The Smartphone Wars: Nod and WinCE

I learned this morning that Windows Phone 7 – the smartphone OS that bricks your phone! – is a skin over a bunch of core components from Windows CE. Which if you came in late, is widely regarded as the second most wretched hive of bugs and villainy Microsoft ever shipped.

(Yes, I said “second most”. Even WinCE could not come near matching the epic failitude of Microsoft Bob, which can only be explained by the justified supposition that its product manager was doing the mattress mambo with Bill Gates. She later married him, and Microsoft Bob disappeared down the event horizon of a black hole created by its own suckiness.)

My initial reaction to this news was “Doomed! What on Earth were they thinking?“. On reflection, however, there is an angle from which this way of slapping WP7 together makes a horrifying kind of sense. Not that I now think it’s any less doomed, mind you…but there are perhaps some useful lessons to be learned if we examine this fiasco from the Microsoft point of view.

It’s 2008. Microsoft’s previous attempt at a new mobile-phone OS, codename “Photon”, has just crashed and burned. Suppose you are the loyal Microsoftie told that Microsoft needs to produce a new smartphone operating system on a deadline of less than three years. You can’t use anything with open-source ick all over it because der Ballmer would have an apeshit freakout and fire your ass the second he found out; there go all your best options. The weenies over in R&D have things they call operating systems, but after listening to a couple of presentations about stuff like fine-grained parallelism and persistent objects you realize they really are research vehicles and adapting one would be as much work as building from scratch.

But you don’t have enough time to build a new OS from scratch. You’re on a tight deadline. And that deadline, for once, isn’t PHB bullshit; Apple shipped in early 2007, and by 3Q2009 there’s also a real risk from Android. In either case a late first ship could find a competitor so entrenched in a supermajority position that all your NRE would have been effectively wasted. If your higher-ups hadn’t held you to a tight deadline, that would have been good evidence that they were on bad hallucinogens.

Note that this is not exactly, or not only, a “Microsoft is incompetent” analysis. Android took four years, 2003-2007, and that was with the tremendous boost it got by re-using the Linux kernel. Building a production-quality OS is not a weekend project, you can’t re-use Linux, and the tempo of today’s smartphone market is brutally fast. As recently as 2004 or so you might have had the luxury of three or four years to first ship, but not in 2008. You have to re-use massive amounts of code to have a prayer of shipping soon enough.

There’s all the WinMobile and WinCE code lying around, and programmers who know it. Trying to salvage something from that wreckage makes sense even if you know it’s mostly shit. Otherwise how are you going to meet deadline?

Reusing as much of WinCE as you can swallow without gagging is the rational-minimax thing to do under the constraints you’ve been handed. That doesn’t mean it’s a good thing, necessarily, as the owners of now-bricked WP7 phones have just found out. What really went wrong here?

Closed source went wrong. I think it’s that simple; we don’t actually need the hypothesis that anyone at Microsoft was incompetent. They were boxed in at every turn by Microsoft’s insistence that we must own it and no open source can be allowed on strategically critical paths. They couldn’t reuse the Linux kernel, and their WinCE equivalent had three orders of magnitude less engineering time in it (judging by calendar years of deployment and the maximum possible size of the dev team). Possibly four orders…

There’s no recovery from this, I think, because having to base WP7 on WinCE wasn’t a contingent, accidental failure. It was an essential one. And, a few years from now when smartphones are almost everyone’s primary computing devices, it may be the particular one that goes on Microsoft’s tombstone.

208 thoughts on “The Smartphone Wars: Nod and WinCE

  1. I’d certainly agree that WinCE’s UI layer is pathetic (and I say that as the owner of a Samsung Omnia II), but is the kernel really that bad as well?

  2. Eric, the recent posts in your blog share a common theme: “how Google will win smartphone wars”, “how Android is best thing since sliced bread”, “how every computing device will be reduced to a smartphone” etc. It’s nice and entertaining read at first, but becomes boring after 8th post. Or 10th.

    Please, do us non-smartphone-freaks a favor! Post a story or two about how open-source bashes closed-source once again (now with a massive hammer: don’t skip the gory details!); how GPSD works despite bad hardware implementations and imprecise standards; how your genius is comparable to that of Asimov; about the non-arms-carrying persons who are “sheep” and too scared to admit it. (Well, the last two are optional :-))

    I also thoroughly enjoyed descriptions of your software projects, be it about INTERCAL compiler, tool to extract data from repositories, or troff->DocBook converter. Is anything going on in these areas? I bet there is, so why Android and smartphone wars get all the glory while it can be devoted to ingenious COME FROM statement? ;-)

  3. Eric,

    The only thing that surprises me about this is that you were surprised. For exactly the reasons that you gave above, I had always assumed that WP7 was WinCE-derived.

    I’m curious as to whether you just hadn’t thought about the man-hours that would have been required to make WP7 a truly new product, or if you had some other reason for this assumption?

  4. I’m not sure it was so much an explicit rational economic decision as an implicit cultural one. Except for Windows NT (and there they reused the APIs), when has MS ever done anything except build on the old versions? I too pretty much assumed that it was a reskinned WinCE derivative from the start.

  5. I assumed Windows Mobile 7 was WinCE derived because, um, every single version of Windows Mobile for the last 10 years has used WinCE as the kernel. Including Photon.

  6. Makes me think about Symbian… I don’t know how bad its much-maligned internals actually are, but if you look at the kinds of things that the Linux developers have been able to do and the kinds of developer resources that have been poured into Symbian, you’d think that they should have been able to fix quite a bit. Symbian has been open source for a year now, which is not long for a famously convoluted OS code base to attract outsider developers. Even so, Nokia had two open source operating systems, both proven on their phones, with Qt on both of them, and then they decided to go with WinCE/WP7 to build an ‘ecosystem’? Hmm.

  7. Other than lack of recent *nix experience, why couldn’t M$ use the BSD kernel? Worked for Apple.

  8. > Other than lack of recent *nix experience, why couldn’t M$ use the BSD kernel? Worked for Apple.

    I think you just answered your own question. Thrice.

    1) It’s nix.
    2) It’s open source (BSD).
    3) It’s what Apple did.

    On Eric’s overall point, it’s not surprising that MS did this. What choice did they have while still being MS? The surprise is that Nokia would be so foolish.

    Is there a word like ‘kamikaze’ in Finnish?

  9. Open-source kernels were not their only option. RTOS and QNX spring to mind.

  10. > Is there a word like ‘kamikaze’ in Finnish?

    No thanks, we’re Lutheran. (Well, not me, personally, but still. Is there a word for it anywhere outside of Japan?)

  11. Somehow I doubt WinCE was something they used regrettably. Remember, the culture inside Microsoft still thinks that their own products are awesome.

    They probably said “we’ve got a kernel, let’s add a .net runtime and a new skin and we’ll throw the number 7 on it because people liked that number on desktop windows” and that was that.

  12. They could have used XP embedded :).

    But first, where have you heard this – is it a rumor, “Halloween memo”, or did someone do a side-by-side of the APIs.

    My first memory of CE was when Microsoft put 4 CDs with the SDK into an issue of Embedded Systems Journal. Of course I threw them away, but 6 months later I had reason to be curious. The problem is that everyone else in our department did exactly the same thing.

    Later, somewhere else they were looking at the severe interrupt timing jitter on a scope and asked one of our other experts what they were doing wrong (as Linux or everything else was rock solid). He basically said it was just CE. Current GPS devices are based on it, though given their price they can’t be charging much for it. (I think when asked, I quoted how much they would have to pay me to even attempt to use it). The PDAs competing with the Palm/Pilot at the time were horrid – handwriting recognition worse than newton, crashy, glacially slow, low battery life, etc. all on something with much more hardware power.

    CE = Crippled Edition. Mainly for the versions of IE, Word, Excel, etc. on the devices.

    My guess would be that CE was altered (cue scene from “Frankenstine”) so that it could run the dot-Net framework and Silverlight and the rest of latest trashy APIs they have come up with (every time they figure out the current one is as bad as MFC, they come out with a totally new one).

    And here I think is the crux of the matter. They could have bought or designed some simple kernel and reengineered it from the ground up to be light, robust, and reliable, but at the same time, they would have to redo dot-Nest, Silverheavy, and their other abominations that can kill a netbook and bring a gamer system to its knees to fit on top of that lightweight but strong framework.

    But the requirement was not so much use WinCE/WinMobile per se, but they were stuck using the heavyweight, complex interfaces and models to the kernel inherited from the WinNT series of mistakes (fond memories of comparing heavyweight VMS, its prototype to lightweight Unix before Windows).

    A 50 pin centronics style SCSI port – with passive termination – in the age of SATA.

    You need to consider the choices for the upper layers – the Dalvik flavor of Java with XML for Android, and objective-c with the plist resources for iOS. Crossing the platform boundary is possible (albeit with jailbreakage), and would probably run decently if attempted.

    Ask yourself what you would need to port the dot-Net and silverlight framework to run on top of these. Or look at the severe fatigue caused by Mono in the pale Moonlight.

  13. Let me reduce the above:

    Microsoft’s current “ecosystem” is a dot-Net monoculture. They needed to run IE and whatever other apps written to that framework. It has to be able to use Visual Studio. You want to prohibit Java or anything else and even make C or C++ difficult. Significant systems (e.g. cert management) are dot-Net.

    So with the only non-negotiable that you will need to run IE(n+1) and Silverlight on the platform, which will in turn require the dot-Net framework, which will need…

    Could you create an underlying infrastructure that would support the above and actually be less clunky than some embedded Windows derivative such as CE in less than a decade?

  14. > Is there a word like ‘kamikaze’ in Finnish?

    Don’t speak Finnish, but they do have a Swedish-speaking majority, and I speak Norwegian, so – close enough? Anyway, in old Norse you have the word ‘feig’, which survives in modern English as ‘fey’, uncaring whether one lives or dies. In modern Norwegian the word has changed meanings and now is best translated as ‘cowardly’.

  15. >> Closed source went wrong. << It's not a closed source issue. If you strive to provide a great good or service then the money follows. When you get distracted by things like money or market share then it all goes off the rails. Open source is immune from the usual corruptors.

  16. It’s worth mentioning that when Microsoft *originally* needed an OS for handheld devices they just went out and found a small embedded-systems company, bought their OS and the people who came with it, and built what they needed on top of it. Apple did something similar for the first iPod – contracted out the OS layer to somebody else who already had one ready. You don’t seem to have listed that as one of the options for a closed-source company.

    I guess they couldn’t do that again now because the bar has been raised so ridiculously high – Apple and Android are waaaaaaaay ahead in this race. Microsoft surely could have shipped *something* in the required timeframe by throwing money at it but it wouldn’t have been something that was positioned to catch up with the leaders in features any time soon. They needed whatever headstart they could get and then some to even stay in the race.

  17. Count me as surprised that esr is surprised. When ’7′ was announced, I thought about what they’d been up to, who they’d bought recently, and tentatively decided they had to have done it in house as a crash effort. they’ve got nothing in house except for what’s sitting on the shelves. (I wondered for a bit if some xbox work could be used, but that’s easy to discount.) plus, they have a cultural resentment thing going, sort of, if everyone else only understood…

    So, um, yeah. CE + go hire someone that wishes they were Frog Design for a skin that will slide by the trade press. Ignore that the trade press doesn’t much matter anymore- that’s how we roll, do it, motherfucker. It bought them Nokia…

    Did anyone expect anything else?

    I’m personally waiting for OLE to be rebranded for sharing shopping lists (wirelessly!), or something.

  18. > Closed source went wrong. I think it’s that simple; we don’t actually need the hypothesis that anyone at Microsoft was incompetent.

    Need some incompetence. If MS has decided on closed source as a strategy, the correct response from the loyal Microsoftie is, “Three years is not enough to produce anything worthwhile,” and then the boss needs to say, “Ah, I guess we’re just boned in this market, then.”

    Though Apparently this is really difficult, judging by the frequency it occurs. It does require two people to not suck simultaneously.

  19. It had to run on ARM. I think that was the roadblock for any XP or Xbox port.

  20. > they do have a Swedish-speaking majority

    …of about 5 % of the population.

  21. The Norse mythology had Berserkers as a close analog to the “wind of the gods”. I would not assume the Fins had something alike.

  22. The winCE kernel is FINE. Really. There is nothing wrong with it.

    The problem is that insufficient effort was put into making the APPLICATIONS work. You know, like the PHONE application, which couldn’t even respond to the call-pickup button in under 2 seconds. Battery life also sucked on all the larger devices (not so bad on the chocolate bar range, but still nothing special).

  23. > … of about 5% of the population
    They get taught swedish in school. They just don’t speak it, why should they?
    Anyway the finns do have a word they’ve lost, or Elop don’t know it, Sisu. Basically is willpower to pass through any challenge. And they aren’t with this, they are leaving an old and crippled system for one that has been declared dead on arrival.
    But I can’t say I’m surprised about the heritage of ’7′. Microsoft have never really been one to make something new. They have always copied, bought or reskinned something and said it was a new product. They’ll be on the way out as well I think. Much like Nokia they are resting old laurels and that doesn’t work in the long run.

  24. Eric:

    Melinda French (later Gates) left Microsoft before dating Bill Gates by most accounts, largely to avoid any issues about “Well, she’s the head honcho’s girlfriend.” The woman has a deadly serious attitude about ethics. (She is largely the reason Gates transitioned from running Microsoft to philanthropy by his own admission.)

    As to MS Bob: When many of the features of MS Bob were mapped to (much better) hardware, they became one of the most successful UIs (if not an operating system shell) of all times.

    I am, of course, referring to The Sims, which bids fair to being the most successful cross platform user interface ever written.

    (I used Bob a couple of times in 1995; having read Snow Crash, it really was an attempt at making a Metaverse UI work, only without networking.)

  25. Was MS Bob really that buggy? I haven’t used it, but from what I understand the main complaint was that it was ill-conceived.

  26. MS could come out with an XBox handheld which is also a phone, speaking of software they already have. If it is as good as the XBox it would sell well, although it would not threaten Android. That does sound like a major development effort, I admit.

    Yours,
    Tom

  27. Max, the main problem with MS Bob stemmed from the following:

    1) It tried to ‘map’ a 3-D ‘house’ UI to make things easier to use to an environment where there was no real 3-D graphics display.
    2) It was SLOW. Much more graphically intensive than any competing UI.
    3) It was a cluttered UI – things were seldom in easy to find or consistently applicable places.
    4) It made Windows power users kvetch because they couldn’t find anything.

    Microsoft tried doing something that was, in retrospect, about a decade ahead of what the hardware could handle.

    While everyone remembers the innovations that succeed, very little credit is given to Microsoft for trying something novel and exploratory.

    Microsoft learned their lesson from Bob: Don’t change UI very much in one incremental update unless you ABSOLUTELY have to. The Office 2007 Ribbon is a case where they pretty much had to change the UI to get what they wanted done…and there are still people, 4 years on, who complain bitterly about it.

    (The Office Ribbon UI is one of the more interesting experiments in UI I’ve seen. It’s MUCH better organized than the menu system it replaced.

  28. >Was MS Bob really that buggy? I haven’t used it, but from what I understand the main complaint was that it was ill-conceived.

    So, it wasn’t “buggy”, it was a bug?

  29. Melinda the manager making the mattress mambo with her manager at Microsoft? How mahhvelous!

  30. I stand by the requirement that WP7 had to be dot-Net + Silverlight, developed in Visual Studio. It goes back to the developers conference in Las Vegas. I caught Paul Thurrott on twit.tv windows weekly and he was saying how great the environment was, and it really bothered me but I couldn’t quite understand why (probably the VS/NT3 dev stuff for CE attempting to cause an unpleasant flashback). Dot-net, Silverlight, and IE and the other bloatware on a mobile device?

    Android used portable technologies. Linux is at home on PPC and ARM as much as x86. So is iOS – going back to the PPC Macs. They started and stayed much lighter, but you can mix and match. There is controversy about Wayland replacing X, but the most blatant but unmentioned fact is that the FLOSS infrastructure is so decoupled and modular that you can carefully rip out the middle layer and replace it with something else and nothing breaks. All the chunks of Gnome that don’t care about the display are elsewhere, and where it attaches is narrow. There’s even a fairly full set of direct framebuffer stuff.

    To go back to the windows conference, let me replay that from the other side that to make it clear: No GCC, or opensource libraries or frameworks. No Java. No Flash, or anything else. Not even the older APIs. It will ONLY be written in C# using something that is or looks like dot-Net using clones or ports of the APIs you know and either hate or love on Windows XP/Vista/7. The web browser will be limited to IE and the APIs that are part of it or that it uses (worse than the way iOS is limited to Safari – I could theoretically port other things to a jailbroken iPhone).

    If you aren’t some kind of strategic partner, you can’t do anything else anyway. Nor should developers care about what is below the API. But this is the Windows API, so to implement it “from below”, means all the original compromises, assumptions, quirks that have accumulated in the Windows system have to be in whatever Microsoft used to implement the API. Mono is under an incompatible license. I suspect if it was BSD it would have been stolen. So they have to use their own library source and bludgeon it so as to run without too obvious of a limp on a mobile device. But the libraries have calls to the system – DLLs, so you have to port them too. And those DLLs? Especially with “IE is part of the OS” mentality? You need something that looks like the windows kernel. And windows CE / windows mobile, for all its other failings, is still Windows.

    So because the front end was VS/C#/.Net/IE/Silverlight, at each lower layer of the system, it forced it to be whatever they did for windows, or else a complete API implementation.

