Robert Pogson

One man, closing all the windows.

Microsoft Squeaks. No One Listens.

  • May 31 / 2011
  • 79
technology

Microsoft Squeaks. No One Listens.

M$ is apparently trying to dictate to the world on what kinds of ARMed systems M$’s “8″ will run. Apparently QualComm is OK with this but many others are not.

Qualcomm:“while Yen walked us through his company’s first SoC developed exclusively for Windows 8 use in desktops and laptops, he explained that Qualcomm pushes the boundaries still further. While based on ARM’s high-end Cortex-A15 ‘Eagle’ design, he claimed that it had been customised to the point that the label no longer applied.

“It’s an A15-class processor,” Yen agreed, “but it’s not a Cortex-A15.” He revealed that the company’s Windows 8 offering will take the form of the MSM8960, a Krait processor featuring two processing cores running at 1.7GHz and an integrated LTE modem. Sadly, it’s a far cry from the quad-core 2.5GHz Krait the company announced at the Mobile World Congress event earlier this year.”

Acer’s Chairman and Chief Executive Officer J.T. Wang said.
“They’re really controlling the whole thing, the whole process,” Wang said at the Computex trade show in Taipei without identifying the restrictions. Chip suppliers and PC makers “all feel it’s very troublesome,” he said.

By trying to freeze the market and dictate hardware compatibility, M$ will only delay its roll-out and restrict itself to a niche. ARM’s ecosystem has already grown to a multicultural environment with many makers of chips and products. M$ cannot force all manufacturers to use Qualcomm’s chips and Qualcomm cannot supply the whole market. It’s too damned big. Delay has happened for months in IT as suppliers try to catch up with demand and Google delayed releases of Android. No one wants to wait for M$ to get its act together while hundreds of millions of units could be sold.

Watch Earth catch fire as 100 million activations of Android happen in short order. By the time “8″ happens there could be hundreds of millions more activations. Even Intel is pushing Android on x86…

Here’s an early product showing Android running on Intel. Who needs M$?

Here’s Intel showing an x86 chip they intend to put in smart phones and tablets running Android, MeeGo etc., anything but “8″.

Thanks to Richard Chapman for the link to Dailytech.

UPDATE
More on this. It has been revealed that M$ has arranged to consult with a few makers of PCs on “8″ but has not included anyone from Taiwan…
“Microsoft has not invited Taiwan-based IT vendors to participate in its Integrated Development Program specifically for developing Windows 8 for use in tablet PCs.

Microsoft has invited chip vendors, including Intel, AMD, Nvidia, Texas Instruments and Qualcomm, as well as a few PC vendors to participate in the program. However, Taiwan-based vendors Acer, Asustek Computer and HTC are locked out, arousing considerable controversy in Taiwan. Microsoft may hope to set rules on its own and the invitation to chip vendors is to save possible trouble, Wang said. But from the point of view of Taiwan-based enterprises, Microsoft is not acting reasonably, Wang stressed.”

see Microsoft should set rules from viewpoint of ecosystem, says Acer chairman

So much for being “partners” of M$… The world should just let M$ sit in the corner.

79 Comments

  1. twitter

    The net loss to society from the Windows monopoly has to be measured in the true difference between the free and non free software world, not just the software costs. The product of monopoly rents, reinventing wheels, and the net losses from intentional waste that non free software always applies are much greater than the sum of each. All of these costs are multiplied by monopoly and artificial scarcity costs which extend to judicial extortion of companies using free software in a way to increase costs, eliminate useful features and remove competitors from the market.

    Non free developers are unable to share or even trust each other, so they are always forced to reinvent wheels. This is why Windows programs bring their own, redundant and inferior, set of tools with them instead of relying on Windows’ even worse internals. That waste is multiplied by each forced “upgrade” that break legacy work.

    Hardware churn is another additional cost. Microsoft is careful to eliminate older equipment from new software offerings in a way that is supposed to promote hardware “refreshes.”

    To get the total cost of Microsoft, we have to subtract all of this waste from the costs people would naturally incur in a free world. The ill gotten gains of Microsoft’s executives is only a tiny fraction of the true cost of Microsoft’s crimes. People would have gotten the work done without Windows and they would have gotten it done with a lot less effort and trouble keeping things working. Anyone who’s used both free and non free software long enough knows this but people who’ve been free long enough tend to forget it. Non free software heartache fades like a bad dream.

  2. Robert Pogson

    That work could have been done on GNU/Linux had the apps been available. The value of what the OS does is the same as that other OS but it costs much less.

  3. oldman

    “I can measure the evil M$ continues to do by subtracting the cost of malware from the gross income of M$ and come to a negative value. ”

    And I can demonstrate the amounts of work that I and others have gotten done over the years using applications built on microsoft OS’s and with microsoft tools.

    And THAT is all that counts.

  4. Robert Pogson

    I can read about the evil M$ has done on US DOJ v M$. I can measure the evil M$ continues to do by subtracting the cost of malware from the gross income of M$ and come to a negative value. It is not worth installing that other OS. $20 worth of GNU/Linux will do a better job and cost nothing for malware.

  5. Linux Apostate

    Well, thankyou at least for saying that. Many people would not be so honest as to admit that sort of bias.

    Criticising Linux may not be cool, but it may nevertheless be the right thing to do. I see Torvalds often criticises Linux, this seems to be a key part of the process by which it improves. The very idea of discouraging criticism strikes me as oddly totalitarian.

    Equally, defending Microsoft may not be cool, but it may still be the right thing to do when the facts are in Microsoft’s favour, as they may be. Microsoft is not all bad.

