)?

    “Many recent migrations show that there is not too much trouble changing 80% of the seats to GNU/Linux. ”

    Actually if you look very closely at most of those conversions there was a non technical factor that facilitated the conversion. Munich was kept going by the idealogical determination of its socialist government. In most places where , there will be no conversion, as much because of business issues as the reality of inertia.

    Remember Pog, “Culture eats strategy for breakfast”

    “GNU/Linux is ready now. “8″ is not ready and won’t be for years.”

    The commercial OS named android is running now Pog, the classical linux desktop is just as dead as ever. As far as windows 8 is concerned I am running it on my desktop in a VM right now and it runs quite well thank you.

  32. Well, well, oldman. I was beginning to think you had died. I haven’t seen many comments from you in a while…

    oldman wrote, “I doubt that the shrinkage will be that much.”

    Many recent migrations show that there is not too much trouble changing 80% of the seats to GNU/Linux. That could be a lot of shrinkage. Businesses need to calculate and compare the costs of endless steps on the Wintel treadmill to a changeover. The numbers look good for GNU/Linux.

    The x86/amd64 platform will be preserved by Moore’s Law. At the same time ARM is already where many people want x86 to be in a few years as far as size, cost, power-consumption, etc. GNU/Linux is ready now. “8” is not ready and won’t be for years. I don’t think people are going to wait around until XP expires before going to “8”. They will go to GNU/Linux and Android/Linux. The legacy apps can mostly be moved to the server and used from thin clients.

    NetApplications’ numbers lie about current levels but they cannot hide the fact that M$’s share of IT is shrinking at least on x86 client machines.

  33. Clarence Moon says:

    I can see no action MS can do to prevent this.

    There are a lot of other things that you do not see as well, Mr. Oiaohm. That is why you are not running a big, successful business like Steve Ballmer.

    After some length of time, you will probably prove to be correct, but it may take 50 or a hundred years. It is much more useful to have a good idea of what is going to happen next year and be able to capitalize on the knowledge.

    You seem to be dedicated to wisahing and hoping for Microsoft to fail, Mr. Oiaohm, but you are not actually doing anything yourself that is at all effective in making that happen and you do not seem to have any benefit coming your way if and when it does happen. Why should you even care?

  34. oiaohm says:

    Clarence Moon
    “For another, it seems to me that, even as things change to become more and more functional as well as more and more compact, the names of the big players are still pretty much the same: Microsoft, Apple, IBM, Sony, Nokia, along with newer prominence for Google, Motorola, LG, HTC, Samsung and others.”

    Ok who are the sharks I am referring to circling Microsoft. Even Nokia is working on Linux.

    Microsoft is not a hardware player. The sharks are the real hardware makers. Some of these are no longer putting the effort into supporting MS OS systems. Windows CE and Windows Phone 7 both released on ARM hardware on time. Windows 8 is going to be late. Hmm wonder if this is a case that device makers are not supporting MS on arm. It is.

    I am not expecting payback against the rats Clarence Moon. Remember the old saying rats are first to leave a sinking ship. The rats have moved most of there shares out of Microsoft.

    Clarence Moon
    “All that you are saying is that technology marches on and some companies may fall by the wayside.”

    This is not what I am saying. I am saying as technology marches on MS profit lines are going to take a pounding. I can see no action MS can do to prevent this. Going into the cloud and other options MS has tried so far to make new income streams is not working.

    Microsoft is in the same place lot of other makers have found themselves in. IBM vs the Clones comes to mind for where Microsoft is now.

    Client division of Microsoft is going to collapse the foundations are cracked. The question is when will it fall and how far. When the Client division collapses so does the Business division.

    Microsoft biggest incomes are at risk.

  35. oldman says:

    “Clarence, you don’t get to define what a personal computer is. ”

    But then again, neither do you Pog. You can speculate on what you hope it will be, but in the end what counts is what sells. Desktops havent been full towers in years – many of the desktops at my shop are small form factor. In fact one does not even need a desktop – a portable will do quit nicely, especially when equipped with USB 3.0 and/or firewire 800 ports.

    As far as I concerned, Mr. Moon does indeed get it mostly right. My take is that there may indeed be some shrinkage in the full function computer market as that part of it that didn’t really need a full function computer succumbs to the lure of tablets. But I doubt that the shrinkage will be that much. Anyone else who still needs to accomplish tasks that require more than what the tablet form factor can accommodate will still need full function computer – either a desktop or a portable. And whether you like it or not it will remain powered by windows.