    This is also not specifically an open/closed source controversy per-se. It is about the coupling and modularity. There are plenty of closed source RTOS kernels and stacks. But because Bill Gates, way back when, wanted to give Netscape users a jarring experience, and merge the browser with the OS, and generally make it hard as possible to break out any component, Windows is not so much an OS as a monolithic program going from the kernel to the application. OpenSource tends to be very loosely coupled because no one can develop any too large piece, and you cannot be sure what the interfaces will be in the future, so they are made small and flexible. Closed source can be either, but Microsoft made very specific efforts to go monolithic when they thought it would help them.

    There is only a Silverlight plug-in for the Macintosh. There is no dot-net nor AFAIK does the Mac group use anything from that codebase. I think it is because they can’t – the pieces used for the windows code base are so welded and inseparable, that you would end up writing a VM. Mono is a porting framework, and it took a lot of effort, but was only possible because they can translate to all the pieces due to their thin APIs. But I don’t think even if you had the source code to IE9 that you could port it to Linux using Mono without recreating an equal sized blob to translate the rest of the Windows baggage.

    When they chose to make WP7′s development chain and frameworks basically what is on W7, they were doomed. Because the only way to do it was to copy over the whole monolith. Wonder why they are talking Windows 8 for ARM? Basically it will be WP7. Windows 8 might just be the point of collapse the whole thing since their big point has been to preserve legacy and backwards compatibility, and they didn’t maintain it from the break back when NT ran on the DEC Alpha.

    Right now you see cheap ARM based android tablets from China for less than a Win7 retail license costs. Perhaps they can port Windows to ARM, but then how are they going to get a windows tax? They had to cut margins to displace Linux netbooks – but the SoC is going to reduce the cost of these devices. 10% is possible, 20% is harder, but who will want to pay 50% to 100% more for the privilege of running Windows? And worse for them, I think there will be an android feed-forward (maybe chromium?) to the x86 PC market.

  31. @tz Those are all very interesting points. In a best case scenario they could lead an optimist to believe that it really is possible for the Windows ecosystem to actually *collapse* in the not too distant future.

  32. But because Bill Gates, way back when, wanted to give Netscape users a jarring experience, and merge the browser with the OS, and generally make it hard as possible to break out any component, Windows is not so much an OS as a monolithic program going from the kernel to the application.

    No, no, no.

    IE won the browser wars because it was better. Microsoft may have wanted to “cut off Netscape’s air supply”, but anticompetitive tactics alone wouldn’t have done it. The writing was on the wall for Netscape in late 1996, before the browser tying began.

    And actually, one of Internet Explorer’s advantages was that Microsoft did just the opposite of what you proposed: they completely modularized the browser and made it a component that could be called and embedded by other applications. This was a huge win as it meant that app developers could rely on the presence of a state-of-the-art HTML engine rather than having to roll their own. If an open-source component exposed the same COM interface that IE does, it would work in place of IE’s engine anywhere there were calls to it from a third-party application. (Imagine Iexplore.exe booting up with Mozilla’s rendering engine?!)

    Microsoft, in general, wants to create the perception that their software stack is more monolithic than it really is. This is to get people to buy into Windows lock, stock, and barrel. Some of their tactics are rather devious — for example, attempting to detect a non-Microsoft component and deliberately breaking or raising an error message if one was found (as was done by Windows 3.1 for non-Microsoft DOSes). There’s other assorted nastiness like dumping C structs to disk and calling them file formats (as Office used to do), making readability of those formats depend on knowledge of the C struct and compiler and compiler version, but I digress. Microsoft may be evil, but they aren’t that stupid. Overall Windows is a very modular and cleanly separated system, thanks in no small part to technologies like COM, which publish and enforce explicit interfaces between system components.

    But the libraries have calls to the system – DLLs, so you have to port them too. And those DLLs? Especially with “IE is part of the OS” mentality? You need something that looks like the windows kernel.

    This, too, is nonsense. Win32 has existed atop at least three different kernels: DOS (if you can call it that), NT, and CE. The NT kernel’s system calls look nothing like the Win32 API, and Microsoft deliberately keeps the kernel undocumented, because app developers are supposed to use the Windows “personality” to talk to it.

    Relatively complete Win32 implementations have existed on Unix kernels — Wine, for example. The problem with projects like Wine is not that the API depends on the NT kernel, but rather finding and implementing all of the bugs and quirks that applications written against Win32 depend on.

    There is only a Silverlight plug-in for the Macintosh. There is no dot-net nor AFAIK does the Mac group use anything from that codebase. I think it is because they can’t – the pieces used for the windows code base are so welded and inseparable, that you would end up writing a VM.

    It’s because they want .NET to be a sales driver for Windows. See above — they want to create the perception that the Microsoft stack is an all-or-nothing affair. Silverlight itself contains much of the .NET standard libs.

    When they chose to make WP7?s development chain and frameworks basically what is on W7, they were doomed. Because the only way to do it was to copy over the whole monolith. Wonder why they are talking Windows 8 for ARM? Basically it will be WP7.

    Windows 8 is being ported to ARM because ARM is the future. The peecee is dead except for office workers, geeks, and maybe hardcore gamers. Smartphones and tabs are becoming the first widespread truly personal computing devices: people carry them everywhere, fiddle with them when bored, etc. Microsoft wants a piece of that. They’re not going to let Linux trammel their chances, and even they recognize that Windows Phone 7 is not a complete solution.

  33. If an open-source component exposed the same COM interface that IE does, it would work in place of IE’s engine anywhere there were calls to it from a third-party application. (Imagine Iexplore.exe booting up with Mozilla’s rendering engine?!)

    Looks like it’s been abandoned, but — http://www.iol.ie/~locka/mozilla/control.htm

  34. >> they do have a Swedish-speaking majority

    >…of about 5 % of the population.

    Ooops! I meant, of course, ‘minority’, but my fingers slipped on the keyboard.

  35. This is the same thing they did with Bing. MSN Live was a joke. They spent a while improving it, but no one gave them a second chance. So they rebranded it Bing. The first day of Bing gave the exact same results as the last day of Live, but the new name drew in users who would never have looked at Live. It worked — Bing now has nontrivial market share.

    So they’re trying the same thing with WinCE/WP.

    Come to think of it, has anyone bothered to look at WinCE recently and check to see if it still sucks?

  36. @Jeff Read;

    If an open-source component exposed the same COM interface that IE does, it would work in place of IE’s engine anywhere there were calls to it from a third-party application. (Imagine Iexplore.exe booting up with Mozilla’s rendering engine?!)

    This actually exists in Wine. In fact, it works well enough that stuff like Google Sketchup’s Web dialogs will work under WIne.

    That last link comes from my blog. If you read the entry carefully, I’m instructing people who want to get Web dialogs working to install IE8 and then uninstall it. What you’re left with is Wine Gecko with a few additional pieces of infrastructure left over from IE8 that you’ll get working Web dialogs.

  37. Windows 8 is being ported to ARM because ARM is the future. The peecee is dead except for office workers, geeks, and maybe hardcore gamers. Smartphones and tabs are becoming the first widespread truly personal computing devices: people carry them everywhere, fiddle with them when bored, etc. Microsoft wants a piece of that. They’re not going to let Linux trammel their chances, and even they recognize that Windows Phone 7 is not a complete solution.

    But they it’s not a war that they can win. It’s much too late for that. Everyone forgets that IBM ported their once-dominant mainframe OSes to PC hardware (see OS/400, the other iOS) but they weren’t able to follow the paradigm shift. In the end, they had to reinvent themselves as a services company to stay alive. MIcrosoft is suffering the same fate as IBM: they are losing their stranglehold. They’ll only be able to retain relevance through support of legacy systems and reinventing themselves as some other kind of company.

    BTW — Steve Balmer isn’t the kind of man who can lead them through such a transition. My prediction is that within the next year or two, Balmer will be replaced by Microsoft’s Board of Directory.

  38. @Morgan Greywolf
    “But they it’s not a war that they can win. It’s much too ate for that.”

    MS lost their monopoly in 2009 after the Vista disaster and the barrage of Linux netbooks an Apple laptops.

    They lost their sense of direction in 2010 after losing yet more developers and directors and missing the ARM boat.

    They will ground this year after losing out everywhere but the desktop. It will become clear this year that MS will be unable to break out of their home, desktop, market ever.

    After realizing that all money spend outside the desktop franchise is pure waste without any hope of return, shareholders will force shedding of all business losing money. So in 2012 MS will start to fall apart.

  39. My prediction is that within the next year or two, Balmer will be replaced by Microsoft’s Board of Directory.

    Volume in drive C has no label…

    Freudian slip?

  40. MS lost their monopoly in 2009 after the Vista disaster and the barrage of Linux netbooks an Apple laptops.

    Windows market share is still high 80s.

    Also, what Linux netbooks? You mean the ones people kept returning because they sucked? Thankfully, you don’t see any of those anymore…

  41. @Jef Read
    “Windows market share is still high 80s. Also, what Linux netbooks?”

    A monopoly means “No Choice”. After Vista people saw they have a choice. When I visit a meeting or conference, I see a lot of mac laptops. There are netbooks because there was Linux. It took MS close to a year to force OEMs to wipe Linux from the netbooks. Which were so popular that they created a complete new market.

    If Taiwanese CEOs offer a public appology for having shown a Linux netbook at a show, you ought to know why you cannot buy them.

  42. >You mean the ones people kept returning because they sucked?

    Urban myth, long since debunked. The return rate on Linux netbooks was in the normal range for laptops and PCs.

  43. Matt> it’s not in the culture to make something this fundamental new and functional. Microsoft simply isn’t capable.

    esr> You can’t use anything with open-source ick all over it because der Ballmer would have an apeshit freakout and fire your ass the second he found out;

    Exactly; it’s cultural, from the top down. A failure of vision. I hesitate to use the word ‘commoditization’ in reference to OSes, but it’s long since happened in the hardware space. Working in that field, I know all too well the insane level of competition (but how do you commoditize something that is free?). Microsoft from the very beginning was not set up to compete, only monopolize through vendor lock-in. So as the incremental cost of deploying an OS has dropped to zero (one Linux LiveDVD will install on infinite machines, or as many as possible before the disk degrades), the only thing preventing change is inertia. MS has been relying on this inertia to extract their rents. Along comes a game changer (smartphones) where MS is late to the game and can’t secure their vendor lock-in to perpetuate the inertia. All the innovation in OSes happens outside Redmond, and The Ballmer’s ego (and other’s, I’m sure) won’t let them explore that realm. NIH will prove to be the downfall of MS.

  44. > A failure of vision.

    I guess this is why it’s so difficult to swallow the spiel about MS and Nokia being a good fit for each other. They’ve both failed in a lot of the same ways: both missed the current smartphone boat, both had an app/music store with a terrible user experience, both had mysterious holes in the mobile apps (e.g. chat missing on Kin phones) etc. A large-scale collaboration between two big corporations seems like just about the last thing that might fix it.

  45. A good metaphore of NoWin phones was a textbook anecdote on plant breeding.

    Tomatoes and potatoes are from the same plant family. So a good cross would give a marvelous double crop plant. They actually succeeded in crossing the two and the result had indeed a mix of qualities from both species.

    The fruits (berries) of the potatoo and the roots of the tomatoo.

    I think the synergy between Nokia and MS will be equaly friutful.

  46. If the speculation in that article is true, then Verizon can’t do anything 4g except Android…

    If Verizon’s T&C for a 4G iPhone were to involve a clause that says something along the lines of “you acknowledge that Verizon has no responsibility or liability regarding the maintenance of phone software or phone software marketplaces that are provided and/or maintained by third parties” then surely it’d be pretty hard to say that it’s Verizon’s fault that iPhone is a closed platform. If the US courts made that leap then Apple could just as plausibly sue Verizon because by not allowing iPhone 4G on the network, Verizon is specifically not allowing “customers … to use … devices … of their choosing in C Block networks”.

    Basically i figure that’d have to go to court and if the court finds for Google in this then Verizon is screwed either way and i’m relatively sure that Apple and Microsoft would both point that out pretty loudly. The only interpretation that isn’t self-contradictory is that Verizon isn’t allowed to do it themselves, perhaps even extending to (for example) giving Apple money to do it for them (though proving that is going to be difficult without whistle blowers).

  47. “There’s all the WinMobile and WinCE code lying around, and programmers who know it. Trying to salvage something from that wreckage makes sense even if you know it’s mostly shit. ”

    Why not Windows 2000? That was the first stable enough version, relatively reliable and with low enough resource requirements to be used in a phone?

  48. @JonB:

    I’m not sure the law works that way. If I’ve agreed that every device attached to my network must allow the user to run software of the user’s choosing, I might not be legally able to “knowingly” sell a device that doesn’t allow that.

    BUT…

    There’s probably absolutely nothing preventing Verizon from selling SIM or microSIM cards :-)

    That way, *they* haven’t sold the offending phone.

    Alternatively, Apple could do a very small about face and allow private app sales some way. I don’t know that they need to make it convenient, although I haven’t read the legal docs surrounding the spectrum sale.

  49. Morgan Greywolf:

    But they it’s not a war that they can win. It’s much too late for that. Everyone forgets that IBM ported their once-dominant mainframe OSes to PC hardware (see OS/400, the other iOS) but they weren’t able to follow the paradigm shift. In the end, they had to reinvent themselves as a services company to stay alive.

    Er, no.

    The OS/400 was a ground-up design aimed at DEC; an explicitly midrange product aimed at medium-sized businesses who were buying VMS and couldn’t be interested in the bigger, more expensive IBM mainframe gear. Originally on custom hardware, these days it runs on IBM’s Power systems. It had some interesting design properties – the integration of a proper RDBMS in the base OS, for one, and a sort of persistent memory IIRC.

    Of course, UNIX killed both of those, so your main point stands.

  50. My expectation is that Microsoft isn’t migrating Windows 8 to ARM for consumers or phones, though that’s not a market they intend to turn down.

    It’s for server rooms. The cost of server hardware has plummeted to the point where electrical costs are now one of the primary areas where cost savings can be squeezed out of servers.

    Microsoft is asking Intel to produce a 16-core ATOM chip for just this market. Lots of server apps don’t need massive CPU processing power; they do need good I/O channels and fast enough disks and fast enough RAM.

    ARM’s architecture looks (from the very much uneducated layman’s perspective I have on it) like it might be scalable on I/O throughput in ways that Intel’s is not.

    Insofar as I am able to tell, more than 2 cores results in a lot of wasted CPU cycles on most platforms other than servers running virtualization. The i5 and i7 platforms from Intel are geared around turning cores on only when needed, then turning them off again to save power…

    In short, the server room is going to get the ‘netbook’ treatment – cheap, commodity hardware that runs at lower voltages and consumes less power, and gives performance roughly equal to a computer that’s about 4 Moore’s Law generations out of date, because that’s about all the performance that’s needed for the task at hand.

  51. There’s probably absolutely nothing preventing Verizon from selling SIM or microSIM cards :-)

    The fact that Verizon has a CDMA network is a pretty big preventer of that.

    GSM is like the metric system: the more the rest of the world standardizes on it, the more the USA retardedly insists on not doing so.

    ARM’s architecture looks (from the very much uneducated layman’s perspective I have on it) like it might be scalable on I/O throughput in ways that Intel’s is not.

    I suspect that I/O throughput is one of the big reasons why IA-32 and AMD64 have so far prevailed in the server room. These architectures come with a raft of I/O standards: the PCI and PCIe buses, “northbridge” and “southbridge” I/O control electronics, IDE host adapters with various DMA modes, etc. ARM systems have been historically optimized for small size, low power draw, and real-time performance, not raw I/O throughput, so oftentimes they are wired directly to whatever device they are intended to control. Imagine my surprise and dismay when, while evaluating a particular ARM SoM, I discovered that its IDE interface consisted of 40 wires tied directly to the CPU’s memory lines. To access the disk, the kernel would bang the lines directly in a tight inner loop in the driver. No DMA support at all, and your CPU got tied up during a disk access. Way too slow for what we were going to use it for.

    Not that all ARM systems are like this, but there doesn’t seem to be the standard ecosystem of hardware support around ARM that there does for PCs (for justifiable reasons). This may change as ARM makes server-room encroachments; however I don’t see that actually happening yet.

    Although if you’re serious about I/O throughput, IBM mainframes are still the way to go.

  52. @shenpen
    “Why not Windows 2000? That was the first stable enough version, relatively reliable and with low enough resource requirements to be used in a phone?”

    Because NT/2000/XP/Vista/7 do not run on Arm. And it HAD to run on ARM.

  53. @Jeff Read
    “Not that all ARM systems are like this, but there doesn’t seem to be the standard ecosystem of hardware support around ARM that there does for PCs (for justifiable reasons). This may change as ARM makes server-room encroachments; however I don’t see that actually happening yet.”

    They know and are working on it:

    ARM dips toe into server chip design
    http://www.zdnet.co.uk/news/servers/2010/05/06/arm-dips-toe-into-server-chip-design-40088862/

    ARM-Based Servers Expected by 2011
    http://www.tomshardware.com/news/Server-CPU-Xeon-Opteron-ARM,10302.html

    ARM Holdings’ 2015 plan: Grab PC, server share
    http://www.zdnet.com/blog/btl/arm-holdings-2015-plan-grab-pc-server-share/44386

    The ARM Cortex-A processors support multi-processing and that delivers the high levels of performance required for server applications. And in this space we have zero market share today and we have an ecosystem that we need to develop. So this is very much a potential opportunity and it’s market growth in the out years.