    Twitter has very obviously projected his own fanboyism onto me. I am no MS fan. I use Linux! At home and work. I even contributed software to it. I only defend what I think is the truth.

  6. Robert Pogson

    Stop defending M$ and that should cease to be a problem…

    It’s OK to describe reality as each sees it in IT, but to criticize GNU/Linux which is doing quite well with a tiny budget is not cool. It is almost irrelevant that GNU/Linux is not perfect compared to the willful damage to IT, markets and competitors done by M$. Defending M$ appears to be some attempt to preserve the monopoly. M$ has enough funds. They can defend their monopoly without the help of fans on the web.

  7. Linux Apostate

    Your relationship with the “search for truth” is obvious from your posting. Whenever I write anything, you see “lies” and “spin”. In fact it is clearly pointless to write anything at all. You are entirely focused on misinterpreting what I write as if it were a defence of Microsoft.

  8. twitter

    Microsoft’s lies are offensive and that is why I bother to comment. It is interesting that people like you think of a search for truth as an attack on Microsoft.

    Please don’t confuse yourself about me not needing to repeat myself. You have not successfully demonstrated that gnu/linux has a problem with ARM or that Microsoft’s attempt to dominate OEMs will be positive for free software. References presented above largely refute the points, as does the decade old gnu/linux ARM community and the obviously flowering and profitable market for Android. It is impossible to spin Microsoft’s long history of ARM failure into some kind of success story or Microsoft’s vaporware into anything concrete. If not taking these lies seriously is an attack on Microsoft, the whole world is attacking Microsoft.

  9. Joe

    “I don’t think Microsoft is dying, by the way. They’re like IBM now, an industry fixture.”

    I don’t think they will go bankrupt any time soon, but they went from being the #1 technology company by market capitalization to #3. They no longer have the pull over the industry that they once had.

    “I don’t think Microsoft is dying, by the way. They’re like IBM now, an industry fixture. They can’t really be killed without killing the rest of the industry.”

    Actually IBM recently surpassed Microsoft in market capitalization. Anyway if Microsoft goes bankrupt it could possibly be the best thing to happen to the computer industry, who had to deal with monopolistic business practices like the OP describes and broken proprietary software for 20 years.

    The only because who would get hurt by Microsoft’s demise would be the Microsoft developers and related ecosystem, but Microsoft is doing a good job of screwing with them all on their own.

  10. Robert Pogson

    Purchase of PCs with that other OS are stagnant while smart phones are booming. There are still about the same number of PCs with that other OS sold per annum as smart phones but the smart phones are growing rapidly (~90%). It’s all over. Get used to it.

    The IT industry does not have to be killed to get rid of M$. The IT industry can change direction much faster than M$. IT will simply leave M$ behind, perhaps rattling around on business desktops. The world wants smaller, cheaper, and faster computers and M$ offers none of those things.

  11. Linux Apostate

    You’ve given up arguing, so I’ll consider this a success. But I’ll remember to quote someone senior on LKML next time I want to argue with FLOSSies. Just saying things that are true is clearly not enough.

    But here’s another question. If you really believe that “Microsoft is lost”, why do you bother to attack it?

    I could ask the same question of Pogson, who is always telling us that the “other OS” is dying and nobody is buying PCs now they can get Android phones. And yet he’s always attacking Microsoft all the same.

    I don’t think Microsoft is dying, by the way. They’re like IBM now, an industry fixture. They can’t really be killed without killing the rest of the industry. In 20 years we will still have plenty of computers with Intel CPUs and Microsoft software, not to mention the descendants of the XBox.

  12. twitter

    The point is moot because Pogson was right about no one listening to Microsoft. It’s game over for those idiots, so Linus and friends will be free to implement rational, free solutions and create real standards. My advice for you is to short MSFT, which has already slid down to $23 but has plenty of downward potential with Windows 7 failing and Windows 8 being a pipe dream, two years away and completely without OEM and vendor support. Microsoft is lost in the past and will fade like a bad memory.

  13. Linux Apostate

    You say: “That’s a simple solution that can be solved in software freedom.”

    I respond: Actually there is plenty of dissent over whether that idea is a good one. For example, Nicholas Pitre: “…this does create pain. you have to make things in sync between the kernel and the mini-kernel (let’s call it bootloader). In practice the bootloader is always maintained separately from the kernel, on its own pace and with its own release schedule. Trying to synchronize independent projects is really painful as you know already, otherwise the user space for perf would still be maintained separately from the kernel, right?”

    I add: Ironic that you should complain about ACPI, given that you’re now essentially supporting the same idea.

    You say: “They are standardizing the icons bezel sizes and other “unimportant” stuff and forcing OEMs to deal only with one chip maker.”

    I respond: And they’re standardising important things as well. Or do you know otherwise? Citations please.

    You say: “The statement… is not a technical argument, it is an assertion of faith that is not supported by historical behavior.”

    I say: You are again wrong, because you just *assume* it’s about faith. It’s not; I only believe that Microsoft wants to make money. It’s all about the huge difficulty of supporting multiple SoCs with completely different BSPs. The kernel hackers know all about that – it is why Torvalds is so unhappy. This *is* a technical argument. In fact it’s so technical that you didn’t even believe it (or understand it?) until I quoted Torvalds confirming what I said.

    The practical, easy solution to this huge problem is to force the SoC manufacturers to standardise… but Torvalds can’t do that, and ARM can’t do that, and the SoC makers don’t want to because they want to keep their monopolies and their lock-in.