  36. The client division is shaky but the others are not. I expect the server division will be next to fall.

  37. Clarence Moon says:

    You call it “disasterous”, Mr. Pogson, I call it just another day at the office. I don’t own any of their stock, but it is up some 30% since the quarter that you reference, so the people who pay their money and stick their necks out on their beliefs about Microsoft are not in your camp. They are making a lot more money overall, too, so it would seem that they have a successful plan for life after Windows, eh?

    All I am saying is that Microsoft continues to be more than a bit player and they are coping with changes to the market. I do not see why that will not continue ad infinitum.

  38. kozmcrae says:

    Previous comment down the spam trap.

  39. kozmcrae says:

    Clarence and his Ego said:

    “By “GNU/Linux” you presumably mean something along the lines of a conventional PC that might run Ubuntu?”

    In reference to Robert’s statement:

    “That started last year with smart thingies appearing in droves on retail shelves.”

    Clarence’s brain, just like his beloved OS, reboots every morning. He still thinks the desktop is King and everything is subordinate to it. He didn’t want to hear that GNU/Linux will be dominate on the retail shelves so he didn’t hear that. He heard just what he wanted to hear — That GNU/Linux will never replace Microsoft on the desktop and Robert was referring to the desktop, which he wasn’t.

    It’s a waste of time and effort exchanging views with Clarence. When you mention ‘X’ he thinks ‘Y’. His reality is skewed. I’m not kidding or making an insult. He’s not on the same planet as the rest of us. Humor him but don’t bother to put much time in to it. He is not worth the effort.

  40. Clarence, you don’t get to define what a personal computer is. Conventional PCs are obsolete. ATX cases are declining in shipments. They were wonderful in their time but we no longer need hard drive, floppies and CD drives on every PC. We don’t need 600W PSUs in most PCs. We don’t need 100W CPUs etc. What’s left runs GNU or Android/Linux just fine and people buy them if offered.

    M$’s client division was down 2% from 2010 in 2011. They had a disastrous Q4 of 2011 down $300 million in revenue from Q4 of 2010, ($4709-$4988)/$4988 X 100 = -5.6%. Operating income was down 10%. At the same time, sales of PCs of the x86 kind were not down at all and counting all personal computers, up a huge amount. That’s not growth, Clarence and, until “8” appears, M$ has nothing going forward, except more of the same. The hard drive shortage is kicking in seriously this quarter, too.

  41. Clarence Moon says:

    What is changing is that M$ is changing from the elephant in the room to just another player.

    I don’t agree with that, Mr. Pogson. I think Microsoft is just as big as ever and it is growing every year. What may be happening, if your theories on your cobbled up statistics are at all accurate, is that the room is getting much larger and the Microsoft elephant is taking up less room by comparison. There are more elephants coming into the room as well, for example Google, Apple, Samsung, and Amazon, so your reference to “just another player” has some merit although it is still a rather exclusive sort of game.

    You are just as strident as Mr. Oiaohm, I think, in caring so much which of the elephants now gets to eat which peanut. Unless we have our personal fortunes tied to any one of these elephants, which I for one do not have, it really does not matter much what happens under the covers as long as what we ourselves have access to enjoying continues to improve as it has for 30 years.

    This year GNU/Linux thingies on ARM and x86 will appear on retail shelves

    By “GNU/Linux” you presumably mean something along the lines of a conventional PC that might run Ubuntu? It would have to be something that competed directly with sales of a more common Wintel netbook/notebook to qualify. Also, it should be a real shelf, as found in the local Best Buy or Target store, not some obscure web site of some nameless Asian company. With those qualifications, when do you see this event becoming commonplace?

  42. Clarence Moon wrote, the names of the big players are still pretty much the same: Microsoft, ….

    What is changing is that M$ is changing from the elephant in the room to just another player. “Big” can change pretty fast when your are leveraged as they are. M$ puts almost nothing into the food-chain except permission to copy and yet has huge revenues. That is not sustainable much longer. Their revenue per PC will drop precipitously as soon as there is competition on retail shelves. That started last year with smart thingies appearing in droves on retail shelves. This year GNU/Linux thingies on ARM and x86 will appear on retail shelves. M$ no longer has the power to prevent that. In the old days, M$ could twist the arm of half a dozen big players and make things happen. The world of IT has become too diverse for them to do that. As my father used to say about the Germans in WWII, “They didn’t have enough bullets to kill everyone…”, M$ does not have enough salesmen to stifle the world any longer. China has thousands of small OEMs who will ship direct to any retailer and M$ doesn’t even know their names. The big players see that competition and do have to offer choice. China is large enough to be its own market and doesn’t need M$ for anything.