  54. Jeff Read:
    >Although if you’re serious about I/O throughput, IBM mainframes are still the way to go.

    I keep hearing this, but I never see a UNIX box that isn’t limited by the physical devices it’s attached to. Have you actually got some kind of verifiable cite for this?

  55. I keep hearing this, but I never see a UNIX box that isn’t limited by the physical devices it’s attached to. Have you actually got some kind of verifiable cite for this?

    When people speak of raw I/O throughput, they’re usually talking about storage. Thing is modern fiber channel SANs will rival the raw throughput of an IBM mainframe. Throw in 2-4 of decent QLogic FC controllers in any commodity PC server running Linux, use Multipath, and connect it to a reasonably configured FC SAN and you’ve got about as much storage throughput as an average IBM mainframe.

  56. > GSM is like the metric system: the more the rest of the world standardizes on it, the more the USA retardedly insists on not doing so.

    IIRC, CDMA uses bandwidth, which is still not cheap, more efficiently than GSM.

    Yours,
    Tom

  57. IIRC, CDMA uses bandwidth, which is still not cheap, more efficiently than GSM.

    You’re right. I think Jeff Read spends way too much time on /.; it’s rotting his brain. Western Europe != “the rest of the world.”

  58. There’s probably absolutely nothing preventing Verizon from selling SIM or microSIM cards :-)

    The fact that Verizon has a CDMA network is a pretty big preventer of that.

    Inside every one of their LTE USB devices is a SIM card…

  59. There’s probably absolutely nothing preventing Verizon from selling SIM or microSIM cards :-)

    The fact that Verizon has a CDMA network is a pretty big preventer of that.

    GSM is like the metric system: the more the rest of the world standardizes on it, the more the USA retardedly insists on not doing so.

    All of Verizon’s LTE devices will have SIM cards. LTE is part of the GSM family of protocols.

    http://en.wikipedia.org/wiki/3GPP_Long_Term_Evolution

  60. >I’m not sure the law works that way. If I’ve agreed that every device attached to my network must allow the
    >user to run software of the user’s choosing, I might not be legally able to “knowingly” sell a device that
    >doesn’t allow that.

    I don’t know, IANALBIPOOTI, it seems to me that the law in question reads that the carrier may not interfere with the device or restrict it or prevent the user from installing the apps they want to install. Nothing in that seems to suggest that if the device itself is already limited that Verzion couldn’t sell it. If you read it the way the article suggests, you could also argue that if vendor X sells the “Ubar Phone” and the “Ubar Phone Lite” both of which use the same hardware under the hood, but “Ubar Phone Lite” maybe has it’s NFC chip turned off, and is sold at a lower price, that Verizon wouldn’t be able to sell the “Ubar Phone Lite”

  61. @tmoney:

    Nothing in that seems to suggest that if the device itself is already limited that Verizon couldn’t sell it.

    The slippery slope is that when a large company (Verizon, WalMart, Home Depot, whoever) supplies things to consumers, those things are quite often engineered specifically for that particular outlet. In other words, there is a good argument to be made that an extremely large reseller is never actually just a reseller, but a joint participant in actually defining and building the device.

    Everybody involved (Google, Apple, the FCC) have lawyers, and they all have different takes on it. Google’s take is (new article) definitely that:


    The Commission’s open access rule is clear that C Block licensees “shall not deny, limit, or restrict the ability of their customers to use the devices and applications of their choice….” The rule thus plainly proscribes a C Block licensee from selling handsets to customers that hinder a customer’s ability to use applications of their choice, and applies to all customers of a C Block licensee.

    Verizon’s position would completely reverse the meaning of the rule such that the open access condition would apply to none of Verizon’s customers, and thereby render the condition a nullity.

    The law could be a little clearer, but one potential problem for Verizon is that laws are often read in conjunction with the record that led up to the law being passed. From the same article:

    At the time, Kevin Martin was the head of the FCC and had advised Congress that “this condition means all handsets will be unlocked and open to all apps,”

    I don’t think Google (or the law) cares if Apple sells a 4G iPhone that works on Verizon. But Google wants to make sure that Verizon itself can’t stock its stores with closed phones that aren’t allowed to drive ad revenue to Google, and (at least in 2008) was busy advising the FCC about what the law meant. Since Google and Verizon are better buds now, I don’t know if they would be pushing this hard now or not.

  62. >But Google wants to make sure that Verizon itself can’t stock its stores with closed phones that aren’t allowed to drive ad revenue to Google,
    >and (at least in 2008) was busy advising the FCC about what the law meant. Since Google and Verizon are better buds now, I don’t know if
    >they would be pushing this hard now or not.

    It would be a pretty significant (and IMHO, very bad and dangerous step) to set the precedent that Verizon can’t sell the products they want to sell and not sell ones they don’t want to sell. The requirement that they must allow any device or app the consumer chooses to bring is a very large step away from they can’t sell any device that isn’t 100% open.

  63. It would be a pretty significant (and IMHO, very bad and dangerous step) to set the precedent that Verizon can’t sell the products they want to sell and not sell ones they don’t want to sell.

    I disagree completely that it would be a particularly significant step. Verizon has a government-granted monopoly. Certainly they paid a bucketload of money for said monopoly. Certainly there were terms and conditions attached, in addition to the requirement to pay more money than any other bidder. If some of the terms and conditions have the effect of turning the auction winner into more of a pure-play bandwidth provider than an end-to-end provider, then google will have achieved its ends in entering the license auction, without having to lay out more than for some lawyers and PR people — a very smart move.

    Most government granted monopolies operate on a sliding scale of evil, but the least the government can do after giving Verizon the role of exclusive gatekeeper to this spectrum is to reserve a few rights to all those parties who are not Verizon.

    Now, as to it being a bad step or not, that all depends on what Verizon wants to do, and the Camel’s Nose, and whatever other fairy tales you can think of. For example, my reading of the rules on spectrum block C says, for one thing, that Verizon can’t sell a device that operates in this spectrum if the device is locked to Verizon’s network. So if a single iPhone 5 works with multiple 4G/LTE vendors, I can see Verizon being permitted to sell it, despite Google’s protestations, and I certainly won’t get too worked up about it. But if Apple makes a 4G/LTE iPhone 5 that only works with Verizon, and sells it through Verizon stores, that would seem to flout the rules even by a fairly liberal interpretation of the rules, and I think it would be wrong.

    In the future, if you have to finance a smartphone through a contract with a true interest rate that’s probably greater than 30%, once you’re done with the contract, the phone ought to still be useful to you, even if you have a falling out with the company you bought the phone from.

    I would like to think this would happen organically, but it’s always hard to say if that will work, once the herbicide of monopoly is applied anywhere near a given market.

  64. The slippery slope is that when a large company (Verizon, WalMart, Home Depot, whoever) supplies things to consumers, those things are quite often engineered specifically for that particular outlet. In other words, there is a good argument to be made that an extremely large reseller is never actually just a reseller, but a joint participant in actually defining and building the device.

    Right and therein is the edge case that would be hard to prove. Actually you raise a fairly easy test that would exonerate Verizon. Is the device sold by Verizon more locked down than equivalent devices(in this case other iPhones) sold by other providers? In the blanket case (i.e. same app store, same effective iPhone with a 4g antenna on it) the answer would be fairly obviously no. (A customer could buy all the same applications on Verizon that they could buy on AT&T)

    I’m not defending iPhone here, i’m just pointing out that interpreting the FCC ruling the way that article suggests could easily be construed as resulting in a situation where you’re breaching the letter of the ruling to try to satisfy the intent.

    Your “Your customers can still use your sim, you just can’t subsidise their closed phone” might work as an argument. Alternatively they could just rule that any closed phone isn’t a valid 4G phone on the basis that it must satisfy the openness constraint. However that imposes a requirement for open platforms which the FCC Ruling specifically declined to do (at this time). So i doubt they’d go there.

    At any rate, IANAL and i don’t want to pretend that I really know what any courts are doing. It just seems to me that interpreting it that way is taking the ruling way further than it was intended. Which just to make explicit, the stated intent, as i read it, is to stop the provider strong-arming a manufacturer into crippling functionality in the phone, be that software (VOIP software) or hardware (WI-FI access).

    that Verizon can’t sell a device that operates in this spectrum if the device is locked to Verizon’s network

    But hang on… thats a different story. My reading is that they CAN’T lock it to their network. If I pull the verizon sim out and put a FoonlyFones sim in, that sim has to work(subject of course to standard technical requirements of actually being able to talk on FoonlyFones’ band). This is irrespective of whether the phone has a closed app store or not. (i.e. it would apply to a Google phone as well)

    For example, providers may not “lock” handsets to prevent their transfer from one system to another.

    Source: p89 of the FCC ruling pdf linked from the article.

  65. This is a very one-aspect-only article. What most people care about is whether the system allows them to do more and in an easy way. I can’t speak about Android but WP7 left iOS in the dust in many ways (WiFi sync, notifications, retroactive autocorrection, Zune Pass, live tiles, hubs…). The OS is stable and does what people need so not sure what exactly you are so angry about. From what I hear, even the development tools are better than on Android (is it all Java?) and iOS.

  66. Just found thess articles from July 2010. Only for entertainment.

    What a difference 7 months make (not even 3 quarters). For those who still want to say that only a clairvoyant could have predicted WP7 dismal performance, and not Nokia. Obviously, “we” (at least I do) think Android alone will sell more than 200M phones in 2011. So even if MS sell 5 million handsets of WP7, they will still have less than 2.5% marketshare.

    Windows Phone 7 sales top five million: Analysts and tech pundits perplexed!
    http://wmpoweruser.com/windows-phone-7-sales-top-five-million-analysts-and-tech-pundits-perplexed/

    In what will likely be remembered as one of the biggest comebacks of this century, Microsoft just released the latest sales figures of their new mobile OS touting 5 million devices sold worldwide for the year ending 2010. This comes as a surprise to many as WP7 as it is now known just launched a mere two months ago in October amid prognostications of failure and wild speculations of Microsoft leaving the mobile space by the end of 2011.

    Yet another ‘expert’ analyst predicts Microsoft’s demise in the mobile market
    http://www.wpcentral.com/yet-another-expert-analyst-predicts-microsoft-s-demise-mobile-market

    In an “article” [by Gerson Lehrman Group] (we use the term loosely here) called “Microsoft’s Mobile Demise” who’s main prediction is that “It is highly likely that after the KIN fiasco, Microsoft will exit the mobile OS space within the next year.”, it is put forth that Microsoft blew their cred with KIN and that Windows Phone 7 just won’t deliver. Full of assumptions, lack of context, splash of straw man and no data and you have yourself some headline making material that lacks any substance.

    And now those other futurists who actually DID get it (almsot) right in July. WP7 was the disaster they predicted. But MS simply cannot give up without losing everything. I think many readers of this blog still think MS will leave the mobile market by the end of this year.

    Microsoft’s Mobile Demise
    http://www.glgroup.com/News/Microsofts-Mobile-Demise-49396.html

    It is highly likely that after the KIN fiasco, Microsoft will exit the mobile OS space within the next year.

  67. #From what I hear, even the development tools are better than on Android (is it all Java?) and iOS.

    WP7 dev sucks. Combine the pain in the ass of Apple’s app store curation with the broken-ass crippleness of the old J2ME shit sandwich and you have something of an approximant. Whole swaths of phone functionality lie inaccessible. (You can’t use sockets, ffs!)

    If Microsoft would uncripple their APIs and provide something akin to a full .NET protected environment for phone apps, they could leverage the best-of-breed Visual Studio dev tools and superior .NET runtime environment and hit a real sweet spot with developers.

    Alas, they screwed the pooch in so many ways.

    Android has a good-enough development environment that works on Windows, Mac, and Linux; extensive and well-documented API coverage; and use of the familiar Java language (though not Java runtime). I really wish they had gone with CLR or something more sensible but what they have is, again, good enough.

    Being open source is a huge win for system integrators though. SoM manufacturers are already supplying Android source ports and builds for their kit.

  68. @Adl
    Nice video. It would be on tooic if we could see such a video for WP7 :-)

    It does visualize what “exponential” actually means.

  69. @JonB:

    … an extremely large reseller is never actually just a reseller …

    Right and therein is the edge case that would be hard to prove.

    Google’s position is (or was in 2008) obviously, that it’s up to the carrier to prove they’re selling open devices.

    Actually you raise a fairly easy test that would exonerate Verizon. Is the device sold by Verizon more locked down than equivalent devices(in this case other iPhones) sold by other providers?

    Consider the case where Verizon slips Apple some moolah to make a cheaper version of iPhone that is locked to Verizon’s network. What happens then? With so many other financial considerations between the large players, how do you even prove this?

    Alternatively they could just rule that any closed phone isn’t a valid 4G phone on the basis that it must satisfy the openness constraint. However that imposes a requirement for open platforms which the FCC Ruling specifically declined to do (at this time). So i doubt they’d go there.

    I agree, and I think Google does too. Whether the letter of the law fully supports it or not, I think Google’s trying to draw a distinction between Verizon having to allow any device to attach to its network, and Verizon not being allowed to itself ship a device that is crippled in any way (including being tied to Apple’s store).

    It just seems to me that interpreting it that way is taking the ruling way further than it was intended.

    Entirely possible, but that appears to be Google’s stance, which is entirely consistent with relegating Verizon to bit-hauler status.

    Which [intent] just to make explicit, the stated intent, as i read it, is to stop the provider strong-arming a manufacturer into crippling functionality in the phone, be that software (VOIP software) or hardware (WI-FI access).

    Right, but google may have a point (seems a bit of a stretch, I agree) based on congressional record, etc. that Verizon really, truly not allowed to ship a phone that works on their 4G network and comes with a walled-off app store.

    that Verizon can’t sell a device that operates in this spectrum if the device is locked to Verizon’s network

    But hang on… thats a different story.

    Yes, I was relating that this is one consequence of my own reading of the regulations, which perhaps isn’t quite as agressive as google’s reading.

    My reading is that they CAN’T lock it to their network. If I pull the verizon sim out and put a FoonlyFones sim in, that sim has to work(subject of course to standard technical requirements of actually being able to talk on FoonlyFones’ band).

    Right, but… Verizon is the only carrier that has those particular repurposed TV channels in the entire country. Microsoft has showed us how technical requirements could easily, for example, make it impossible to create a phone that would talk on the adjacent band cost-effectively. My original point in mentioning this is that, absent the rules, if Verizon wanted to make mischief, they could make it expensive for people to switch by making/selling devices that just don’t work on other networks. The rules seem to clearly prohibit this sort of behavior, however.

  70. @ADL

    The world is on fire for Android, especially after the Galaxy release!

    Nifty piece of software. Seems very useful for all sorts of multidimensional demographic data (place, time, and some third, interesting variable).

  71. Some Guy: Odds that RIM would license QNX to MS to compete with RIM? Zero.

    Beyond that, I’m with Mark Erikson and the first comment; yes, the CE UX/shell are horrible*, but the underlying kernel/core’s badness is not obvious.

    And I’ve actually played with a WP7 phone, and the UI/UX layer was top-notch.

    (* I’ve owned two or three handhelds running various version of CE or Windows Mobile. They mostly sucked ass.)

    (There was an implication in the parent post that “the phones got bricked on update because of CE” [somehow] … but it’s not clear why that would be relevant. It’s not like bad updates causing bricking are something CE is known for, to my knowledge, or something that other platforms have Just Never Experienced.

    It’s a horrible black eye and might well end up killing WP7′s chances, but… why blame that on the CE kernel?

    Or was that not intended to be implied, and either a result of ambiguous wording or bad reading on my part?)

  72. >There was an implication in the parent post that “the phones got bricked on update because of CE”

    That implication was not intended. Lacking knowledge or reasonable grounds for conjecture, I don’t have a theory about the cause of the bricking.

  73. Consider the case where Verizon slips Apple some moolah to make a cheaper version of iPhone that is locked to Verizon’s network. What happens then? With so many other financial considerations between the large players, how do you even prove this?

    There’s a certain level where, the ruling not withstanding, if a manufacturer wants to shoot itself in the foot like that then surely that’s beyond the FCC’s remit. If a company sells a phone that is technically incapable of talking on any part of the 700mhz band other than Verizons then yes you could argue that Verizon has “locked” the phone without actually locking the phone. However this falls into Caveat Emptor in many ways. So long as they can’t (and don’t) block valid un-crippled devices then let the marketplace determine what’s right.

    which is entirely consistent with relegating Verizon to bit-hauler status.

    Oh I think the ruling is specifically pushing an agenda of relegating 700mhz providers to bit-hauler status. I just don’t think it’s saying(or should be saying) “you can’t sell a phone that has an app store that rejects applications”. It IS saying “you can’t reject a phone because it has an app store which doesn’t reject applications”. The second I have no problem with (more competition between app stores, competition isn’t decided by the bit hauler). The first seems like the FCC trying to decide the competition before it’s been won (even if I approve of the direction it’s deciding). I want Openness to win because it’s better, not because the FCC pulled a fast one and got away with it.

    Right, but… Verizon is the only carrier that has those particular repurposed TV channels in the entire country. Microsoft has showed us how technical requirements could easily, for example, make it impossible to create a phone that would talk on the adjacent band cost-effectively.

    Hmm valid point. My reading of that section (point 224 on page 89 of the pdf) basically says “we’re watching you, you do dodgy stuff and we’re stepping in”. In particular, “the Commission will ensure the sufficient openness of any network management practices and selected technical standards in the event the approach outlined above proves unsatisfactory”. I’m not saying there’s no room for shenanigans (the libertarian thought process assumes you can’t legislate to remove shenanigans anyway right?) but there’s scope for the FCC to step in if Verizon doesn’t do the right thing.