    So who can?

    I think we all know the answer to that. When the Linux engineers complain about the diversity of ARM platforms, complaints and workarounds are the limit of their capability. When the Windows engineers complain about the same things, they talk to Steve Ballmer and tell him something must be done.

    Because only one company can say “Make these changes, or no Windows for you.”

    Some things require an empire.

    But then, what have the Romans ever done for us?

  14. twitter

    Thank you for pointing to the kernel list, which shows what problems exist and how they are being dealt with in freedom. The problem is that ARM people are new to software freedom and are relatively inexperienced coders. The solution, according to Linus is to,

    somebody really REALLY needs to think
    about how to get those crazy board details out of the kernel entirely. Having per-board drivers for real hardware is sane – having to have per-board detail files for clock chips is just crazy. Split off that thing a “Linux ARM second-stage bootloader” project that has the per-board tables or something. Don’t pollute the main kernel with crazy details like this.

    That’s a simple solution that can be solved in software freedom.

    This has nothing to do with what Microsoft is doing, despite many typically misleading and dishonest assertions. The statement,

    I don’t doubt that Microsoft wants to standardise the ARM platforms in order to benefit Windows 8. My contention is that this benefits all other OSs as well… as a necessary side-effect.

    is not a technical argument, it is an assertion of faith that is not supported by historical behavior. Microsoft “standards” have always made life harder for free software. I’ve named ACPI, Winmodems, USB and more only to be ignored or told nothing was wrong. Hmph.

    Thankfully, some semi-technical arguments are finally made.

    It’s about whether or not a particular kernel binary can boot on multiple systems, a luxury that you enjoy on every PC, but not on ARM devices.

    This is not accurate, but besides the point.

    OEM changes like icons, graphics and preinstalled apps don’t matter at all. The differences that *do* matter are all within the SoC – it is here that the architecture can be different and incompatible. This doesn’t benefit the user at all – it just helps the SoC maker, because it makes it harder for the OEM to adopt a different SoC. This is one reason why the SoC makers hate standards.

    This completely contradicts what Microsoft is doing. They are standardizing the icons bezel sizes and other “unimportant” stuff and forcing OEMs to deal only with one chip maker. This assures that SOC makers will be free to keep making things difficult for gnu/linux and other universal software which has so far done a better job of providing cross platform application uniformity. Users do care that they get the same applications on x86, ARM, MIPS, PowerPC, etc and that’s exactly what free software delivers right now, while Microsoft has only managed to provide functionally deficient platforms such as WinCE with next to no user level applications.

    The reaction is predictable. OEMs, who love standardization, are outraged by this collusion by Microsoft and SOC makers. Not surprisingly, they have revolted and Windows 8 has been pushed back by more than a year.

  15. Linux Apostate

    One of the things I have been trying to communicate to you is that x86 hardware is diverse in a different sense. There, the diversity can be dealt with using device drivers because it is at a higher level. The basic stuff is always standard. You may have differen PCI devices but the PCI interface is always at the same physical address.

    But on ARM the diversity is at a lower level. The basic stuff is *not* standard. Device drivers are *NOT* the solution.

    Read further through the thread. Torvalds and the others are discussing these very issues and how to deal with them. It’s better that you go there, rather than listening to me, because Torvalds speaks with an authority that I just don’t have (even if I use my real name). The problem is not to do with newcomers to the kernel, though a better kernel design might be able to handle the problem better.

    Here’s a particularly good bit:

    Torvalds: “The long-term situation should be that you should be able to have ONE binary kernel “just work”. That’s where we are on x86. Really.”

    Nicholas Pitre: “But X86 is peanuts. Really. There was one machine called the IBM PC at some point that everybody cloned, and the rest was totally irrelevant. Then came that thing called Windows that reinforced this hardware
    monoculture as it was used for the ultimate conformance testing. This is damn easy in that case to produce a kernel that works virtually everywhere. On ARM there is simply not such thing as a single machine design to clone, and a closed source test bench to design for.”

  16. Robert Pogson

    The user does not care what is on the SoC except the performance he gets which is a choice of chip that the OEM makes. The OEM gets all the help he needs from ARM and the maker of the chip. This is all a red herring. The applilcations and the user interface are determined by the app writers and the OEM and those can be the same no matter what the chip is. There are a bunch of layers of abstraction between those design elements and the hardware.
    “ARM right now i a nightmare, and most of it is because ARM hardware manufacturers are morons. But the way the ARM tree is then laid out has made that even more painful, and the decision to put all the crazy board details in the kernel tables instead of trying to have a per-board boot loader that fills in the details is just crazy.

    That is because the folks working for the OEMs are new to the kernel. That will all be sorted out. It has nothing to do with the diversity of the hardware but how people approach it. Linus could just refuse some requests until folks standardize how they use the kernel. When the OEMs get tired of hacking the kernel for each device they will standardize. It does not require M$ or anyone to dictate. The diversity of x86 hardware is much greater than for ARM and no one seems to think that is a problem.

  17. Linux Apostate

    This did give me a chance to think about *how* I can communicate why certain sorts of diversity aren’t good.

    The first thing to say is that OEM changes like icons, graphics and preinstalled apps don’t matter at all. The differences that *do* matter are all within the SoC – it is here that the architecture can be different and incompatible. This doesn’t benefit the user at all – it just helps the SoC maker, because it makes it harder for the OEM to adopt a different SoC. This is one reason why the SoC makers hate standards.