    All those small players add up to the biggest of big players. According to IDC, “other” now ships 39% of PCs.

  43. Clarence Moon says:

    the blood is in the water

    I am sure that you have no idea how downright silly you look when making such comments, Mr. Oiaohm! All that you are saying is that technology marches on and some companies may fall by the wayside. It is entirely possible that one of them may be Microsoft, too, but that hardly rises to the sort of personification that you are making.

    I am elated at the prospect of new and intriguing things, myself, whether they come from Microsoft or some other company. I have a Kindle Fire and a Nook that I have used a lot for reading on vacations and almost as much for Angry Birds. I am dusting off my engineering education and looking into writing apps for the Kindle, too. It will give me something to do after I get used to being retired.

    If you are looking for payback, though, as you seem to expect to get from watching the “sharks” surround a drowning Microsoft, I think you will be disappointed. For one thing, the villains involved, such as Gates and Ballmer, are totally insulated from any financial ruin and have long ago moved the bulk of their fortunes to safer havens than MSFT shares.

    Q. “If you won the lottery, would you quit your job?”
    A. “No. I would have to go into the office to quit.”

    For another, it seems to me that, even as things change to become more and more functional as well as more and more compact, the names of the big players are still pretty much the same: Microsoft, Apple, IBM, Sony, Nokia, along with newer prominence for Google, Motorola, LG, HTC, Samsung and others.

  44. oiaohm wrote, Question is how much will cross the good enough line so displacing PC units with small arm computers in the form of phones or small boxs?

    Don’t forget. When “the box” is this small, it can become part of the circuit-board inside a monitor, keyboard, or even a mouse. Soon, it might be in a dongle/cable/connector. Moore’s Law works. At the moment, we are at 28-40nm. 10-12 nm is within reach. Chips could become a fly-speck. Connectors could become the limiting factor on size. Maybe everything will become some tiny connector, like mini-S/PDIF.

  45. oiaohm says:

    Clarence Moon factor you are missing is android kernel is merging main-line.

    So platforms that support android also will support general Linux systems.

    Next factor is raspberry pi and rhombus tech.

    Raspberry Pi first 10 thousand units sold out in seconds. There is a good chance the next units will go equally as fast except with a 100+ thousand volume of the next run.

    This is setting a new base point in the market the sub 50 dollar system.

    You could not see how a phone could work on a office desk its about time you take your blinkers off Clarence Moon. The Raspberry PI is weak. Tech following it will be stronger. Question is how much will cross the good enough line so displacing PC units with small arm computers in the form of phones or small boxs?

    Raspberry pi and Rhombus Tech are both going to be releasing the designs of there boards.

    MS up until now has been able to fight multiplier tech like SoftXpand and others. Problem is once MS accept it in business this will reduce MS income with no possibility of recovering it.

    Ticking time bomb right. Something a person like you would not get Clarence Moon is that mobile phones running Android use the same kind of software as what SoftXpand is to support hdmi screen connected by USB port. Merge of Android kernel and Linux Mainline kernel + what is going on with Wayland will equal SoftXpand feature in all Linux distrobutions. So time is limited for how long MS can keep on not providing this feature. This feature will rip Microsoft revenue to shreds. Expected is somewhere between 1/2 to 1/8 of there current revenue on desktop from the change. This allows for home users not using it.

    Clarence Moon the blood is in the water and the sharks are circling Microsoft. So far the sharks have not started taking big bites and are still just nibbling.

    If each of the time bombs go off lower entry level PC price at sub 50 dollars, desktop multiplexer software have to be provided by default and more competition from android and linux desktop distributions due to merge. Result is not small little bang.

  46. Kozmcrae says:

    Clarence and his Ego said:

    “Anyone who might be interested in saving money in that regard could more easily just time share a conventional PC.”

    But not anyone interested in saving time.

  47. oiaohm says:

    Clarence Moon as soon as someone say Units and IDC numbers they are bogus. So that claim is invalid since IDC does not provide unit numbers only revenue.