    Something to keep an eye on to be sure.

  74. I don’t have a theory about the cause of the bricking.

    According to this microsoft update the failure rate was 10% and of that 10%, half were because of bad internet connections or lack of space for a backup.

    As usual… mind the sharp spinning sensation while reading that link.

  75. > lack of space for a backup.

    Back it up on Microsoft servers over the air. There would probably be a limited number of different files to back up, so a clever method wouldn’t even use much storage.

    Yours,
    Tom

  76. According to this microsoft update the failure rate was 10% and of that 10%, half were because of bad internet connections or lack of space for a backup.

    That doesn’t explain much.

    If there is a bad internet connection, then the update file should fail to be downloaded in its entirety and then pass checksum validation. It therefore would never be designated as the default for the boot loader.

    If there is no space for a backup, the update should display a message to that effect and abort without damaging the current, known-good-working firmware.

    There is just no good reason why a botched upgrade should brick a device (absent a catastrophic failure of the flash memory chip). It takes so little to design it so that the worst case always allows the previous firmware to be booted as a fallback.

  77. Japanese cellphone manufacturers were caught flat-footed just like Nokia. The difference is some (or maybe most or all) of them now understand they have to go Android.

    Remember articles like these? We were never going to catch up:

    http://www.indianexpress.com/news/our-cellphone-future/13735/
    http://www.nytimes.com/2005/11/17/business/17japancell.html

    A bit later, now that the iPhone was out, the rest of the world might have been ready for the Japanese onslaught:

    http://www.nytimes.com/2009/07/20/technology/20cell.html

    Of course, that never materialized. Now, the Japanese are planning on using Android:

    http://www.nytimes.com/2011/03/03/business/global/03android.html?src=twrhp

    Some of them may be a bit late to the party, but when NEC, Kyocera, Sharp, and Sony Ericsson all realize that (a) they have to field Android to stay competitive nationally, and (b) fielding Android will also let them compete globally (which they haven’t done very well or at all in smartphones), it really shows just how far out in left field Nokia is.

    BTW, I think all those brands have instant name recognition in the US…

  78. >(b) fielding Android will also let them compete globally

    With the obvious consequence of exerting some serious price pressure on the established players. Going to be fun…

  79. Verizon is the only carrier that has those particular repurposed TV channels in the entire country.

    Not quite true – Verizon is the only carrier that has a nationwide license in those bands (The Block-C license per Wikipedia). ATT has licenses in the 700 mHz band, as does some other regional carriers. ATT, in fact, just took some 700 mHz spectrum off Qualcomm’s hands (still subject to regulator approval, IIRC) – the old UHF channel 55 & 56.

    Verizon and AT&T were the big winners. For $9.9 billion, Verizon won 109 licenses nationwide in the A, B and C blocks; each C-block license covering a huge geographic area. AT&T won 227 B-block licenses for $6.6 billion. Dish Network won 168 E-block licenses for $712 million, while competitor DirecTV did not bid at all, and King Street Wireless picked up 152 licenses in various A and B blocks. The remaining licenses were won by several other organizations including Paul Allen’s Vulcan Spectrum LLC, which secured A-block licenses in Oregon and Washington State.

    Read more: http://www.answers.com/topic/700-mhz-spectrum#ixzz1FPK9uBlB

    http://www.cellularmaps.com/700_auction.shtml has some instructional maps.

    Suffice it to say that based on this, I doubt Verizon will be the only carrier offering an iPhone capable of operating using the LTE system in the 700 mHz bands. It might be that Google may be able to force them not to offer it, but ATT would be foolish as hell to make much effort to prevent people from buying service from Verizon to operate a phone acquired elsewhere with.

  80. Verizon is the only carrier that has those particular repurposed TV channels in the entire country.

    Not quite true – Verizon is the only carrier that has a nationwide license in those bands (The Block-C license per Wikipedia).

    The context was a discussion about whether Verizon could “accidentally” have a handset made that only covered the nationwide spectrum that Verizon now owns in the C-block, e.g. not the whole band. That handset could be a valuable commodity for customer lock-in.

  81. After such an interesting and looooong discussion about the legal struggles of providers inside the US, here a few old reminders why the future of smartphones is so important. Who wins the race to the next mobile phone standard, wins the world, literally.

    Mobile phones in use worldwide top 5.0 billion: study
    http://www.physorg.com/news198405924.html

    The world’s 5.0-billionth mobile phone subscription was recorded on July 8, the company said in a statement, with the market having increased almost seven-fold in 10 years.

    “In the year 2000, about 720 million people had mobile subscriptions, less than the amount of users in China alone today,” Ericsson said.

    How Many Cell Phones in the World
    http://www.vagabondjourney.com/travelogue/how-many-cell-phones-in-the-world/

    As of 2009, 60% of the humans on earth are reported by the UN to possess a cellular telephone, 75% of which are in developing countries. There are now 5 billion cell phone subscriptions current active in the world. Just ten years ago, this number was, relatively speaking, next to nil. This technology has not only proliferated itself evenly over the face of the globe, but it has inveigle itself deeply within the cultural matrix of almost every country on the planet. Without regard for race, culture, background, or even location, the cell phone has become the prime medium through which people communicate on earth.

    Cell phone subscriptions to hit 5 billion globally
    http://reviews.cnet.com/8301-13970_7-10454065-78.html

    Along with the surge in cell phones, demand for mobile access to the Internet has skyrocketed. The ITU expects the number of mobile broadband subscriptions to surpass 1 billion around the world this year, a leap from 600 million at the end of 2009. The organization predicts that within the next five years, more people will hop onto the Web from laptops and mobile gadgets than from desktop computers.

  82. MS/Nokia is betting Nokia WP7 to be the third horse in the race. As usual, they are lying to us, and possibly to themselves.

    Nokia CEO makes the case for Microsoft deal
    http://reviews.cnet.com/8301-13970_7-20031686-78.html

    They are lying, because they really know there never is a third horse in the race. Their marketeers know, because that was already described in an influential book in 1994. So when setting out a product strategy to become the third horse, they knew there was no such position. So we can ask why they did it anyhow. I think the answer is in my previous post: Who wins the smartphone race wins the race to serve the world’s computing needs. The day MS gives up, is the day their decline and fall will be prominently visible and their share prices will fall down to represent only their net value of their cash value.

    They violate quite a number of the “The 22 Immutable Laws of Marketing” (Al Ries and Jack Trout, HarperCollins, New York, 1994).
    http://www.bizsum.com/articles/art_the-22-immutable-laws-of-marketing.php
    http://www.1000ventures.com/business_guide/marketing_22laws.html

    Especially law 8. the law of duality – in the long run, every market becomes a two-horse race.

    Some others:
    1 It’s better to be the first than it is to be better.

    2. the law of the category – if you can’t be first in a category, set up a new category you can be first in. (Apple is good at this)

    3. the law of the mind – it’s better to be first in the mind than to be first in the marketplace

    MS are struggling to get the others right too.

  83. The context was a discussion about whether Verizon could “accidentally” have a handset made that only covered the nationwide spectrum that Verizon now owns in the C-block, e.g. not the whole band. That handset could be a valuable commodity for customer lock-in.

    I didn’t call it out, but Verizon owns noticable amounts of of A- and B-block licenses as well. A C-Block-only phone is locked out of additional bandwidth in major markets – which would make the device perform worse in those markets than devices with access to the entire band, as well as being unable to roam on other network’s 700 mHz spectrum. If I may get a little historical: Sprint, whose “native” coverage was entirely in the 1900 mHz PCS bands on a nationwide license, and who ordered phones from a vendor that only sold them to Sprint, still had their phones built with 850 mHz capabilities as early as 2003 (when I bought one). I’m not saying that Sprint’s decision-making process at the time is like Verizon’s today, but it is instructive to note that potential performance drives business decisions.

    More importantly, lets see if Apple is willing to continue building Verizon a separate device. They are famously unwilling to maintain a large number of SKUs; it was one of the arguments against the CDMA iPhone, no? And that’s the only handset that matters.

  84. That doesn’t explain much.

    Sure it does! It explains volumes about the incompetence of Microsoft demonstrated by the update program!

    If there is a bad internet connection, then the update file should fail to be downloaded in its entirety and then pass checksum validation. It therefore would never be designated as the default for the boot loader.

    Unless the update program either A) doesn’t do any such checksum validation (bad), or B) starts overwriting the firmware before the download has even finished (implies A).

    Don’t laugh. I’ve seen actual devices with upgrade systems that are this brain damaged.

    If there is no space for a backup, the update should display a message to that effect and abort without damaging the current, known-good-working firmware.

    You’d think that program would grab the download size from the HTTP server it’s downloading from and then compare that against the available storage space, refusing to download the update if former < the latter. You'd think that because if you were writing such an update program, that’s what you’d do. But apparently this isn’t the case.

    So there we have it.

    Either A) Microsoft PR is lying or B) Microsoft’s crack development staff is really that incompetent. (Note that this is not an exclusive or, either) I don’t know which one I find more disturbing.

  85. Morgan, I could certainly believe that MS is incompetent enough to neglect all the (obvious) things that you mentioned.

    I will give another example by Nokia. Nokia requires digital signatures on Symbian OS applications that access “sensitive” APIs of the Symbian OS, say, audio and network control.

    It is not that hard to have your application signed, in fact most of the requirements are pretty trivial (such as “the application must not interfere with emergency calls”). The signing is done by so-called “Test Houses”, which are contracted by Nokia. There used to be 3 of them, now there is just one.

    Now, several years ago, a specific software was signed by a test house, whose designers described it as a backup manager. It wasn’t. Actually, it was a spy program, and now it is called FlexiSpy and openly marketed as such.

    But the signature can’t be revoked! Once done, there is no way in Nokia’s Series 60 phones to revoke “bad” applications, even though firmware updates are done quite often.

    They just freakingly forgot that option when they designed the signing system.

  86. >So there we have it.
    >Either A) Microsoft PR is lying or B) Microsoft’s crack development staff is really that incompetent.
    >(Note that this is not an exclusive or, either) I don’t know which one I find more disturbing.

    Alternative C) We’re reading the statement wrong. While our discussion here is particularly about the bricking of the phone, the statement from Microsoft appears to be a more general one about all the failures being experienced, not just the brick. The way I read the statement, 90% of people updated with no problems. Of the remaining 10% “nearly half” experienced a failure in the form of not having enough space, or not having a good internet connection. That doesn’t imply that those people experienced bricked phones, just that the update failed. Incidentally, when I update a linux machine and the updated gives back “Not enough space” in the error message, that’s still a failed update, even if the device is still in perfect working order after the fact.

    It seems to me, what Microsoft is saying is somewhere between 5 and 7% of users experienced a failed update of some sort other than not enough space or a bad internet connection. Still not great, but that doesn’t say anything about how many of that 5% actually got a bricked phone vs some other failure, and it doesn’t say anything at all about whether or not they do proper programing a check the checksums and space before they start writing.

    I never thought I’d find myself defending Microsoft, but it appears the blood in the water might be turning off some critical reading skills in some folks.

  87. Back it up on Microsoft servers over the air. There would probably be a limited number of different files to back up, so a clever method wouldn’t even use much storage.

    Actually its worse than that because this is working from physical machines (PC or Mac, possibly laptops). So apparantly some portion of 5% of people with WM7 phones don’t have enough space on their desktop/laptop machines to backup a smartphone…

    Either A) Microsoft PR is lying or B) Microsoft’s crack development staff is really that incompetent.

    Maybe i’m just cynical but my mind immediately accepted the hypothesis that MS is lying and then went through the thought of “if thats what they’re willing to admit, what the hell is so bad that willingly admitting such incompetence is preferrable”.

  88. In other news:

    Verizon will pull unlimited iPhone data plans
    http://www.digitaltrends.com/mobile/verizon-will-pull-unlimited-iphone-data-plans/

    That was quick. Shortly after introducing the iPhone, Verizon will cut out its unlimited plan and force users into tiered pricing.

    Less than a month after introducing the iPhone, Verizon announced it will be pulling its $30 a month unlimited data plans for the Apple smartphone by summer. Reuters reports that at an investor meeting today, Verizon CFO Fran Shammo claimed that it simply is “not a long-term solution” for the phone contracts. Tiered pricing will increase Verizon’s profits.

    As could have been expected, Shammo admitted that the unlimited data plan was originally implemented to lure in new users for the iPhone’s launch, and that this summer it will follow in AT&T’s steps and begin tiered pricing. AT&T was reportedly mulling reinstating its former unlimited plan for the iPhone when Verizon announced it would be ending its rival’s exclusivity with Apple, but there’s been no further comment on it. AT&T offered unlimited for three years, pulling the option in July 2010 after insisting a limited number of users were hogging bandwidth on the infinite data plan.

  89. @tmoney:

    It seems that Microsoft is being purposefully vague about the number of users with bricked phones. The only number they will commit to is a “small number”. No one said that the “over half” of the failed updates were bricked, but they didn’t say they weren’t bricked, either. Nor did they say what percentage of phones were bricked. Nor did they say anything about what caused those phones to be bricked.

    The statements from Microsoft were very carefully worded. That’s all.

  90. It seems to me, what Microsoft is saying is somewhere between 5 and 7% of users experienced a failed update of some sort other than not enough space or a bad internet connection.

    So the “10% bricked” figure is bogus?

  91. >The statements from Microsoft were very carefully worded. That’s all.

    Of course they are. Ask anyone who does customer service and they’ll tell you the moment you make an official announcement about any sort of problem, suddenly everyone is having that problem whether they are or not. Hell, just ask Toyota how that works out. If I were Microsoft, I wouldn’t talk about bricking either until I had a specific list of models, serial numbers, carriers and computers or some combination there of to have a nice limited set to talk about.

    >So the “10% bricked” figure is bogus?

    I would bet dollars to donuts it is. Of course, I’m one of those people that thinks the word “bricked” gets thrown around too often. Bricked means unrecoverable without new hardware or some sooper sekret device from the manufacturer. If on the other hand it just means you have to go into sooper sekret reload mode for your phone and restore from there, that isn’t bricked anymore than screwing up your grub configuration means your computer is bricked.

  92. @tmoney:

    Often for embedded devices “bricked” means unrecoverable without a JTAG. Since most customers aren’t going to be bright enough to even open up the phone, let alone manufacture the usually custom JTAG cable they’re going to require to unbrick the phone, put the phone into “recovery mode”, or configure the necessary TFTP software on a PC, Unix box, whatever.

    These days there are relatively few devices that don’t have some failsafe method of recovery, it’s just that you need a technician (or a hacker ;) to do it for you.

  93. I’m fairly certain I said that requiring specialized hardware to recover would count as “bricked” in my book, but setting up a piece of software or going into a recovery mode is not “bricked” to me. Most consumers aren’t bright enough to fix a boot problem on a computer either, but we don’t call that bricked.

  94. Since most customers aren’t going to be bright enough to … put the phone into “recovery mode”

    There is simply no excuse for the device to not be easily put into “recovery mode” by holding down a particular button while powering on. This should be documented in the back of the user guide under “In case your phone won’t power on at all”.

    And the button-press may not even be required. If the phone is powered up while attached to a USB cable with a computer on the other end, the phone could prompt whether to “connect to this computer?” or continue with normal firmware load.

    Once it’s connected to the computer, it should present the main filesystem (preferably NOT including the boot loader, which would be kept hidden unless Super Secret Recovery Mode were engaged) as a disk drive, to which the firmware updater would be able to rewrite the newest firmware that caused the ‘brick’ (being able to determine that it has yet to be marked as known-good) and resolve most such instances.

    Heck, it might not even be a bad idea to have a way to put a micro-SD card in the phone containing a specially-named file that the phone could boot from even if the onboard flash is hosed somehow.

    This is all just me spit-balling on how a boot loader could be as robust as possible and preserve the ability to get an alternate firmware file loaded some way under the worst-case scenarios. The people who are paid to design this stuff should do at least as well.

  95. There is simply no excuse for the device to not be easily put into “recovery mode” by holding down a particular button while powering on.

    Sure there is: deterring the loading of unauthorized firmwares.

    “Bricking” is a useful deterrent that prevents even technically competent people from loading their own FW if they are timid enough. Most manufacturers have a vested interest in ensuring that only OFW is loaded. Of course, if your own OFW is broken, it’s egg-on-face time. But Microsoft can just pass the buck to handset makers; after all, 99% of the time it’s not Windows, it’s the driver.

  96. Interesting article on Apple’s costs relative to its competitors:

    http://www.businessweek.com/news/2011-03-02/apple-has-edge-on-tablet-rivals-with-ipad-costs-report-says.html

    Bottom line: (According to the article) Apple has fine-tuned its supply chain, and has lower selling costs than its rivals, and “The company is also pricing iPad more aggressively than its other products …”

    Reading between the lines, it appears Apple may have decided to draw its Maginot line in the tablet space rather than the cell phone space. Certainly they appear to occupy higher ground there than in cellphones right now.

    It will be interesting to see how this plays out. Everybody wants a cut of the huge component market, and they won’t all be able to sell through Apple. Expect really cheap Android tablets sooner rather than later. I could be wrong, but I can’t see this playing out like the iPod market, especially if Google convinces the music companies they really need another competitive sales channel.

  97. >I could be wrong, but I can’t see this playing out like the iPod market, especially if Google convinces the music companies they really need another competitive sales channel.

    Easy sell, I think. It’s no secret that the record companies and studios are very unhappy with the degree of leverage Apple has over them via iTunes. They’re past masters of rent-collecting by sitting on a chokepoint, and they’re not happy about someone else doing it to them.