    The second thing is that it occurred to me that I should really go find someone on LKML saying the same things. This is tricky because all of what I’ve been saying is completely obvious to the kernel hackers and doesn’t need to be stated. But here’s a recent thread that might be of interest:
    http://lkml.org/lkml/2011/3/30/379
    http://lkml.org/lkml/2011/3/30/525

  18. Linux Apostate

    Yes, that confirms what I said. You think it’s a good thing to have a “dozen or so” incompatible, proprietary hardware standards.

  19. Robert Pogson

    Nonsense. There are a dozen or so chip manufacturers for ARMed CPUs for gadgets and they provide some kind of standard interface for their OEMs. At the application level, the apps do not need to know what chip they run on because Dalvik is the standard virtual machine. The place where the OEMs customize is in the intermediate code such as icons that appear by default, backgrounds and such. No OEM is likely to change much even though they could because it is a lot of work. It is advantageous to advertise Android x.y and they cannot do that if they rewrite everything. There are a hundred OEMs using these SoCs and most of them will run what Google gives them.

  20. Linux Apostate

    Twitter: If you’re going to accuse people of trolling, and you are the *first* person to have made such an “ad hominem” accusation in this thread, then you have to expect them to take it personally.

    Yes, I’ve checked. Your very first post here is a personal attack on me. That’s why I took it personally. So much for your claim that “I usually like to stay out of personal stuff.”

    You said: “Try answering some of my arguments and I might consider you something other than a troll.”

    Which seems unlikely, since your definition of “troll” is apparently “someone with a different opinion to me”. However, you then say:

    “The most interesting thing you might tell me is how a company can make something using gnu/linux and not release the source code… There’s a long history of free software users making ARM devices do what they want…”

    This is actually a complete change of subject on your part. My claim is not that Linux *can’t* run on a device, but that a *generic* Linux kernel cannot be produced to run on *all* the devices, unless a common standard permits a single BSP.

    Of course you can “win” any debate if you change the argument to one that favours yourself. But it’s cheating. Although you have “won”, you haven’t really won, and smart people will know that.

    “You might also try to tell me why you think Microsoft “standards” for ARM will be any different from other free software hostile “standards”… Where does your optimisim come from?”

    This point has been dealt with at length, but particularly on Jun 3rd, 2011 at 11:00 am:

    “I don’t doubt that Microsoft wants to standardise the ARM platforms in order to benefit Windows 8. My contention is that this benefits all other OSs as well… as a necessary side-effect.”

    In short, it’s a technical argument. And I have tried to explain it at length.

    Your subsequent posts reveal that you still haven’t really understood what’s being discussed, because you think that the issue is open versus closed source, copyleft versus copyright, but it isn’t. It’s about whether or not a particular kernel binary can boot on multiple systems, a luxury that you enjoy on every PC, but not on ARM devices.

    Others (and myself) have tried to point out that not all Microsoft standards are bad, which is an easy, non-technical concept, and your reply has been essentially “Nuh-uh”. Even though you admit that you have very little recent experience of Microsoft platforms and standards.

    To his credit Pogson didn’t do any of this. He does seem to understand the issue, but he thinks it is better for every SoC to have a different and proprietary architecture, just as long as this makes life difficult for Microsoft (Jun 2nd, 2011 at 7:51 am). If this also makes things difficult for Linux distributions, as it inevitably will, then never mind – the devices come with Android anyway, and that’s sufficiently FLOSSy to be good enough.

    I think this is a bizarre perspective for a Free Software advocate, but it *is* on topic and it *is* a response to the actual discussion. Your (i.e. twitter’s) failure is changing the subject, then insisting that I should defend some other argument that’s completely unrelated to what I was actually saying, and then calling me a troll when I don’t.

  21. Yonah

    Apple was harmed? Cry me a river! Companies compete with each other. It’s business. Some win, some lose. After reading about how you were treated the last time you were in an Apple store (they were too nice to you in my opinion) I’m surprised you’re defending them now. SCSI termination was a real pain in the ass. I had a SCSI zip drive. Yeah, it was faster, but chaining devices together and screwing with ID numbers was gay. Good riddance.

    Reasonably popular references? You’ve spent the last 10 years patrolling the Internet searching for any story you can find reporting any problem or expressing any bad opinion anyone has with Microsoft operating systems. You even link to opinion bloggers like SJVN. He has something bad to say about Microsoft? Shocking!

  22. twitter

    Sigh, more personal stuff. Microsoft boosters and other trolls do that.

    You mean you don’t know? Vista was released in 2007! Does your Windows knowledge really stop at XP?

    I have not been so unfortunate as to have to use, much less fix or administer Vista/Vista7, but I have learned a lot about their failures. My Vista Failure Log and Windows 7 Failure Log are reasonably popular references that have shown up in the first page of Google search results. Fortunately, Vista/Vista 7 was never adopted by businesses so most people don’t have to put up with it.

    Yes, I’m that twitter. You would have known that sooner if your were more interested in learning than you are in defending Microsoft. Most of the links I provide go to my journals where I’ve collected documents of interest.

  23. oldman

    “hat’s not quite the same as “Both ACPI and USB were developed with Microsoft”.”

    But it is close enough to be a distinction without a difference, Pog.

  24. twitter

    Are you trying to tell me that Apple was not harmed by OEMs moving to USB and ATA instead of Firewire and SCSI? Are you trying to tell me that USB was as good as Firewire or that ATA was as good as SCSI? Are you trying to tell me that ACPI was not designed to screw gnu/linux and that it did not blow up in Microsoft’s face? These are all laughable assertions that you indirectly make by insulting me and by saying how happy you were with inferior tech.