    W3Techs and Security Space are the only two that currently are doing Unit numbers and they have the numbers basically reversed. ~60 percent Linux vs ~40 percent windows.

    Windows is so high in revenue due to how much windows licenses is. Hardware makers in fact make more per unit selling the Linux Servers.

    IDC yes is gross revenue not profit to hardware makers.

  48. oiaohm says:

    Clarence Moon IDC is revenue not units. So the number is very grey.

  49. Clarence Moon says:

    I am sad to see that you have to go to such lengths, Mr. Pogson. I cannot imagine that there are people who would go to such ends to simply spoil your site. You provide a lively forum with much more opportunity for comprehensive argument than any other Linux vs Windows site I have seen.

    That said, I still do not see much merit in this case. As you note, the end customer has a number of options beyond MiniFrame or Microsoft.

    I had given up on reading the lengthy complaint and had not reached the section you cited, so you have a point. Even so, I would argue that it is all anecdotal and turns on the monopoly power issue in part. It would be interesting to see this issue re-visited in a Federal court now that Apple is supplying x86 software and Linux is more openly seen as a viable alternative to Windows. I would imagine the rise of ARM and Android would be significant factors in any decision.

    As a sales guy, I know how customers give all sorts of reasons for not buying your product. It is human nature to want to have a neutral excuse for not following through on the sorts of promises made to one giving a sales pitch. The buyer is always trying to seem enthusiastic even while giving the same sort of attention to your competitor. In the end, the loser(s) have to be sent off, but not in anger since it may be necessary to renew negotiations in the future. Perhaps the JP Morgan story has merit, but I doubt that the whole truth is being told here.

    Microsoft’s reply does not seem to be based on this issue, though, and they cite a bushel basket full of prior decisions that support their right to do what they think proper in line with promoting their wares. Do you dispute any of these?

  50. Read the complaint. M$ told several business customers and OEMs not to use MiniFrame’s product.

    148. On or about July 13, 2011, JP Morgan Chase (“JP Morgan”) contacted MiniFrame and inquired about MiniFrame’s SoftXpand line of products. JP Morgan had tested SoftXpand and was interested in implementing SoftXpand in approximately 80,000 JP Morgan Chase computers that were running Windows XP.

    152. JP Morgan Chase also noted that they had already contacted Microsoft regarding the proposed arrangement, and that Microsoft had approved it. This may have occurred, in part, because JP Morgan Chase already had a site license from Microsoft for its use of Windows that was based on a “per number of users” cost.
    153. On or about July 19, 2011, possibly the very next day, Microsoft contacted. JP Morgan Chase and revoked its previous approval of this arrangement. Microsoft’s reasoning for the revocation was based on the Single User Restriction language in the EULA.

    So, this was not a case of M$ being cut out of any revenue at all but they wanted to exclude the competition. JP Morgan was going to have 160K users and only wanted 80K PCs with 160K monitor/keyboards/mice.

    151. JP Morgan Chase also informed MiniFrame during the July 18, 2011 communications that they estimated the total cost of purchasing 80,000 new touch screen computers running Windows XP, plus the cost of then migrating 160,000 to Windows 7 was roughly $100 million per year over a three year period. The use of SoftXpand would save JP Morgan Chase at least 50% of those costs.

  51. Clarence Moon says:

    It seems to me that this case has no merit at all, Mr. Pogson, unless and until Microsoft takes some sort of action to bar MiniFrame from selling their product. The only complaint stated by MiniFrame in the filing seems to be that they were trying to do a deal with HP and HP opted for a Microsoft product instead.

    All that exists at the moment is an allegation that the “single-user at a time” clause in the EULA would apply to MiniFrame’s product. Since Microsoft is not asserting this clause in any action, it seems doubtful that it is causing any problem. More likely, the clumsy configuration of a single PC with multiple, cabled monitors and keyboards would be much more likely to be the cause of MiniFrame’s inability to market. Anyone who might be interested in saving money in that regard could more easily just time share a conventional PC.

    One interesting claim that you might be interested in discussing was:

    Microsoft has a monopoly in the server operating system market.

    According to the International Data Corp (“IDC”), in the 4 th quarter of 2009, Microsoft’s
    Windows Server Operating Systems was the No. 1 server operating system in the world, both in
    terms of units and revenue. In terms of units, Microsoft’s Windows Server Operating Systems
    accounted for 73.9% of all server operating system sales, followed by Linux at 21.2% and Unix
    at 4.3%.

Leave a Reply