    If Google can dissemble sufficiently well about the inevitable downstream consequences (e.g, any streaming DRM being cracked within months or weeks of the Android port) they’ll bite. Google has to be retaining at least a few biz-dev types who can lie sufficiently smoothly, and perhaps even a few ignorant enough to think they’re telling the truth.

  98. While you scamps are off unblocking your phones, the rest of us are writing software.

    The boomers are finally dying, and with them, the command line will finally die. (and good riddance to both.)

    Computers are about creation, analysis, and playing about. The old ways are passé, and moot. Die aging hacker, die!

    Witness that with the ascendancy of smart phones and tablets, GPSD solves a problem nobody has now.

    Ncurses? Dead now that CRTs are dead.

    Popclient^Wfetchmail?: Who has anything but a trivial use for this canard?

    As for Linux, it has been reduced to a mere kernel with the rise of Android.
    Android is the new distro, the new hotness. Linux as a desktop has failed, Maemo failed, and Android will be crushed by IP lawsuits. Google will be forced to retreat to the browser as ChromeOS, but having tripped, few will trust Google’s strength.

    If Android is so great, how is it that Apple can ship for a year and introduce it’s second generation iPad while you can’t find Android tablets in the stores yet?

  99. >Witness that with the ascendancy of smart phones and tablets, GPSD solves a problem nobody has now.

    Your rant is strictly ignorance talking.

    Take GPSD as an example. GPS receiver chipsets report packet protocols over over TTL-level serial interfaces; even if you mount them on a tablet or smartphone motherboard and the packets never go to an external port, something still has to interpret the take. And, in fact, several cellphones and tablet devices that I know about (along with probably many more that I don’t) use GPSD internally.

  100. >It’s no secret that the record companies and studios are very unhappy with the degree of leverage Apple has over them via iTunes.

    They sure do like the money, though. iTunes was responsible for a huge increase in their back-catalog sales.

  101. >They sure do like the money, though. iTunes was responsible for a huge increase in their back-catalog sales.

    But the fact that Apple effectively owns the customer is surely not giving them a warm, fuzzy feeling.

  102. >>Witness that with the ascendancy of smart phones and tablets, GPSD solves a problem nobody has now.

    >Your rant is strictly ignorance talking.

    Could be wrong, but I figured his tongue was so far into his cheek that he had to brace on the wall to keep from tipping over.

  103. > But the fact that Apple effectively owns the customer is surely not giving them a warm, fuzzy feeling.

    More to the point, they have to be absolutely drooling over the sheer number of Android devices out there.

  104. >>> As for Linux, it has been reduced to a mere kernel with the rise of Android.
    >>Your rant is strictly ignorance talking.
    > I figured his tongue was so far into his cheek

    or his head was up his ass…

    In other news, distros now at each other’s throats. http://lwn.net/Articles/430098/

    (so maybe he has a point? dunno.)

  105. Artists do not need record companies, they need and audience. However, record companies need to keep artists and their audience separated or else they have no income. Any music market that would allow artists to sell their music without the intermediation of the record companies would kill these companies.

    So the record companies are between a rock and a hard place. Because in such a market, the most consolidated level will cream all the profit. And currently, that would be the online market. But if the record companies foster competition between music markets to de-consolidate them, they would also give an incentive to the online markets to open up directly to the artists. After which the record companies would be de-intermediatized.

    So they hate iTunes as it creams off the profits, but they hate open markets even more, because they would leave them no income at all.

  106. Artists do not need record companies, they need and audience. However, record companies need to keep artists and their audience separated or else they have no income.

    I think thats a bit over general.
    I think you’d find that if record companies disappeared tomorrow, the vast majority of bands would go nowhere for the same reason that the vast majority of VCs composed entirely of hacker geeks tend to go nowhere. it would be far more effective for a band (in this model) to hire someone to handle the administrivia for them (which would include organising studio time, organsing tours and making sure the books balance).

    Given that having administrative staff would take money, record companies would thus, at minimum, become “Musical” VCs in a world of direct commerce from the artists. And thats ignoring the whole collective bargaining, advertising and support for the “lesser talented” band members angle.

    Like the DRM vs Piracy argument, I find this kind of doom and gloom to be suspect at best and outright canard at worst. Not that this stops companies frightened of change from saying it of course.

  107. @JonB
    “Given that having administrative staff would take money, record companies would thus, at minimum, become “Musical” VCs in a world of direct commerce from the artists. And thats ignoring the whole collective bargaining, advertising and support for the “lesser talented” band members angle. ”

    But that is already outsourced by the record companies. A record company does nothing but supply banking services: They invest risk capital in exchange for all profits from sales of music. All other services are outsorced to outside agencies.

    Artists make their money initially from concerts and merchandising, not from “music sales”. Only a happy few who are able to generate excessive sales are able to force some money from the record companies. The same holds for movie studios. Just recently there was an example of the studio accountants managing to book at a $167M loss on a Harry Potter movie grossing almost a billion dollars revenues.
    http://www.deadline.com/2010/07/studio-shame-even-harry-potter-pic-loses-money-because-of-warner-bros-phony-baloney-accounting/
    “Harry Potter and The Order of the Phoenix, the 2007 Warner Bros sequel. Though the film grossed $938.2 million worldwide, the accounting statement below conveys that the film is still over $167 million in the red.”

    An artists agent bureau can easily hire studio, management, and PR services for artists. They do not need the parasitic overhead of the record companies. The only thing that keeps artists from doing just that is the fact that the record companies control all publication (payola) and distribution channels. You will simply not be played on the radio nor be available in the shops or iTunes without the record companies.

  108. Now is the Winter of his discontent made glorious summation:
    > Artists make their money initially from concerts and merchandising, not from “music sales”.

    That’s the received wisdom, true. Many bands find otherwise. It intrigues me that people seem to think that popularity in one is separable from popularity in the other, or that one has much bigger margins than the other. Perhaps it’s because it looks like the venue doesn’t take as much as a label. But a venue also doesn’t pay for your transport, food or accommodation, doesn’t pay for the gear that you blow up doing shows, doesn’t pay for the front of house sound guy you have to have so you don’t sound awful, and yet will also claim a share (25%? 30%? sometimes more) of your merchandising revenue because your table is in their foyer. For smaller-to-medium bands, touring’s expensive, and likely a break-even proposition at best. Make 10% on a record that cost you $30,000 to make and that sells 100,000 internationally, and it looks pretty good compared to spending three months living on chips so that you can say you made a profit that day. I always find strange that people don’t see that most bands are their own VC, both in terms of investment capital and supply of services, and that touring is an investment strategy with not much immediate cash pay-off, a.k.a. building and maintaining an audience.

    > You will simply not be played on the radio nor be available in the shops or iTunes without the record companies.

    Maybe in your town. Depends on what you mean by “the record companies”. In my town, I can tune into breakfast radio and hear records released by record companies you’ve never heard of, because “the record company” is one guy in a rented house. That guy doesn’t look parasitic to me. If he’s nice, he’ll give me some money to make a record and I won’t have to take out another loan.

  109. @Yet Another Darren
    “Make 10% on a record that cost you $30,000 to make and that sells 100,000 internationally, and it looks pretty good compared to spending three months living on chips so that you can say you made a profit that day. ”

    Sounds great. Except that you will not make a penny until the “costs” of the record company are paid out. And these costs, somehow, seem to expand endlessly.

    RIAA Accounting: Why Even Major Label Musicians Rarely Make Money From Album Sales
    http://www.techdirt.com/articles/20100712/23482610186.shtml

    “That report suggests that for every $1,000 sold, the average musician gets $23.40.”

    And in my country, you will not be played on the radio unless you are already very, very popular, or on a contract.

    Payola! (or, How to undermine your own argument to Congress)
    http://arstechnica.com/tech-policy/news/2010/07/payola-or-how-to-undermine-your-own-argument-to-congress.ars

    “Despite their public stances, music labels seem to agree—they just can’t stop paying radio stations to promote certain songs.”

  110. You will simply not be played on the radio nor be available in the shops or iTunes without the record companies.

    Thats not quite true. I can’t say this first hand but accounts i’ve read say that iTunes has a very much alive and well indie section. Sure the labeled section is vastly more popular but if you think about it, i’d suggest thats more diagnostic of why record companies aren’t screwed in a market involving millions of tune stores.

  111. @JonB
    “I can’t say this first hand but accounts i’ve read say that iTunes has a very much alive and well indie section.”

    I have no first hand knowledge, just from the papers and articles, but I understood that it was very difficult to compete on price on iTunes.

    You would expect that unknown bands would compete on price, as authors do in eBooks on Amazon. But that does not seem to materialize in music, except on Free (CC) sites like Jamendo
    http://www.jamendo.com/en/

  112. While you scamps are off unblocking your phones, the rest of us are writing software.

    The rest of who? Just whom do you think the most productive software producers in the world are? I’ll give you a hint: it ain’t Microsoft. Apple remains productive due to its heavy use of open source software. Too bad they want to lock their customers into a world where open source doesn’t matter.

    The boomers are finally dying, and with them, the command line will finally die. (and good riddance to both.)

    Uh, what? One doesn’t follow from the other. I’m no boomer, not by a long stretch of the imagination. I do Unix systems engineering for a living. Believe me, the command line isn’t going anywhere. Even Microsoft has admitted as such with its software. Ever heard of PowerShell? And every Macintosh sold today still comes with a terminal window, BSD command line tools, and the GNU toolchain.

    Computers are about creation, analysis, and playing about. The old ways are passé, and moot. Die aging hacker, die!

    Because hackers have never changed their methods or tools, right? That’s laughable considering the hacker tendency towards neophilia.

  113. >You will simply not be played on the radio nor be available in the shops or iTunes without the record companies.

    There are several “digital distribution” companies that act more like an agent than a traditional “record company” for purposes of dealing with Apple and Amazon and such. For instance, cdbaby lets you upload an album everywhere for 39 bucks upfront (a “handling and setup fee”) and thereafter takes 9% of what Apple pays them, passing the rest on to you. Artists end up making about 60 cents per track or 6 bucks per album, so they don’t have to sell a lot to break even

    There’s also dittomusic, tunecore, and others. Some pass through *all* the royalties but charge a flat “maintenance fee” of, say, $20/album.

  114. @Glen Raphael
    “Artists end up making about 60 cents per track or 6 bucks per album, so they don’t have to sell a lot to break even”

    I am glad that things are not as bleak as I was led to believe.

  115. > I am glad that things are not as bleak as I was led to believe.

    IANAL, err, Band Manager.

    I expect they are still pretty bleak if you are not good at self promotion. Selling 200 copies each of ten tracks at .60 per track is only $1200. With 2000 devoted fans that goes to $12000 for the whole band and their sound guys, etc. You need 20,000 devoted fans before you can live off it, I think. And getting 20,000 devoted fans requires some talent at self promotion and some work I think.

    The power laws of network effects apply to music, right?

    Yours,
    Tom

  116. [url]http://www.infosecurity-us.com/view/16360/droiddream-trojan-is-a-nightmare-for-thousands-of-android-users/[/url]

    Closed source apps from obscure companies running on an open source platform. When you have the source code available to you, it’s definitely easier to find bugs or security holes. Looks like it has been the goodwill of the Linux community and a general aversion towards closed source applications which has prevented Linux machines from getting infected by trojans. This unfortunately, does not apply to Android, and there are groups which would highly benefit from this.

  117. And getting 20,000 devoted fans requires some talent at self promotion and some work I think.

    I was probably bad at stating it but this is kind of what I was getting at. In the most charitable interpretation of the future where the band has all the power, there’s still promotion and administration work to be done and it would be counter-productive to get someone from the band to do that. To me its logical and reasonable that the “record company” entity would naturally handle that side of things rather than waste band members time which would be better spent making more content/doing more promotions.

    I think my original point was that the idea that all the record companies do is act as a monetary gatekeeper between musical artists and audiences just feels naive and overly-simplistic to me.

  118. Winter: Unknown bands don’t really compete on price outside of the CD market (where they still aren’t really “competing”).

    What I mean is, there is no market “for music” in a general sense; it’s not fungible.

    Someone after a Big Label Name (“The Rolling Stones” for old people, “Justin Bieber” for tweens, whatever) is not going to think “hey, I could save a few bucks by buying some other band’s music, or some other genre!” – they want a specific thing most of the time, and either buy it or don’t.

    Unknown bands can, however, compete for sales by the fact of producing a product that is interesting in its own right, or that people might find they prefer to the big-label product.

    At $.99 a song and $9.99 for an album, there’s no need to compete on price (since the price is so low), especially with preview letting you show your quality before you take any money.

    (I won’t take most music even if it’s free. Things I really like, I’ll pay $9.99-14.99 for without qualm. That sort of heuristic is not uncommon.

    Which is why “indie” bands and labels can, if they can get people to know they exist and that they produce a good that is desired, make tolerable money.

    This is far easier with online sales than with physical media; get your troo kvlt back metal album onto iTMS and anyone in the world, more or less, can download it in a few minutes. Put it out on CD and LP and people have to mailorder or live in one or two cities to have any chance of acquiring it, assuming they can ever find out about … especially with the typical “limited edition” which tends to sell out before most of your prospective purchasers ever hear about it.

    This also applies to Amazon’s store and any other competitors, naturally.)

  119. He’s got the jack frost blues:
    > Sounds great. Except that you will not make a penny until the “costs” of the record company are paid out. And these costs, somehow, seem to expand endlessly.

    I honestly don’t know what you think you’re trying to educate me on, or why. Do you really think that the people who actually make records don’t know that most record companies are run by suits who try to skin them alive? The simple fact is this: you don’t have to sign to BMG to make a record, and you never did. The fact that computer programmers have recently taken up an interest in the music industry and now seek to instruct on the horrors that are major label deals says a lot more about what those computer programmers have been listening to since 1975 than it does about what the music industry is like.

  120. @YAD
    “Do you really think that the people who actually make records don’t know that most record companies are run by suits who try to skin them alive?”

    Nope, I was trying to explain why I think the “Record Labels” are not interested in any online music market that allows free access to artists to sell their work directly to their audience.

    Any information I have about the Music Industry is from performing artists who are quite happy to explain why they never ever make money on music sales and why you sign a contract because there is no other way to get access to the distribution channels or radio play lists. And you do that while you know that they might simply use that contract to prevent you from publishing music (or even performing the music) that competes with some of their established artists.

    Personally, I think there is a scarcity of brands, but no scarcity of bands. So I think that there is little future for musicians who want to make big bugs. Too many contenders.

  121. MS does not believe in their own WP7

    Microsoft’s new ‘Bing deals’ fails to support Windows Phone 7
    Android: Check. iOS: Check. WP7: Hang up, caller
    http://www.theregister.co.uk/2011/03/04/bing_deals/

    “Being an ex MSFT share holder and former employee… this is really disappointing. Great feature and I’m sure it rocks, but I’ll only get to use it if I ditch my WP7 device and move to Android. Is that what you’re telling me?”

  122. And the Windows Tablet is a loooooonnnnnggggg way off. No real contender I think.

    Microsoft tablet OS to see light of day in ‘autumn 2012′
    Ballmer does the three-toed sloth dance
    http://www.theregister.co.uk/2011/03/04/microsoft_tablet_2012/

    While Apple and, to a lesser extent, Google are forging ahead with their fondleslab strategies, Microsoft apparently plans to wait until the 2012 back-to-school season before hitting the release button, according to the news agency.

    Citing people with knowledge of the plans, Bloomberg reported that public testing of Windows 8 would get underway towards the end of this year, which roughly fits in with Microsoft’s typical operating system roadmap.

  123. Closed source apps from obscure companies running on an open source platform. When you have the source code available to you, it’s definitely easier to find bugs or security holes.

    How does the appearance of DroidDream demonstrate this?

    Looks like it has been the goodwill of the Linux community and a general aversion towards closed source applications which has prevented Linux machines from getting infected by trojans.

    Not at all. Android is more suspectible to malware and trojans because its security model does, in fact, leave much to be desired, a point that I had argued against with Jeff Read some time ago, but that I’ve since come around to accepting as true, even before the appearance of any Android malware.

    In closer examination of the system, the SD card mounted under /mnt/sdcard is a huge problem. It uses the VFAT file system, which lacks proper POSIX permissions and furthermore allows the execution of binaries mounted there. Any application can read or write all files written there.

    Furthermore, there is no way to fine-tune permissions: you have to accept or deny all permissions for each application you install. And many applications seem to request too many permissions.

    It does seem that the use of the VFAT filesystem is Android’s Achilles’ Heel. This is avoidable, however, if one were to root the phone and format SD card as an ext2 or ext3 filesystem.

  124. And the Windows Tablet is a loooooonnnnnggggg way off. No real contender I think.

    Um… don’t want to rain on your parade but Windows tablet has been around for a good 5-10 years. And people can say “but it was windows” (and in some cases i’d agree) but the reason why it flopped wasn’t because the OS was bad(it wasn’t any worse than windows normally is) but no-one could think of a reason why they would want one.

    This is the reason why you have to give some form of props to the apple RDF. it literally re-animated a product category that previously no-one wanted.

  125. In closer examination of the system, the SD card mounted under /mnt/sdcard is a huge problem. It uses the VFAT file system, which lacks proper POSIX permissions and furthermore allows the execution of binaries mounted there. Any application can read or write all files written there.

    I’m confused.
    If i have permission to run “./trojan” under vfat then surely i’ve got permission to run “chmod a+x trojan;./trojan” under ext3. Or if they don’t allow chmod there’s always File.setExecutable .