    The malicious intent and problems with ACPI power management are well detailed here and links to OSX and Windows failures are included. Included is a letter from Bill Gates telling his people to make ACPI Windows only, some links to trouble that non free software had with the system and a quote from Intel’s people complaining about how poor Microsoft’s implementation was for XP. Vista’s trouble with ACPI is infamous. I have a few youtube videos of Windows 7 power management flunking out in my Windows 7 sucks video collection.

    Firewire and SCSI remained expensive relative to Microsoft’s inferior “standards” throughout their production. If you got USB to work with 98, you enjoyed transfer speeds of 8 mbps while Firewire cranked out 400 mbps. People who bought ATA drives were stuck with a poor subset of SCSI that delivered 5 to 20 MBps transfers while SCSI was churning along at 40 to 80 easily. Google’s studies of hard drive performance showed no difference between SCSI and ATA or SATA reliability, which proves that the only difference between the drives was the controller card used.

    Talking to you guys is like falling through a time warp to the mid 90s before anti-trust cases spilled Microsoft’s malice out in public. After reading how little they respect their users, partners and employees, “anything but Microsoft” is good advice. Better advice is to look for software freedom and to demand anti-trust investigation of Microsoft’s obviously anti-competitive contracts.

  25. Robert Pogson

    Quoting Wikipedia:
    “The USB is a standard for peripheral devices. A group of seven companies began development on it in 1994: Compaq, DEC, IBM, Intel, Microsoft, NEC and Nortel. The goal was to make it fundamentally easier to connect external devices to PCs by replacing the multitude of connectors at the back of PCs”

    That’s not quite the same as “Both ACPI and USB were developed with Microsoft”.

  26. Linux Apostate

    Well said, Yonah.

    Twitter seems to be the Fred Phelps of FLOSS advocacy.

    I don’t doubt that Microsoft wants to standardise the ARM platforms in order to benefit Windows 8. My contention is that this benefits all other OSs as well, not because Microsoft somehow wants to help Linux, but as a necessary side-effect.

    I have attempted several times to explain why this would be a necessary side-effect, but I think I’m wasting my time, so I can’t be bothered to try to explain again in full. Briefly, Microsoft needs one version of Windows 8 to work on all devices, because the alternative is a fragmented platform like Windows CE. As a side effect, that means that one version of Linux must also work on all of them. That is what I have argued. I am right.

  27. oldman

    “It’s a pattern he won’t ever break from. ATA isn’t a Microsoft standard, it was developed by Western Digital.”

    IN other words, Yonah, he is the jerk from slashdot?

    Say it isnt so….!

  28. Yonah

    Both ACPI and USB were developed with Microsoft and including five companies and seven companies for each respectively. But, Microsoft wouldn’t need to do anything at all to be the problem. By simply mentioning the name alone, Twitter decides that technology is obsolete, anti-competitive, illegal, wrong, immoral, you name it. It’s a pattern he won’t ever break from. ATA isn’t a Microsoft standard, it was developed by Western Digital.

    I was using USB with Windows 98. I had a HP flatbed scanner that used USB. No problems. Still using USB today with Windows 7 to load files on my portable AV player, transfer files to my phone, and connect 3 joysticks. Mei wen ti (no problem). DRM scheme? More like DRM schizophrenia!

    I once added an ATA/IDE hard drive to my Amiga 500. Much cheaper and easier than SCSI. Anything a company produces that Twitter doesn’t like only exists to harm competitors. Just like Coca-Cola. Their soda wasn’t meant for people to enjoy. It only exists to harm Pepsi.

  29. Robert Pogson

    My knowledge of that other OS is limited to viewing sad performance on a few machines since XP. Half the world is still using XP according to some popular web stats, so I am probably not alone. It’s great. Half the world has not yet had the “KoolAid”, and still may be saved from M$.

  30. Linux Apostate

    “I presume Vista/Vista7-8 are even worse because of Vista’s paranoid DRM scheme”

    You mean you don’t know? Vista was released in 2007! Does your Windows knowledge really stop at XP?

  31. twitter

    USB was horrible. People should have just used IEEE 1394, aka Firewire. If Microsoft invented USB they did a poor job of implementing it. Windows 95, then 98 were supposed to work with it but did not. Windows XP is still clumsy with USB, pretending to install new drivers every time a mass storage device is shoved in for example, and I presume Vista/Vista7-8 are even worse because of Vista’s paranoid DRM scheme.

    A similar comparison can be made between SCSI and ATA drives, where SCSI was the rational standard and ATA was the cheap and inferior copy designed to harm competitors. Apple happened to be the victim in both cases because Microsoft’s inferior standards were bullied onto the market, leaving Apple as well as Unix machine makers without economies of scale in their chosen and superior tech.

    Free software has mastered almost all of Microsoft’s nasty standards but they remain inferior and their intent was clearly anti-competitive. When Microsoft is gone the hardware market will get better.

  32. Linux Apostate

    Even if ACPI is an example of a bad Microsoft standard, that doesn’t mean all Microsoft standards are bad. What about USB, for instance? Created by a consortium of hardware manufacturers… and Microsoft. And it’s of great benefit to Linux and Mac users.

    It is a shame that this discussion could not have been about a Google effort to standardise Android platforms, since then we could have avoided this paranoia about Microsoft, which has made it very difficult to communicate. Without that we could have concentrated on technical issues and benefits and maybe even learned something.