    And even if they can run the file, how is that different from running a file under any other linux system? One would hope that phones don’t use kernel modules, and we’re running as a less privileged user. What can that file do thats dangerous?

  126. @JonB
    “Um… don’t want to rain on your parade but Windows tablet has been around for a good 5-10 years.”

    Yes, for Intel.

    But people do not seem to want Intel based tablets. They want ARM based tablets. And Windows on ARM, that is WinCE or WP7.

    No Windows tablet here I think.

  127. In closer examination of the system, the SD card mounted under /mnt/sdcard is a huge problem. It uses the VFAT file system, which lacks proper POSIX permissions and furthermore allows the execution of binaries mounted there. Any application can read or write all files written there.

    You can’t execute stuff from a filesystem mounted noexec. And you can have proper POSIX permissions on a vfat filesystem; it’s just that the entire filesystem gets the same permissions at mount time.

    And not all applications can write to the SD card. Write access is afforded only to members of a certain group which means that only apps which the user has granted write access to can write to the SD. (Each app gets its own username app_nn which it runs as.)

    In principle, the security system in place is designed to restrict permissions on a per-app basis to the SD card and other system components. In practice, however, Android is the most vulnerable to malware mobile platform.

  128. Actually, this should go in the Bricking article. But I think you want to read this too

    Microsoft blows Windows Phone update, again
    Don’t touch that Zune
    http://www.theregister.co.uk/2011/03/04/windows_update/

    Samsung users who held off updating after hearing about the problems last time are being told to hold off again as the fixed fix isn’t really fixed at all.

  129. Android security = users value usability over security. Hence a file system that can be read to and written by anyone and their pet dog’s OS; and consequently has no real security. Sideloading files by mounting as a SUB mass storage device is convenient as hell.

    OR was I supposed to be surprised by a dog bites man story?

  130. Ian Argent, as Jay Maynard indicated, if sideloading via USB mount were actually more convenient and preferable the iPod and iTunes wouldn’t hae dominated as thoroughly as they have.

    Once again, fandroids miss the point by viewing the world through geek-colored glasses.

  131. About Android security: What you want is to protect the phone against its owner. Because the owner can install anything this is a losing strategy. They solved that in the OLPC laptop’s Bitfrost with a layered defense. Children could install anything, but had to jump through some hoops of technical insight to get there.
    http://wiki.laptop.org/go/Bitfrost
    No lockdown
    Though in their default settings, the laptop’s security systems may impose various prohibitions on the user’s actions, there must exist a way for these security systems to be disabled. When that is the case, the machine will grant the user complete control.

    In more Android news

    Google’s Android Spurs More App Jobs Than iPhone
    http://www.businessweek.com/technology/content/mar2011/tc2011033_078030.htm

    Software engineers skilled in making applications for Google’s (GOOG) Android mobile-operating system this week became more sought-after than developers for Apple’s (AAPL) iPhone, according to online job board Dice.com.

    Employers requested experience or skills with Android in 987 job postings on Dice as of Mar. 1, more than the 970 jobs asking for iPhone expertise, Bloomberg Businessweek.com reported today. The number of available positions mentioning either Android or iPhone surged more than threefold from a year ago, when the site listed 273 Android-related jobs and 312 iPhone-related jobs.

  132. @Jeff – Apple is coasting on legacy network effects as far as iTunes “superiority” goes. You have to have a computer via a network connection to activate an iPhone, and I believe an iPad as well. You can pick up an Android and go. (Since both require you to create an account with their mothership, I think it’s a wash, there).

    Apple dominated because of content, not because the iTunes Way is easier. They don’t have a content monopoly any more. If your device behaves like a USB drive, the apps used to load media across don’t have to do any work other than copy files. VFAT means not having to worry about permissions (the way my various NTFS-formatted external drives trip me up when I put them in different machines).

    USB mass storage Just Works, and without having to have a specific vendor’s app on the host machine.

  133. What we all thought is now outed: MS bought a share in Nokia Smart Products for $1B. If they sell 20 million phones, that is $50 subsidy per phone. But Nokia has to pay license fees for every WP7 phone sold.

    Odd. This way, the incentives for Nokia are to sell as few phones as possible. Now the question is whether the loss in share value and the costs associated with the decommissioning of Symbian and Meego are not in excess of $1B.

    Microsoft ‘paid Nokia $1bn’ for WinPho 7 deal
    It’s an ad, ad, ad, ad world

    http://www.theregister.co.uk/2011/03/08/nokia_microsoft_deal/

    Bloomberg reports that Redmond will ship that $1bn to Espoo, Finland, in support of Nokia’s efforts to “promote and develop” smartphones based on Microsoft’s as-yet-unloved mobile operating system in a deal that’s set for “over five years.”

    And speaking of unloved, Bloomberg also notes that Nokia’s shares have shed over a quarter of their value since the announcement of the WinPho 7 deal less than one month ago.
    …….

    In addition, one of the chatty tipsters noted that the payout was a bribe/inducement to keep Nokia from selecting Android as its go-to mobile OS. Looking at it another way, a source said, the choice of WinPho 7 will give Nokia the opportunity to “stand out” from the crowd of Android phones increasingly flooding the market.

  134. >Odd. This way, the incentives for Nokia are to sell as few phones as possible.

    Possibly not. It depends on whether the $1bn is a lump payment up front or delivered over five years in tranches that are explicitly or implicitly tied to sales goals.

  135. @esr
    Not quite clear whether any tranches are sales related, but you might be right and MS might have the upper hand in the deal:

    The sources also said that Nokia will pay Microsoft a licensing fee for each copy of the operating system it installs on its handsets, but that the $1bn-plus payout will begin before any of those fees cross the pond on their way to Redmond.

    …..

    Microsoft’s CFO Peter Klein certainly thinks the deal was worth a cool billion. Speaking at the ISI Annual Conference shortly after the Nokia deal was announced, he said: “Certainly this deal is a long-term, multifaceted deal which has lots of commitments by both companies, and there’s going to be search revenue and there’s going to be royalty [revenue]. There’s going to be joint go-to-market opportunities. And obviously both parties are aligned, and we think this has great long-term financial opportunities.”

    Perhaps so. But as with any such mega-deals between industry heavyweights, implementation is all – and if Bloomberg’s sources are correct, Nokia has a $1bn head start.

  136. >Bloomberg also notes that Nokia’s shares have shed over a quarter of their value since the announcement of the WinPho 7 deal less than one month ago.

    So, that’s a promise of one billion over a couple of years, versus a loss of about ten billion in market valuation.

    Does NOK still have any shareholders who give a shit?

  137. @Some Guy
    “So, that’s a promise of one billion over a couple of years, versus a loss of about ten billion in market valuation.”

    Just maybe, this might not have been planned. :-)

  138. Just for the record, another write up on the deal. Nokia gets, indeed, only part of the money up front. The further decline of MS stocks is added as a bonus. They are still on track for a 2012 breakup I see. There is the obligatory “buy” advice from some analyst.

    Microsoft Is Said to Pay Nokia More Than $1 Billion in Deal
    http://www.businessweek.com/news/2011-03-07/microsoft-is-said-to-pay-nokia-more-than-1-billion-in-deal.html

    Microsoft shareholders want the company to salvage its mobile-software business while also reining in costs. The company doesn’t break out results for its mobile-software unit, and instead groups them with the profitable Xbox video-game business, making it difficult to evaluate the financial performance of phone software.

    Chief Executive Officer Steve Ballmer has come under pressure from investors and his own board to improve sales of mobile software after the company lost market share to Google and Apple. Microsoft stock has declined 7.8 percent so far this year.

  139. More interesting news: Deutsche Telecom, exit stage left.

    “Sprint, T-Mobile USA Reportedly Talking Tie-Up: Sprint Nextel (S: 4.68, +0.20, +4.58%) has reportedly held discussions about acquiring Deutsche Telekom’s T-Mobile USA in a deal that would marry the third- and fourth-largest U.S. wireless providers.”

    Looks like assessment of Sprint may be making a move to get out of that “also ran” status they’ve been stuck in for the last decade.

    I’m wondering if the fact that this gives them a much bigger Android play has a lot to do with this (I’d guess a big YES).

  140. @winter:

    > This way, the incentives for Nokia are to sell as few phones as possible.

    That doesn’t follow, unless the royalty rate is set so high that Nokia loses money on every phone.

  141. The company doesn’t break out results for its mobile-software unit, and instead groups them with the profitable Xbox video-game business, making it difficult to evaluate the financial performance of phone software.

    Lawl. Redmond accounting.

  142. Today’s reason to read slashdot…

    http://mobile.slashdot.org/comments.pl?sid=2027146&cid=35417892

    Nokia: We’ve had a good think about it and we’re going to start developing for Android
    Microsoft: What would it take for you to start using and developing for Windows Mobile?
    Nokia: *Has a think* *Pinky moves towards mouth* ONE BILLION DOLLARS!
    Microsoft reps: *look at each other, shrug shoulders* Yeah, OK, I can’t see any reason why we can’t do that..
    Nokia: Err, OK, I guess we’re using Windows Mobile then….

  143. @Patrick Maupin
    “> This way, the incentives for Nokia are to sell as few phones as possible.

    That doesn’t follow, unless the royalty rate is set so high that Nokia loses money on every phone.”

    Reasoning left out:
    1 Nokia get $1B, irrespective of the number of phones they sell
    (probably wrong, but lets assume)
    2 Nokia has to sell incompatible, backward phones no one wanted at a negative premium
    (marketing speak)
    3 Nokia has to pay a license fee for every phone sold, increasing the negative premium
    4 Nokia can drastically cut costs by not selling WP7 phones and increasing income by selling Symbian/Android phones

    The only thing missing were the <SARCASM> tags.

  144. As I wrote above, there is no third horse in marketing. law 8. the law of duality – in the long run, every market becomes a two-horse race.

    And HP is recognizing that too. They try to put WebOS on all their PCs, probably running in VirtualBox. Another company seeing the end of MS coming, and coming fast.

    HP to put a WebOS in every PC
    Platform shootout extends to the desktop
    http://www.theregister.co.uk/2011/03/09/webos_hp/

    It seems more likely that it will be some sort of WebOS execution environment, making WebOS applications run on top of Windows rather than requiring the user to pick which OS to use. WebOS developers currently run a WebOS machine on emulated hardware, using the Oracle-owned VirtualBox. Tighter integration should be possible and would create a nested environment into which users could install safe (signed) WebOS applications, but still resort to Windows apps where necessary.

  145. Interesting data points:

    This game does better financially on WP7 than Android:

    http://www.everythingwm.com/fruit-ninja-over-7-times-more-revenue-on-windows-phone-compared-to-android/2011/03/09/

    This game does better financially on Android than on Apple:

    http://blogs.computerworld.com/17941/android_ios_app_profit

    Both of these data points, of course, go against the conventional wisdom.

    A few possible explanations spring to mind, related to number of available apps in each market (saturation), frugality of users in different markets, method of extracting money (game sales vs. in-game sales vs. advertising), etc.

  146. @Patrick Maupin
    I see that Android had more new apps in February than WP7 has in total (30k vs 10k).

    So competition will be fierce on Android, while there will be a lack of apps on WP7.

  147. >So competition will be fierce on Android, while there will be a lack of apps on WP7.

    Not sure that conclusion is justified. If the willingness of the tiny population of WP7 users to buy pay apps is as incredibly high as Patrick Maupin’s cite suggests, more WP7 apps will be written. App writers only care about userbase size as a proxy foe expected revenue.

    I find the WP7 figures weird and hard to explain, myself. But we have to deal with reality as it is, not as we’d like it to be.

  148. So competition will be fierce on Android, while there will be a lack of apps on WP7.

    If the willingness of the tiny population of WP7 users to buy pay apps is as incredibly high as Patrick Maupin’s cite suggests, more WP7 apps will be written.

    It will certainly be interesting to see if/when any kind of equilibrium point is reached. To the extent that developers notice that particular data point, I would expect that a lot of easy-to-port applications will be ported fairly quickly. That in turn will drive up the total number of apps available to WP7 users, which may help to recalibrate per-app revenue expectations in the WP7 world, and lead to better decision making for developers thinking about hard-to-port or native apps for WP7.

    a priori, I would expect that, if every app had to be crafted from scratch for every platform, the number of apps available should scale, somewhere between logarithmically and linearly, according to the number of end-users on the platform. Conversely, if the total cost of porting was zero, I would expect the number of apps available to converge to identical values across the app stores.

    This, of course, assumes relative constant demographics across platforms. Obviously, Symbian has a lot of really low-cost cellphones, and RIM has a lot of corporately locked down cellphones, so those would be outliers, even if the platforms offered reasonable functionality.

  149. @esr
    “Not sure that conclusion is justified. If the willingness of the tiny population of WP7 users to buy pay apps is as incredibly high as Patrick Maupin’s cite suggests, more WP7 apps will be written. App writers only care about userbase size as a proxy foe expected revenue.”

    Obviously. But in an exponentially growing market, there will be a considerable lag in available apps. Equilibrium is nowhere in sight.

    If we simply assume that each handset owner has the same budget, an Android owner can choose from 150K apps, and iPhone owner even from more (?), but a WP7 owner only from 10K.
    http://blogs.computerworld.com/17941/android_ios_app_profit

    An owner is somewhere up to 15 times more likely to chose a particular WP7 app than the corresponding Android app. USA market share numbers are currently 29% for Android, 27% RIM and Apple, and 10% Windows (mobile + WP7).
    http://www.computerworld.com/s/article/9213258/Android_smartphones_top_iPhones_BlackBerries_for_first_time_
    http://blog.nielsen.com/nielsenwire/online_mobile/who-is-winning-the-u-s-smartphone-battle/

    And the MS numbers are rather disconcerting as their marketshare pre WP7 was almost 11%
    http://www.comscore.com/Press_Events/Press_Releases/2010/12/comScore_Reports_October_2010_U.S._Mobile_Subscriber_Market_Share

    WP7 marketshare is slipping fast, MS total marketshare has now declined to 8%:
    http://www.1800pocketpc.com/2011/03/09/wp7-marketshare-slipping.html
    http://www.theinquirer.net/inquirer/news/2032210/microsofts-marketshare-falls-thanks-wp7

    MS themselves claimed that 2 million WP7 handset had been sold to carriers by January
    http://www.eweekeurope.co.uk/news/questions-remain-over-windows-phone-7-sales-19331
    http://www.eweekeurope.co.uk/news/microsofts-phone-market-share-shows-need-for-wp7-hit-20409

    Given that there were at least 40 million Android handsets out there in January, this means the WP7 market is only 5% of the Android market.

    So if your app has a 15 times bigger chance to be chosen on WP7 than on Android which has a 20 times bigger market, you will effectively sell only 75% as many apps on WP7. So if you can earn only 33% more on each sale on WP7, you still will earn more on WP7.

    Sounds plausible.

  150. Wrt my previous post, I was thinking what the equilibrium size of the app-store would be compared to the market size.

    If we simply take:
    Net-Profit = sales * (Prize – Variable-Cost) – Development-Cost or N = s * (P – V) – D

    We assume that Net profit is equal on both platforms, as are Development costs and Variable costs. Prizes are net prizes after the “commision” of the market has been removed. Sales are simply taken as the probability that the app is bought by the users. Total number of apps bought #TotalBought = (Budget/Prize * #consumers) and the sales of any single app s = #TotalBought/#apps. For simplicity, we assume that the number of apps bought per consumer is constant Budget/Prize=#A

    Compare WP7 market with Android, that is the sales are simply the total number of apps sold divided by the number of different apps (#apps)
    s_WP7 = (#A * #consumers_WP7)/#apps_WP7
    s_And = (#A * #consumers_And)/#apps_And

    Equilibrium for a new app is reached when net-profit is equal on both platforms, that is when (D and V equal)
    s_WP7 * (P_WP7 – V) – D = s_And * (P_And – V) – D

    (#A_WP7 * #consumers_WP7/#apps_WP7) * (P_WP7 – V) = (#A_And * #consumers_And/#apps_And) * (P_And – V)

    #apps_And/#apps_WP7 = [(#A_And * #consumers_And) * (P_And - V)] / [(#A_WP7 * #consumers_WP7) * (P_WP7 - V)]

    Simplifying #A_And = #A_WP7 (consumers buy a certain number of apps, independent of price) and P-V ~ P (V small) we get a first approximation of

    #apps_And/#apps_WP7 = (#consumers_And * P_And) / (#consumers_WP7 * P_WP7)

    This has a large Duh factor.

    Lets make the number of apps bought per consumer prize sensitive #A = Budget/Prize factor, but taking only the average prize.

    #apps_And/#apps_WP7 = [(B/P_And) * #consumers_And * P_And] / [(B/P_WP7) * #consumers_WP7 * P_WP7] = #consumers_And / #consumers_WP7

    That is, if budgets are fixed, the number of available apps will simply scale with the number of consumers. Irrespective of the prize of the apps.

    So, with only a 5% market share, WP7 will only have 5% of the apps of Android.

    Note that the “commission” taken by the market will decrease the number of apps sold, and so decreasing the number of apps available. The only way out for MS is to subsidize development. By decreasing D they can increase the number of apps that are developed. By how much?

    If we look at the first formula and adapt it to the total market

    #apps_WP7 * (s_WP7 * (P_WP7 – V) – D_WP7) = #apps_And * (s_And * (P_And – V) – D_And)

    it is obvious that subsidies will have to be a sizable fraction of total development costs of ALL apps. That is a lot of money.

  151. Sorry for this avalanche of posts. But rereading my previous posts I see they are rather chaotic and unintelligible. Just to summarize:

    Given the numbers from the links posted, if an app is prized only 33% more expensive on WP7 market than on Android, it can still make more profit on WP7.