  33. oldman

    “Actually, what’s funnier is hearing Twitter calling people trolls given his long history of abuse and sock-puppetry on Slashdot. ”

    Interesting reading Yonah. Frankly I had’nt run across this kind of internet vermin. Pog tends to dump the abusive kinds fairly quickly.

    However I can’t quite believe that our twitter is the jerk that is described in the posts.

    perhaps it is someone else…

  34. Yonah

    I’ve never had any problem with ACPI. My Windows 7 machine sleeps and wakes when needed. Why can’t Linux developers get on the ball and make it work? I mean, after all, open source programmers are 10 times more intelligent than their closed source counterparts. Like an army of McGuyvers with keyboards, they are.

    “The people who did not want you to hook up your 300 baud modem still wants to charge for services that people can better provide for themselves and still abuses their position to have things the way they like it.”

    Just out of curiosity, have you read the Unibomber Manifesto? Seems to be right up your alley. Evil, immoral capitalists and their wasteful refrigerators! You can preserve food the old fashioned way by packing it in salt. It’s better and more compatible with a moral society.

  35. twitter

    Google is not the only supplier of Android because Android is free software. It is not copyleft and that causes problems but no one makes people get Android from Google.

    Most copies of Android are jails because Google did not use GPL3 to avoid Tivoisation, but this can mostly be attributed to the malicious demands of Telco providers. The people who did not want you to hook up your 300 baud modem still wants to charge for services that people can better provide for themselves and still abuses their position to have things the way they like it.

    The best way to really avoid fragmentation is to use a copyleft license. This is the BSD and Unix lesson. Again, this is a carrier not a Google issue though Google may be able to fix the problem with TV White spaces.

  36. oldman

    “Having one supplier for the OS is a really bad idea when the supplier wants to “get value” from everyone’s work and give nothing in return.”

    So you are OK with Google being the sole supplier of Android Pog?

    As far as standards are concerned, I don’t care who creates the reference platform. Let google as the owner maintainer of android, take the lead with the vendors.

  37. twitter

    Microsoft is not advancing standards, they are telling OEMs that they can only work with a single chipmaker or they will be punished. Microsoft “standards” have never been real standards, they have always been “extensible” and are more designed to thwart competition than reduce costs by simplification. ACPI is a poster child for this but all the other Microsoft non standards I’ve mentioned above are equally awful. The only way a person can consider Microsoft standards good is if they think whatever Microsoft says is right because they used to have a monopoly on desktop software and some people liked it. Microsoft’s moves will reduce competition, increase costs and have little or nothing to do with well thought out technical standards. It’s something the DOJ should investigate.

    People should have software freedom on all platforms, of course, ARM included. The copyleft license of Linux insure that users will get that freedom because they will have the maker’s modifications. There is no such freedom with Windows or OSX and people should avoid those platforms. If Microsoft were to give people their software freedom, I’d be happy to recommend Windows once the community had done the hard work of dragging it into the 21st century and Microsoft had adequately compensated all of the people they have robbed over the years.

    The news we are discussing shows that Microsoft is determined to maintain their monopoly at any cost and fail before they show users any respect. Microsoft is trying to freeze the market with Vista 8 vapor while bullying hardware makers behind the scenes with these restrictions and judicial extortion. This is obviously aimed at raising the costs of Android and other flavors of Linux and keeping free software along with all other Microsoft competitors away from users.

  38. Linux Apostate

    But you could have that alongside platform standards, which would make it easy to have a choice of operating system once you’d bought the device.

  39. Robert Pogson

    Not happy, but happier. Competition is good and as long as consumers can make a choice of platform, it’s good for everyone. Having one supplier for the OS is a really bad idea when the supplier wants to “get value” from everyone’s work and give nothing in return.

  40. Linux Apostate

    You’d really be happy if the users of ARM laptops, tablets and phones were locked in to the software provided by the OEM, provided that the software was not Microsoft?

  41. Robert Pogson

    “Anything but M$” is a great idea. The world should not depend on a single point of failure and a parasite at that.

  42. Linux Apostate

    See, I think standards are such a good thing that it doesn’t matter where they come from.

    For instance, I think about the PC revolution. It wouldn’t have happened if the cloners hadn’t been able to build hardware to the same standard as the original IBM machines, to run the same software. We might never have had any PCs, and before you tell me that would be a good thing, consider what machines we would have instead. Each one would be a proprietary design with a proprietary OS preinstalled. Getting a machine without that OS would be as difficult as getting a Macbook without MacOS.

    And that’s not all, because there’s a weird inconsistency here. When it comes to PCs you are all for Free Software and allowing the user to choose his OS. But when it comes to smartphones and tablets, users should take what they’re given, and it’s fine if the OEMs make it as difficult as possible to install anything else by refusing to cooperate with a standardisation process.

    Just what is the principle here, anyway? Is this just “anything but Microsoft”?

    I thought I was dealing with a simple lack of understanding about the technical issues. But I suspect that you don’t care about understanding those technical issues at all. You are happy for the OEM to make the choices for you, provided they don’t choose Windows. Are you sure you’re a free software advocate?

  43. twitter

    I usually like to stay out of personal stuff but a guy named “Linux Apostate” being surprised that people “suggest” he’s trolling is funny. Try answering some of my arguments and I might consider you something other than a troll.

    The most interesting thing you might tell me is how a company can make something using gnu/linux and not release the source code which can then be used by the community to make that device work the way they want it to. Tivoization and gpl violations are the only two that I can come up with. Outside of that, distributions should be able to make whatever they want of it, even if it requires a custom kernel as is alleged. There’s a long history of free software users making ARM devices do what they want despite Microsoft having forced the whole PDA market to switch to WinCE and other trash.