    The equilibrium number of apps available on a platform scales linearly with the number of consumers, independent of net prize. So if WP7 has only 5% of the market share of Android, it will also have only 5% of the apps. Any commission taken by the market-place decreases the net-prize and therefore decreases the number of apps available.

    If MS would like to increase the number of apps on WP7 with subsidies, they would have to fork out an amount of money that would be comparable to the total development cost of all apps on their platform.

  152. “I learned this morning that Windows Phone 7 – the smartphone OS that bricks your phone! – is a skin over a bunch of core components from Windows CE.”

    [Citation needed]

    (Apologies is someone already posted this; I skimmed the comments and didn’t see anything.)

  153. FORM 20-F
    ANNUAL REPORT PURSUANT TO SECTION 13 OR 15(d) OF THE
    SECURITIES EXCHANGE ACT OF 1934
    For the fiscal year ended December 31, 2010
    Commission file number 1-13202
    Nokia Corporation
    http://www.sec.gov/Archives/edgar/data/924613/000095012311024458/u10545e20vf.htm

    We expect the transition to Windows Phone as our primary smartphone platform to take about two years. While we transition to Windows Phone as our primary smartphone platform, we will continue to leverage our investment in Symbian for the benefit of Nokia, our customers and consumers, as well as developers. We and Microsoft have entered into a non-binding term sheet, and the proposed Microsoft partnership remains subject to the negotiation and execution of definitive agreements.

    Nokia outlines new strategy, introduces new leadership and operational structure

    On February 11, 2011, Nokia outlined its new strategic direction, including changes in leadership and operational structure designed to accelerate the company’s speed of execution in the intensely competitive mobile product market. The main elements of the new strategy includes: plans for a broad strategic partnership with Microsoft to build a new global mobile ecosystem, with Windows Phones serving as Nokia’s primary smartphone platform; a renewed approach to capture volume and value growth to connect “the next billion” to the internet in developing growth markets; focused investments in next-generation disruptive technologies; and a new leadership team and operational structure designed to focus on speed, accountability and results.

    Nokia and Microsoft have entered into a non-binding term sheet, however, the planned partnership with Microsoft remains subject to negotiations and execution of definitive agreements by the parties and there can be no assurances that definite agreements will be entered into. The future impact to Nokia Group’s financial statements resulting from the terms of any definitive agreements will be evaluated once those terms are agreed.

    As of April 1, 2011, Nokia will have a new operational structure, which features two distinct business units in Devices & Services business: Smart Devices and Mobile Phones. They will focus on Nokia’s key business areas: smartphones and mass-market mobile phones. Each unit will have profit-and-loss responsibility and end-to-end accountability for the full consumer experience, including product development, product management and product marketing.

    Starting April 1, 2011, Nokia will present the financial information in line with the new organizational structure and provide financial information for three businesses: Devices & Services, NAVTEQ and Nokia Siemens Networks. Devices & Services will include two business units: Smart Devices and Mobile Phones as well as devices and services other and unallocated items. For IFRS financial reporting purposes, we will have four operating and reportable segments: Smart Devices and Mobile Phones within Devices & Services, NAVTEQ and Nokia Siemens Networks.

  154. From Nokia’s SEC sheet:
    “focused investments in next-generation disruptive technologies;”

    That must be the holy grail for every investor. Sounds a lot like “buying winning lottery tickets” as a business strategy.

  155. Nokia is trying to hold on to its MeeGo developers by offering the a bonus for staying until the end of the year. Not exactly surprising since Elop’s announcement said that the platform was going nowhere.

    http://mynokiablog.com/2011/03/02/nokia-offers-50-bonus-to-meego-developers-if-they-stay-until-years-end-to-deliver-nokia-n950/

    I don’t know how much of all of this was deliberate, but moves like this pretty much ensure that the wrong people will stay to collect the bonus. Nokia’s MeeGo phone looks very much to be dead on arrival, probably by design.

  156. An account from the “inside” why Nokia could not get it’s smartphone UI in order.

    Why Nokia failed: ‘Wasted 2,000 man years’ on UIs that didn’t work
    For want of a nail, the Kingdom was lost?
    http://www.theregister.co.uk/2011/03/10/nokia_ui_saga/

    With both Linux and Symbian platforms, 80 per cent of the code did not need to change to make Nokia competitive once again. With Symbian, the code had been written over many thousands of man-years, and only the top 20 per cent (at most) needed to be refreshed. Yet Nokia couldn’t deliver this. For want of a nail, the kingdom was lost.

  157. In other Mobile Phone OS related news. Not particularly relevant in day to day uses. But watch MS’ response, which can be summarized as “Our next version will be great”.

    Pwn2own: BlackBerry and iPhone Fall, Chrome and Firefox Untouched
    http://www.esecurityplanet.com/news/article.php/3927911/Pwn2own-BlackBerry-and-iPhone-Fall-Chrome-and-Firefox-Untouched.

    Apple’s iPhone and Research in Motion’s BlackBerry were both successfully exploited by security researchers. The iPhone and BlackBerry join Microsoft’s IE and Apple Safari as technologies that researchers were able to exploit. Google’s Android and Chrome, as well as Mozilla’s Firefox, all emerged unscathed.

    ….

    “We have confirmed that IE 9 RC is not affected by the vulnerability used in the Pwn2own contest,” Microsoft Security Response tweeted. “IE 9 officially releases on Monday.”

  158. This is so cool! A dual screen hinged Android phone. This kind of innovation is the good side of fragmentation. I understand what Apple usability partisans are saying when they talk about the beauty of a closed system. It works great for the Wii or XBox 360 or the Playstation too. I get it. But I’m not willing to sacrifice the innovation that comes from an open system.

    Yours,
    Tom

  159. @James and Tom DeGisi
    WP7 == WINCE

    Basically, it is almost inevitable. WINCE is Microsoft’s only ARM OS. And it will be for some time. So there is no alternative.

    I have no experience nor knowledge about WINCE or Windows Phone, so I have no idea what MS did with WINCE to get WP7.

  160. @Jeff Read:

    From the article:

    If Google is proven wrong, pretty much that entire software stack — and also many popular third-party closed-source components such as the Angry Birds game and the Adobe Flash Player — would actually have to be published under the GPL.

    That’s not even wrong. More to the point, if it were right, what’s the fine line that keeps all apps that run on Linux from meeting the same fate?

    In order for anybody to sue google over the headers, that party (not Oracle or some other random third party) would have to (a) want to damage google/Android/Linux and (b) own the copyright on some (more than de minimus) code in the header file that got copied/transmogrified by google’s tool.

    While there’s certainly a small possibility that there are a few copyrightable elements in the header file (such as inline functions that actually have some literary merit, as opposed to simply being obvious functional wrappers on top of underlying kernel functions), the chances that someone who cares actually owns copyright on enough of those to amount to anything seems remote. (The comments in the header file could certainly be copyrightable, but I assume, perhaps wrongly, that the program strips those.) If anybody does own enough to make a stink, a small (from google’s perspective) donation to linux kernel development would probably be all it takes to fix up those guys.

    In any case, google’s answer can legitimately be the same as the entire world’s answer to SCO. Show me the needle in the haystack. Show me what is yours that you think I stole, and I will rewrite it. There is no way that google will be blindsided into any kind of injunction on this. The need to show immediate, irreparable (as in can’t be fixed by a later cash infusion from google) harm will be insurmountable.

    Google could be in serious trouble, but as someone who thought they were completely batshit insane when they bought youtube, I have to say on reflection that google’s copyright lawyers, while aggressive, seem to know what they are doing, and that, unlike youtube, this issue falls squarely within the bounds of almost 20 year old caselaw (Sega v. Accolade).

    BTW, even the FSF wouldn’t touch this one (except for, perhaps, a small negotiated settlement). They don’t want to run the risk of any caselaw that says you can link non-GPL code to GPL code, because that would be harmful to the party line.

  161. >unlike youtube, this issue falls squarely within the bounds of almost 20 year old caselaw (Sega v. Accolade).

    Patrick is correct. Google “scenes a faire” for discussion; Google is probably bulletproof on this one. Even if someone sues them, the threshold for a TRO will not be met.

    I’ve been the lead forensic analyst in an IP case where scenes a faire involving Unix header files was applied as a key defense, so I’m not just speculating about this; I’ve been where the rubber meets the road and contributed to legal briefs addressing a nearly indistinguishable fact pattern.

  162. Thats FM again. He has been shown to be completely wrong every time so far.

    Anyhow. Linus has set explicite borders of what APIs can be linked to (including header files as far as I know). And Linux devs have yet to sue someone over header files. They seem to insist they are not copyrightable. At least that was the opinion when SCO tried the exact same trick.

  163. The original FUD story was not from FM, but from a former attorney of MS (who changed his bio to hide this fact, see Groklaw)
    Edward J. Naughton
    http://www.brownrudnick.com/bio/bio.asp?ID=512&ForwdName=Edward+J.+Naughton

    This was what the original bio looked like (as unearthed by PJ)

    :
    - Co-counsel defending Microsoft against a putative consumer class action alleging that it had violated wiretapping statutes and common law privacy rights by designing Windows to permit third parties to place cookies on computers. Obtained dismissal of complaint….

    - Represented Microsoft in several dozen lawsuits against resellers and corporate end-users of counterfeit, infringing, and unlicensed software.

    SJVN cites some well-known specialists ;-) to debunk the story:

    Does Google’s Android violate Linux’s Copyright?
    http://www.zdnet.com/blog/open-source/does-googles-android-violate-linuxs-copyright/8497

    Now, I am not a lawyer, but Naughton is talking largely about the Linux kernel header files, and back in the days when SCO was fighting with IBM over Linux’s copyrights, the issue of header files came up. At the time, Eric Raymond, then president of the Open Source Initiative (OSI), told me there was a good reason why Unix and Linux’s header files looked the same: “Do you know that there is not one bit of executable code in those files? They’re pretty much all macros and declarations forced by POSIX and other technical standards.” Linus Torvalds also told me at the time that much of the material in the header files was there “because that’s specified by several standards, not Unix per se-you’ll find those error names in any operating system that has a C compiler.”

    ….

    Eben Moglen, the Columbia University law professor who knows more about the GPL and IP law when he’s asleep, then I ever when I’m wide-awake on my best day, told me, “I would say that the issue is a little less complex and a little less dire than it might seem on first acquaintance, while the facts are not quite as simple and therefore the narrative not quite as compelling as one might be led to believe.”

  164. Car informatics around the mobile phone.

    Because it has Nokia in it:

    Has the Battle for the Digital Car Been Won?
    http://www.consortiuminfo.org/standardsblog/article.php?story=20110318094909562

    Instead, what we have evolved towards is a new and far more versatile mobile platform – the smartphone. Rather than becoming some kind of new desktop, the automobile has instead become the equivalent of a dumb terminal. And what the Car Connectivity Consortium is all about is enabling that dumb terminal to interact with its CPU – a mobile device.

  165. And the man who wrote the GPL disagrees with Naughton, FM cs about kernel header file copyrights. Now, the people who wrote the code and the man who wrote the the license disagrees with their interpretation. Now who would be right, FM or those who actually participated?

    Anyhow, we must thank FM for helping getting all the FUD so easily busted.

    Re: Gauntlet Set NOW!
    http://lkml.indiana.edu/hypermail/linux/kernel/0301.1/0362.html

    That’s not the FSF’s view. Our view is that just using structure
    definitions, typedefs, enumeration constants, macros with simple
    bodies, etc., is NOT enough to make a derivative work. It would take
    a substantial amount of code (coming from inline functions or macros
    with substantial bodies) to do that.

  166. @Winter:

    That would be a (welcome) about-face by RMS. For some of his earlier thinking on the topic, see Why CLISP is under GPL.

    However, I doubt it was so much an about-face as a correct reading of his current audience. For example, about a month after the RMS snippet you quote, Eben Moglen responded to questions for a slashdot interview:


    The language or programming paradigm in use doesn’t determine the rules of compliance, nor does whether the GPL’d code has been modified. The situation is no different than the one where your code depends on static or dynamic linking of a GPL’d library, say GNU readline. Your code, in order to operate, must be combined with the GPL’d code, forming a new combined work, which under GPL section 2(b) must be distributed under the terms of the GPL and only the GPL.

    I deconstructed this a few months after it was written (over 7 years ago).

    Basically, RMS’s (and by extension, the FSF’s) schizophrenia is a direct result of the cognitive dissonance arising from the twin goals of wanting to coerce authors into releasing new code under the GPL, while simultaneously wanting users to enjoy “maximum freedom”, never quite realizing that the best authors are, in fact, users with an “itch to scratch.”

  167. One last comment on Linux and the GPL.

    IF there was no userland exception in the Linux kernel license, then it’s entirely possible that people could be in violation, not for distributing the kernel header files, but for distributing the actual kernel itself, in conjunction with other Android code.

    BUT if Linus had taken that tack then, it would be a non-issue at this point in time, because nobody would know what Linux is.

    AND if the kernel team were to attempt to take that tack now (which they most assuredly won’t, in any case, because they don’t believe there is an issue), then laches and equitable estoppel would stop them dead in their tracks.

  168. @Winter:

    And, one last comment on RMS, FSF, and the GPL.

    Anyhow, we must thank FM for helping getting all the FUD so easily busted.

    In one sense, FM is simply channeling old FUD from RMS. There is a good argument to be made that in the 90s, RMS/FSF actually slowed down adoption of free software quite dramatically with the fearmongering about how the copyright worked. Yes, the FUD led to some license successes, like getting CLISP under the GPL umbrella, but those successes were only among people who already believed in free software, and just wanted to get on with programming. The damage done, however, lasted well over a decade.

  169. Sorry, the layout of the previous post was messed up, please remove and use this one

    @Patrick Maupin
    “However, I doubt it was so much an about-face as a correct reading of his current audience. ”

    I cannot read RMS’ mind, so you could be right. Personally, I suspect it is a blend of two different reasonings. A better understanding of the boundaries of “derived work” under US law developed over time and a change in strategy given changes in the world.

    The fact that language tools and methods used are irrelevant for copyright was there from the beginning. Handling people who try to outsmart the law is part and parcel of a judge’s job. They will not even blink if you try to come up with some smart reasoning why your translation of other people’s work must be “new and non-derived” because you used X instead of Y. Most of the time they do not care at all whether you use X or Y. If anything, the GPL was designed to catch such people early.

    @Patrick Maupin
    “One last comment on Linux and the GPL.”

    Linus also debunks this FUD (this is Groklaw’s coverage):
    http://www.groklaw.net/article.php?story=20110322014831856

    “It seems totally bogus. We’ve always made it very clear that the kernel system call interfaces do not in any way result in a derived work as per the GPL, and the kernel details are exported through the kernel headers to all the normal glibc interfaces too.

    “The kernel headers contain various definitions for the interfaces to user space, and we even actively try to make sure that the headers can be used by user space (and try to mark which of the headers are expected to be usable in such a way). Exactly because we know user space needs those details in order to interact with the kernel.

    “So I haven’t looked at exactly what Google does with the kernel headers, but I can’t see that they’d want to do anything fundamentally different from glibc in this respect,” Torvalds wrote.

    Such a declaration from the copyright holder(s) is legally binding in the USA (as far as I can understand that labyrinth they label “justice” in the US). For good measure, below is roughly the same from Andrew Morton.

    the first
    reply by Andrew Morton

    If there really is inlined functionality which we do not wish made available to non-GPL modules then it should be either uninlined and not exported or it should be wrapped in #ifdef
    GPL.

    the
    second reply by Andrew Morton

    Use of an inline function is just that: usage. It matters not at all whether that function is invoked via inline integration or via subroutine call. This is merely an implementation detail within the code which provides that function.

    Such functions are part of the offered API which have global scope,
    that’s all.

  170. And finally, this is the information FM et al refused to show in their FUD pieces.

    Please note that the headers EXPLICITELY state that they contain NO copyrightable information. The onus is on the FUDders to show that they actually DO contain copyrightable stuff.

    This is found in the headers:
    http://mrpogson.com/2011/03/17/oh-please-not-more-fud/

    *** This header was automatically generated from a Linux kernel header
    *** of the same name, to make information necessary for userspace to
    *** call into the kernel available to libc. It contains only constants,
    *** structures, and macros generated from the original header, and thus,
    *** contains no copyrightable information.

    (emphasis mine)

    And this is how they were created:

    http://www.netmite.com/android/mydroid/bionic/libc/kernel/README.TXT

    Bionic comes with a set of ‘clean’ Linux kernel headers that can safely be
    included by userland applications and libraries without fear of hideous
    conflicts. for more information why this is needed, see the “RATIONALE”
    section at the end of this document.

    these clean headers are automatically generated by several scripts located
    in the ‘bionic/kernel/tools’ directory, which process a set of original
    and unmodified kernel headers in order to get rid of many annoying
    declarations and constructs that usually result in compilation failure.

    the ‘clean headers’ only contain type and macro definitions, with the
    exception of a couple static inline functions used for performance
    reason (e.g. optimized CPU-specific byte-swapping routines)

    they can be included from C++, or when compiling code in strict ANSI mode.
    they can be also included before or after any Bionic C library header.

  171. @Winter:

    A better understanding of the boundaries of “derived work” under US law developed over time and a change in strategy given changes in the world.

    Except that they still use their own definition of derived works when it suits them. See, for example, what the FSF still has to say about Eclipse plugins, or about plugins in general.