    You might also try to tell me why you think Microsoft “standards” for ARM will be any different from other free software hostile “standards” Microsoft has come up with over the years, like ACPI, Winmodems, OOXML, older document formats, mshtml, things associated with Outlook and IIS, Samba2, and so on and so forth. Do you think the “standards” microsoft imposed on netbook makers, such as small screens, large hard drives, CPU and memory limitations, were a good thing for gnu/linux? Really? Where does your optimism come from?

    Microsoft ruined the PDA market 10 years ago and they’d like to to it again. They hit companies with patent lawsuits on pen writing and eliminated graffiti. They also applied retail pressure to run off Sharp and flooded the market with terrible, Windows based PDAs that no one wanted. The effort cost Microsoft and partners a lot of money then and it will cost them a lot of money now. I don’t think they have the money now and I don’t think partners will be dumb enough to do it all over again. Efforts to eliminate netbooks have only partially succeeded and will fail completely as Google starts selling Android netbooks.

  44. Robert Pogson

    FLOSS is about FREE SOFTWARE not hardware from a cookie cutter. The OEMs or ARM can sort things out as needed. M$ is not God.

  45. oldman

    “It’s clear that Twitter doesn’t understand either, and furthermore has no apparent interest in understanding, just in slinging insults ”

    As I have said before, Mr. twitter is just noise.

  46. Linux Apostate

    Actually, I’m quite offended by the suggestion that I am trolling.

    It is unfortunate that I have apparently not succeeded in explaining why it is hard to get Linux working on a range of ARM platforms, and why standards would make this easier. I’d thought I was getting somewhere in the discussion with Mr Pogson but alas he appeared to give up (Jun 1st, 1:45 pm) before any understanding could be reached.

    It’s clear that Twitter doesn’t understand either, and furthermore has no apparent interest in understanding, just in slinging insults while saying barmy things like “Microsoft with ‘standards’ only manages to half way run on x86″.

  47. twitter

    Pogson does not need to know C or assembly to see through the trolling above, which claims black is white and white is black. GNU/Linux runs on ARM and sells by the million despite the alleged lack of a reference platform. GNU/Linux, BSD and other free software seems to do very well with ARM, MIPS, Spark and so on, while Microsoft with “standards” only manages to half way run on x86. Examples of free platforms that came long before Apple and Android are Familiar Linux, Zaurus, Angstrom. There are plenty of others and there will be plenty more in the future as long as everything remains free. The only platform I’m aware of with the fragmentation worried about is Windows, where every computer blighted with it comes with an OEM CD that will work only for that system and the hardware is so nasty that generic Windows installs might fail. It is only free software that is able to navigate devices both new and old.

    The answer, oldman, is always freedom and never restrictions like Microsoft is trying to impose as “standards”. No one familiar with ACPI, Microsoft’s Power Management Poison will wait for Microsoft to make things complicated only to have gnu/linux perform better anyway. As long as chip and board makers cooperate in freedom, they will avoid Microsoft licensing costs and everyone will avoid Microsoft’s restrictions. If the software is free, everyone will be able to get it, modify and share it. Economies of scale do work for standardization and those standards too will be free.

    The part about Tiwanese OEMs being locked out is an interesting, evolving story that should be watched closely. Future Microsoft partners pay especially close attention to what Microsoft is doing to old partners.

  48. Linux Apostate

    Yes, it’s a curious argument for our host to make. He says that PC users should install Debian… but smartphone/tablet users should just accept what they’re given?

    What’s the point of FLOSS if the user can’t customise, can’t install a new kernel, a new distribution?

  49. oldman

    “There isn’t a lot of need for third parties to install Android on devices that shipped with Android and geeks can take care of themselves.”

    The funny thing is Pog, I have probably had more experience that you have writing code in Real languages (C and various (x86, 68K, system z) assemblers not to mention writing real time device drivers for D/A and A/D converters. Yet I wouldn’t presume to make the ad hominem statements about what should and should not be feasible in a given environment that you make.

    Just because you can conceptualize at a high level what may be possible in a situation like this, doesn’t make you right.

    The devil is in the details, and with all due respect, you haven’t shown that you have the relevant experience to comment accurately.

  50. Robert Pogson

    There isn’t a lot of need for third parties to install Android on devices that shipped with Android and geeks can take care of themselves.

  51. Linux Apostate

    How does that help a third-party FLOSS developer to write an installer that will work on more than one sort of device?

  52. Linux Apostate

    But *how* does the BIOS find a storage device? What sort of operation does it carry out? Where does it look?

  53. Robert Pogson

    Just like the BIOS on an IBM-compatible PC does, it enumerates the storage devices. The hardware is so fast that it takes only a few seconds to try out lots of possibilities.

  54. Linux Apostate

    You’ll have to elaborate on that. What do you mean, “hunt for a storage medium” and “checks things out”? What do you think this process would involve?

  55. Robert Pogson

    done:=false;
    hunt for a storage medium;
    n:=0;
    repeat until done
    begin
    examine storage medium[n];
    attempt to run some code from the storage medium that checks things out;
    done := success ();
    inc(n)
    end;
    boot from the storage medium;

  56. oldman

    “It may well be that OEMs will install Linux so the installer for individual geeks will become less important.”

    The only thing that is being installed on these systems so far is the commercial OS Android and when it finally arrives the commercial OS windows 8.

    No classical Linux is being installed, Pog.

  57. Linux Apostate

    It isn’t that easy. You can’t make an installer that will identify the hardware, because you can’t run anything on the hardware without first identifying it.

    Of course none of this matters if users are happy with using the preinstalled software and relying on the OEM for updates, as with WinCE, WP7 and Android.

  58. Robert Pogson

    It’s easy with FLOSS. Take some Linux that runs on the hardware and use it as a base. All the installer has to do is recognize or be told which platform is used. Many distros do that now for x86 and amd64. You either pick an installer or the installer identifies the platform.

    It may well be that OEMs will install Linux so the installer for individual geeks will become less important. 90% of users of hardware are not interested in installing an OS. For an OEM it is no problem to install Linux on their hardware.

  59. Linux Apostate

    So you don’t think it’s a problem if there is no generic kernel?

    Then how would you make an Ubuntu installer that would work for all ARM-based laptops?

  60. Robert Pogson

    “The origin of the phrase is most commonly attributed to Wind River Systems for its VxWorks embedded operating system because of its extensive device driver offerings, but it is in wider use in the industry now.”

    It’s still not a problem. For the embedded folks, it’s business as usual. For the software folks, that’s all hidden by abstraction layers. For the OEMs, geeks take care of it.

  61. Linux Apostate

    But do you see that the problem isn’t drivers?

  62. Robert Pogson

    If you are hoping for M$ to do anything good for ARM, don’t hold your breath. They will miss no opportunity to add M$-only features.

    All the suppliers of ARMed chips provide SDKs which take care of the low-level hardware issue. For each hardware maker the task is the same. Choose a platform, SDK, and customize. M$ is not on the scene, nor even on the horizon. They are probably irrelevant except for foolish companies like Nokia.

  63. Linux Apostate

    Actually, with respect, you don’t understand. Drivers actually have nothing to do with the problem or the solution. It is all about the board support package (BSP).

    The BSP does *contain* some drivers, for things that the OS needs immediately and cannot load at a later stage, e.g. the interrupt controller and the timer. But its real purpose is to specify things like the location and size of usable RAM, the interrupt mapping and the locations of peripheral controllers. Without this information the kernel cannot even begin to boot, let alone get to the stage where it is able to load drivers.

    There is one basic BSP for Linux on all x86 PCs. Find it here. This is possible because, at the lowest level, all x86 PCs conform to a single standard.

    There are more than 40 BSPs for Linux on ARM devices, and that only counts the ones in the mainline kernel. Find them here.

    Why won’t one BSP do? Because the physical memory layout, the interrupt mapping, the timer, the interrupt controller, and the peripheral controllers are not defined by any standard. Ergo, every ARM device has to have a Linux kernel all of its own. There is no generic kernel that works everywhere.

    The hope is that Microsoft will standardise the BSP so that one generic kernel does, in fact, work everywhere. Like on a PC. This is a good thing which will benefit Linux and Android in the long term.

  64. Robert Pogson

    GNU/Linux or Android/Linux recognizes hardware through the drivers, ie if a driver exists it can be loaded at boot time and do the job. That’s what allows me to use a single image for 5 different motherboards in the system where I last worked. That’s foreign for Microsofties who would need one image for each but it’s one of the great features of Linux. Further, if code for a driver is submitted to the kernel group, it will be maintained by the kernel group and anyone who uses that hardware, so the load on the hardware manufacturer is lessened, not increased.

  65. Linux Apostate

    If every model from every manufacturer requires a different Linux kernel, then you rely on the manufacturer (or a third party) to keep releasing updates.

    I don’t quite understand why you don’t think this is a problem.

  66. Robert Pogson

    Nonsense. The Linux kernel takes care of the hardware. All else is a standard API from Android. Linux has proven itself capable of operating with all manner of hardware better than that other OS.

  67. Richard Chapman

    Okay “oldman”, let’s give Apple the leadership this time around. Now how do you feel about “standardization”?

  68. oldman

    Of course Pog, let a million reference platforms grow, each incompatible with each other! These manufacturers will learn the hard way as each produces its own version of hardware, each with its own firmware and kernel. NO standardized drivers for hardware.

    Its 1983 all over again!!!

    Of course to a hacker/geek like yourself, this is heaven and freedom. To myself, its yet another reason that ARM will go nowhere beyond the Embedded OS for smart phones and tablets.

  69. Linux Apostate

    I told you so, did I not :) It was inevitable that there would have to be some standardisation, some convergence between the radically different ARM-based SoCs in order to accommodate Windows and its applications. This standardisation is actually a good thing, not just for the people who might buy these Windows machines, but also for the people who might later buy them with Linux or Android.

    Acer never complained about having to make PC motherboards that conform to the PC standard. This is just the same.

  70. Richard Chapman

    My guess is that at some point, maybe at Ballmer’s departure, Microsoft will see the writing on the wall and backtrack a little. Then backtrack some more. Then issue a strongly worded memo to the hardware manufactures with a new set of guidelines that essentially signal Microsoft’s abdication of their reign of terror over them.

  71. Robert Pogson

    It’s probably more than a year away. That’s a lot of ARMed units shipped/activated.

    One of M$’s 247 patented methods of achieving/maintaining monopoly is to announce stuff that may or may not be delivered or on time. When all the PC makers were its “partners”, that froze the market and excluded competition. There are dozens of smartphone, tablet and ARMed chip makers who are not M$’s partners and M$ doesn’t have enough money to buy them all out. Nokia has agreed to be frozen and will have a big loss to show for it, much bigger than the cash M$ threw to them.

  72. Ray

    I thought windows 8 won’t be out for another year or two.

Leave a comment