    Now, there is possibly a small kernel of a public service there. If party A writes an Eclipse plugin that uses party B’s GPLed library, and then either party A or some other party bundles the plugin with Eclipse and distributes it, then the distribution of the program as a whole arguably violates the license on party B’s software.

    But if party A were to GPL the entire plugin that used party B’s library, and distribute it with a note warning about bundling it with the actual Eclipse distribution, then I don’t think that party B would have a cause of action against party A or anybody who took party A’s advice.

    The fact that language tools and methods used are irrelevant for copyright was there from the beginning. Handling people who try to outsmart the law is part and parcel of a judge’s job.

    Absolutely. But I only included that part of the quote from Moglen because the sentence starting “The situation” would be practically unreadable without that background. It is interesting that you spent more time responding to and reinforcing the one thing that Moglen said that I agree with, than you did addressing his, and the FSF’s still ongoing licensing power-grab. They publicly state, over and over, that the GPL is more viral than it actually is (and it’s plenty viral enough).

  172. @Patrick Maupin
    “They publicly state, over and over, that the GPL is more viral than it actually is (and it’s plenty viral enough)”

    I have had this discussion here before. I refuse to redo it.

    My point is that “derived” work goes as far as the law and courts let it go. And I simply ignore other peoples opinions on that. If the FSF see a broader interpretation than you do, let them prove it in court. I have absolutely zero intuition about the workings of the US legal system.

  173. > If the FSF see a broader interpretation than you do, let them prove it in court.

    My whole point (which I’ve made many times, but certainly made demonstrably over 7 years ago) is that they will continue to claim a broad interpretation (which they are still doing) but that this will never be tested in court by them. Given that they have proved other things in court, and that there are ample opportunities for them to test this in court, should they wish to avail themselves of them, I think my conclusion is pretty sound.

  174. As we are going cold Turkey with no new Android post in a month, here some recent snippets. I think that in the end the numbers of apps in the respective stores will be determined by the number of users (actually, total summed budget in market minus the “tax” from the app-store). So if Android market share overtakes iOS, so will the Android app store.

    Apple and Android app marketplace insight
    http://mybroadband.co.za/news/software/19099-Apple-and-Android-app-marketplace-insight.html

    The most recent report from the App Genome Project is based on data from August 2010 until February 2011. Over this period the number of applications in the Android marketplace increased by 127%. In comparison, the Apple store increased its numbers by 44%. Right now Android has around 100 000 applications in its marketplace. Apple added around 16 000 applications a month to hit 350 000 applications in February 2011.

    App Genome Report – February 2011
    https://www.mylookout.com/appgenome

    In support of recent handset shipment data covering Android’s tremendous growth, we found that even though there are more total apps in the Apple App Store, the Android Market’s relative growth rate is nearly three times that of the Apple App Store. If each market continues to grow at the same rate, the Android Market will have more apps than the Apple App Store by mid 2012.

  175. And some more about tablets.

    Android, Not Apple, to Lead Huge Tablet Market in 2014: RBC Analyst – eWeek
    http://www.androidtabletworld.com/android-news/android-not-apple-to-lead-huge-tablet-market-in-2014-rbc-analyst-eweek-2

    By 2014, Abramsky expects that more than 400 million people will own tablets, with 185 million units shipping in 2014. And while Apple may for now dominate the field?holding more than 90 percent of the worldwide market share, according to ABI Research?Abramsky projects that 40 percent of 2014?s sales will be tablets running Google?s Android OS.

    ……

    Abramsky, in a chart in his 88-page report, emphasized the potential for tablet growth by comparing tablet owners to those of other markets. While tablet and smartphone owners total approximately 394 million users, broadband subscribers are 555 million strong, PC owners run to 1.3 billion and Internet users 2 billion, while overall mobile subscribers have passed the 5 billion mark.

  176. @tmoney:

    This isn’t the first time that google has delayed the release of Android source. They have done this periodically from the start, when they don’t want people coding to internal interfaces that will be changing, for example. There is no question that, from the perspective of the outside world, it is more a cathedral model than a bazaar model. Of course, one of the benefits of a cathedral model is that you can polish your stuff a bit more before you show it off, but one of the downsides is that you’re expected to.

    Google certainly inhabits this in between world, driven by, among other things, the hardware schedule demands of its closest partners. But some of the uninformed, unthinking vitriol at slashdot, for example, would be amusing if it didn’t play straight into Apple’s and Microsoft’s hands. I have seen, in the same comment, allegations that google should be doing development in the open (bazaar-style) and that they shouldn’t have released the (allegedly broken) HoneyComb software yet!

    I don’t claim to know all google’s reasons for the delay. It may simply be, as google suggests, that they want to clean things up before the “Ice Cream” release. Or it may well be, as some at slashdot allege, that one goal is to let some of the bigger manufacturers (who monetarily helped development) get a headstart on shipment of HoneyComb hardware.

    Obviously google has their own internal business reasons to do whatever it is they do, but this certainly isn’t any sort of ethical issue. But I will be extremely surprised if google’s business issues drive any kind of significant change in their Android business model, and I’m actually kind of amazed at the number of people who (obviously on purpose, in some cases) misunderstand how google interacts with open source, and treat this as proof of some underlying honesty issues and/or the beginning of the end.

    Google has handed the FOSS world the dominant cellphone software stack on a silver platter. This is huge, and nothing google does in the future can detract from this. FSF-fetishists don’t like it because it’s not properly viral, and point to that fact as evidence of google’s evil behavior, conveniently ignoring that Aladdin, TrollTech, MySQL, and countless others have extremely similar business models, where they develop in a cathedral and push software out to the world without taking contributions. The difference, of course, is that those companies charge rent to those of their customers who want to use the software in proprietary ways, while google derives their profit in a more circuitous fashion. But to an FSF-fetishist, although proprietary software licenses are evil, it is permissible, even encouraged, to consort with those enemies for the express purpose of rent-seeking to support more GPLed programs, but completely ignoble to give your efforts to the world with no strings attached.

    What the purists fail to note, however, is that once google open-sourced Android, they put the same sort of competitive pressure on themselves as any of those for-profit GPL/proprietary companies. If they stop innovating with Android, others will pick up the slack. Of course, if at some point google wanted to take a breather, they could hand off a lot of the maintenance work to others — a tactic that would be sure death for those GPL/proprietary companies.

  177. @tmoney
    “Bogus how?”

    See Patrick Maupin’s response, except where he falls in the trap set by the journalists (you must admire the craftmanship here). But I had different reasons to consider it bogus.

    Journalism thrives by conflict. Bad journalists create fake conflicts to improve readership. And there is the divide and conquer motive of political reporting. Nothing better than support the purists of the opposition and get them up against other purists. But that is just form. It is the content that is bogus.

    So read the opening sentence:

    In the great mobile-device wars, Google (GOOG) has portrayed itself as the open-source crusader doing battle against the leaders in proprietary software—Apple (AAPL), Microsoft (MSFT), and Research In Motion (RIM:CN).

    Google never said anything like that. The word crusader betrays the intention of the article. It sets up commercial competition as a religious crusade. It also belittles the FOSS community.

    There are at least two parties needed for a conflict, so who is Google in conflict with?

    Dave Rosenberg, a longtime executive in the open-source software world, describes Google’s moves as an affront to hard-core open-source enthusiasts but adds that he isn’t surprised.

    Who? Executive of what?

    Maybe this Dave Rosenberg?
    http://news.cnet.com/8301-13846_3-20016140-62.html

    Dave Rosenberg dishes up “Software, Interrupted” with nearly 15 years of technology and marketing experience that spans from Bell Labs to multiple start-up IPOs to open-source enterprise software companies. He is co-founder of MuleSource and currently serves as the general manager of Hardy Way. He is an adviser to Canonical, IT Database, Puppet Labs, Riptano, and SOBA Labs. He is a member of the CNET Blog Network and is not an employee of CNET.

    Is he really this central in FOSS. Why not are more noted blogger or journalist, eg, SJVN, PJ, RMS, Linus, Morton? Further down we will get a quote from Eben Moglen, the one man that is up in purism to a level with RMS. But he will not exactly be fuming out of the mouth about this.

    And then we get this:

    It may also provide justification for critics of Android, who argued that Google created what amounts to a Wild West of mobile software by allowing people to do what they pleased with Android. Some of the early Android tablets, for example, looked silly when compared with the iPad, mostly because Android hadn’t been built for these types of devices.

    What critics? MS, Apple? So, Google not releasing software prematurely is promoting a Wild West of mobile software? And notice the dishing of Android tablets as “silly”.

    Now the play to the purists, divide and conquer:

    Open-source purists often contend that projects should be developed from the ground up, in front of the public, with people free to pick and choose the code they want, when they want it. This is the model Intel (INTC) has backed with its open-source rival to Android, called MeeGo.

    If these purists are so mad, why could the authors not get any of them to deliver a quote? Because the purest purists, the FSF, actually do not advocate this policy. For the real purists, the GPL only kicks in after release. The FSF is fine with keeping stuff in-house until you feel like getting help. And whatever you can say about Intel, they are not FOSS purists. Intel just wants to offload development costs.

    Ah, but they got Eben Moglen to quote:

    “It’s usually a mistake,” Moglen says. “Long experience teaches people that exposing the code to the community helps more than it hurts you.”

    So is Eben raving mad about Google’s treason, fuming out of the mouth? Actually, Eben considers it “inefficient”. It is bad for Google’s productivity.

    More “treason” to follow:

    In many cases, Google improves these products to meet its needs yet doesn’t turn the changes over to the broader open-source community, preferring to keep its secret sauce secret.

    This is well known, and one of the reasons the AGPL was formulated. The authors hope we did forget by the time we reach the end of page two that Google was supposed to be the long-time Crusader of Free Software, who now have fallen off their believes and switched to the enemie. Because, this fact alone would have told every reader that Google is keeping some things secret if they can. And that holding back an unfinished release of a major update in their software would be the least of the FOSS communities worries. Especially as few outside of Google are actually working on this code.

    And to show that they are really unbiased, the authors end their article with something positive to say about Google in a real anticlimax that few will read.

    Still, Google has opted to release some advances to the public and has received praise for these efforts. In particular, it provided the framework for a new type of data analytics system, called Hadoop, that has enjoyed wide adoption by big Web companies and traditional businesses.

    Compare the last two paragraphs with the first, and you see why I think this article is bogus.

  178. @Winter:

    See Patrick Maupin’s response, except where he falls in the trap set by the journalists (you must admire the craftmanship here). But I had different reasons to consider it bogus.

    What falling in the trap? I was referring to the back-and-forth on slashdot when I wrote “FSF-fetishists” and then shortened it to “purists” once, which was possibly a mistake, since you seemed to think I was talking about different people in the last paragraph than in the penultimate one. If you read the slashdot coverage, you will have to admit that these FSF-fetishists abound, and if you think honestly about my previous writings, you will have to admit that I am not prone to circuitous name-calling — when I think the FSF, or Stallman, or Moglen, or whoever, has personally fucked up, I name them directly.

    FWIW, you’ve done a fine job deconstructing the article, but yellow journalism happens regularly, and I was pretty much ignoring the article itself. I think google is far too big to stay above the fray, but if google ever does succumb to allegations that it is not open enough by changing its behavior (perhaps for the worse), that will be a direct result of what happens on places like slashdot, not what happens in businessweek.

  179. @Patrick Maupin
    Ok, sorry, I was somewhat loose with the reading. I think the article IS important as it is intended to FUD Android users and developers.

    My point was that the article was clearly intended to fan the “flamewars” between Google and FSF/GPL purists and Android developers. I normally shun /. as I find the SNR too low. The article itself has no substance that would give any reason for starting a fight again. So the authors resort to rhetorics. And I think they did a good job.

    I found it remarkable that you wrote:

    “FSF-fetishists don’t like it because it’s not properly viral, and point to that fact as evidence of google’s evil behavior,”

    “What the purists fail to note, however, is that once google open-sourced Android, they put the same sort of competitive pressure on themselves as any of those for-profit GPL/proprietary companies.”

    That must be purists from the /. crowd. What I heard from Bradly Kuhn’s blog/podcast was that the FSF people use Android phones because it is the “most free choice”. RMS does not use mobile phones as a matter of principle, but that is not related to Google. So the FSF are well aware that Android is a path to a completely free mobile phone software stack. The FSF want more, but don’t we all want more (of whatever we want)?

    So the article succeeded in getting you to rant about some immature /. posters, stoking the fire between “purists” and Google.

    As I said, that was based on a fast reading of your post. As always, all interpretations are my own and do not necessarily reflect opinions held by those whom I respond to.

  180. @Winter

    I still don’t see how that makes the article bogus. Hyped? Sure, pretty much all journalism is these days. As for the charge that Google hasn’t portrayed themselves as a bit of an OSS champion, I refer you to Mr. Rubin’s response to (IIRC) the whole Apple v Adobe kerfluffle:

    “the definition of open: “mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make”"

    But none of this is really surprising. In fact, I would be more surprised if Google didn’t slowly move to a more “closed” process like this. It appears the most truly successful uses of open source are a sort of Cathedral with a Bazaar on the front lawn model, we can see this in Red Hat, Mozilla, MySQL, Apple and many others. Compare to a completely open Bazaar model like say wikipedia, where anyone really can contribute [and then note that Wikipedia has been slowly closing their doors over time]. When your product becomes more than just a product, when it becomes a brand, there are certain benefits to a cathedral. Closing your doors, even just a little, gives you better control over how your brand is perceived. As Android and Google become more powerful brands, Google is going to want to exercise more control of what “Android” actually is.

    Despite the differences, it is apt to look at Android and the lessons to be learned from windows. Microsoft made it so that in the consumer’s minds, Windows was the PC. Although there were always manufacturer brands (dell, HP, gateway), if you asked someone what type of computer they had, you had a better chance of getting the windows version than the name of the manufacturer. Now look at what sort of headaches that managed to bring Microsoft. How many problems do you think were attributed to Windows, than an honest evaluation would have determined it was the hardware?

    So it can be with Android, and that appears to be what Google is afraid of here. They don’t want people putting Honeycomb on phones because they don’t know that it will work well, and they know if it doesn’t work, that Android will take the blame for it, not Samsung or HTC.

    I think you’re going to see a lot more of this. In fact, it wouldn’t surprise me if in a year or two, you see Android, and Android Community Edition or some other distinction between Android (Cathedral) and Android (Bazaar).

    Now, honest question because I don’t know:

    If Android has GPL components in it (which I think it does), and we have Honeycomb shipping (which we do) is Android in violation of the GPL until the Honeycomb source is released? Or are the GPL components segmented enough that Google can provide the source to those components, without the source to Honeycomb and still be in compliance with the GPL?

  181. @winter:

    So the article succeeded in getting you to rant about some immature /. posters, stoking the fire between “purists” and Google.

    Yes. Unfortunately those immature /. posters will have a much wider audience that google cares about than the businessweek article.

    @tmoney:

    When your product becomes more than just a product, when it becomes a brand, there are certain benefits to a cathedral. Closing your doors, even just a little, gives you better control over how your brand is perceived. As Android and Google become more powerful brands, Google is going to want to exercise more control of what “Android” actually is.

    I agree completely. Also, note that this sort of thing happened with Aladdin GhostScript — paid version before GPL version. The only difference is that here we have a bunch of hackers itching to get the source, either to the Xoom device they bought, or so they can put it on another device they bought. Hence all the noise at slashdot. The thing is, that the only reason this happens with google, and can’t/won’t happen with TrollTech/MySQL/et al, is that with google, there is immense pressure for a manufacturer to ship something once the hardware’s ready, and with those other companies — well, they just can’t ship the software until the software’s ready :-)

    If Android has GPL components in it (which I think it does), and we have Honeycomb shipping (which we do) is Android in violation of the GPL until the Honeycomb source is released? Or are the GPL components segmented enough that Google can provide the source to those components, without the source to Honeycomb and still be in compliance with the GPL?

    It’s my understanding this whole recent made-up kerfuffle about the kernel header files is simply because the Linux kernel is, in fact, the only part of Android that is GPLed. Since the kernel itself has a license exception for the parts above it, and since nobody who contributes to Linux (well, maybe except Oracle :) seems to mind Linux kicking ass in the cellphone space, it’s going to be awfully hard to gin up a GPL lawsuit over google’s practices in this area.

  182. @tmoney

    Android = Linux (GPL) + Dalvik (Apache) + More Userland (Whatever)

    If you cannot get the Linux under Xoom, then please scream blue murder at Google and sue them. Dalvik is for Gogole to do with what they like.

    I completely agree that Gogole keeping back on Android 3.0 is inconvenient and a reason to nag them.

    Given that Google are the only ones actually working on it, I cannot see who has a “moral right” to get it. They write it and they have the say.

  183. In contrast to the previous bogus article, this is the way to handle Google’s Android 3 policy.

    Be outraged if you feel like it, but keep it on topic. And note that this article actually tells us why the author thinks Google should act differently without dragging in opinionated quotes from “representatives of the community”. So the reader can actually make up her mind about whether or not she agrees.

    Google’s decision to delay Android 3.0 code is unacceptable
    http://www.zdnet.com/blog/open-source/googles-decision-to-delay-android-30-code-is-unacceptable/8559

    Google’s commitment to open source has been questioned in the past, in spite of its mega contributions to Mozilla, standards work and decision to use Linux. Yet it always seems to weather the controversies.

    But this? This decision to delay is simply unacceptable.

    Android is what it is today because of Linux and because of Google’s open source following — developers, organizations, customers and vendors. Delay the open source release? No. Justice delayed is justice denied.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre lang="" line="" escaped="" highlight="">