The Smartphone Wars: WebOS, we hardly knew ye

The business press is abuzz today with the news that HP is pulling the plug on its WebOS smartphone and tablet lines. This won’t be any huge surprise to people who’ve been following the discussions on Armed & Dangerous; WebOS has looked terminal to us for a long time.

Still…WebOS didn’t suck, technically speaking. It was certainly better constructed than the turd-with-frosting that is WP7. It’s worth taking a moment to reflect on the circumstances of its demise, and what its difficult history tells us about the future.

The cool thing about WebOS was that its architecture was beautiful. If the scuttlebutt from my friends who grokked it is true, it was the actual realization of Marc Andreesen’s dream of the browser becoming the entire OS. Apps were written in the browser using HTML5, CSS, and JavaScript.

WebOS’s problem was that the coolness stopped there. The source was closed, with all the usual bad effects including higher defect rates and lower developer interest. The actual implementation was prone to user-visible bugs. HP lacked the aggression and marketing skill to actually sell the devices to the public – actually, and even more damningly, they never seemed to figure out what the products were for.

Apple’s smartphones have a narrative hook: “We’re the experience designers for the cool kids.” Android’s smartphones have a narrative hook: “You are in control and the source is open.” RIMs have a narrative hook: “Serious devices for serious business.” Microsoft has one too, even if it’s just “We’re Microsoft. Resistance is futile.” WebOS never found its hook.

RIM is next to the wall, probably. WP7 should already have been terminated for extreme failure (Samsung’s own-brand Bada OS is actually outselling it), but it isn’t a normal product bet – it’s Microsoft’s forlorn and only hope of staying relevant in the smartphone-centric future of computing. Thus, it will probably continue bleeding cash until it takes Microsoft down with it.

Apple’s vulnerability shouldn’t be underestimated. Yes, they’re making money hand over fist and there has been adulatory press lately about record market cap and large cash reserves than the U.S. Federal government…but Android tablet market share has climbed from zip six months ago to 20% today, and if I were Apple’s planners I’d be worried about the fact that their smartphone sales aren’t growing any faster than total market volume.

Some sort of larger shakeout seems to be going on. WebOS kicked the bucket at about the same time that Apple’s recent (and admittedly small) market-share gains petered out and Microsoft actually recorded a tiny gain in its userbase. Only Android’s growth rate seems unaffected.

Gartner Research, which is generally very conservative and kind to market incumbents, has said it thinks consumers “in mature markets” (whatever that means) have essentially stopped buying dumbphones in favor of low-end Android handsets. Could it be that we’re also passing out of the era that smartphone platforms can grow without directly taking on each other?

I’ve written before that I wasn’t expecting that transition until mid-3Q2011…but on the unusual occasions that I’ve gone wrong in forecasting this market it has generally been by getting the direction of change right but underestimating the pace. It might be that Android, Apple, Microsoft, and RIM are now entering scorpions-in-a-bottle time.

If so, the likely outcome is the same as we’ve seen in other technology markets with strong network externalities. There can be only one…major incumbent. A whale, with a minnow or two in its shadow. Maybe Android should invert the Twitter fail whale into a success cetacean?

145 comments

  1. Hmm. Looking at your first chart, the Jun2011 score is:

    Android – 40
    Apple – 26
    RIM – 24
    MS – 6

    for a grand total of 96% of the total market. Was HP the other 4%? Would it be informative to still show it as a (moribund as of August) trend line? Or was it not even managing 4%?

  2. “but Android tablet market share has climbed from zip six months ago to 20% today”

    we’re quoting again shipments over sales. as we recently learned from the HP story, the difference can be rather large:

    “The source was closed”

    Compared to what?

    1. >Compared to what?

      No “compared”. It wasn’t released under an OSI-conformant license. This does tend to depress developer interest.

  3. The 20% market share for Android tables is based on shipments, not end user sales. Apple reports sales, not the number of tablets sold into the channel. That’s comparing, if you’ll forgive me, apples and oranges.

    Apple needs good, serious, focused, successful competition. Calling them vulnerable is true in the sense that their market share has nowhere to go but down. But I’ll believe that an Android tablet vendor is providing that competition when they start reporting end user sales.

    1. >The 20% market share for Android tables is based on shipments, not end user sales. Apple reports sales, not the number of tablets sold into the channel.

      Unless these tablets had a return rate upwards of 50%, that’s still pretty threatening sales growth. And they didn’t, because that would certainly have made the news.

  4. “I’ve written before that I wasn’t expecting that transition until mid-3Q2011…but on the unusual occasions that I’ve gone wrong in forecasting this market it has generally been by getting the direction of change right but underestimating the pace. It might be that Android, Apple, Microsoft, and RIM are now entering scorpions-in-a-bottle time.”

    Uh…isn’t it 3Q2011 now? Everything is going according to plan. You have forseen it….

    1. >Uh…isn’t it 3Q2011

      Yes, but a good six weeks before I thought we’d see clear signs of the transition. Which is a significant amount of time in this market.

  5. Eric: what’s your thought on this: In light of Google’s stated goal to make the browser be the OS, is there any hope at all of getting all that WebOS beautiful architecture and browser-*is*-the API coolness into Android?

    1. >Eric: what’s your thought on this: In light of Google’s stated goal to make the browser be the OS, is there any hope at all of getting all that WebOS beautiful architecture and browser-*is*-the API coolness into Android?

      Not for a long time, if ever. They’re already committed to a different architecture and it’s in heavy production.

  6. I just looked to day and according the FAQ you can’t access the microphone at all, and for the camera you need to run a web page with javascript to store a picture for you to get to it.

    Over a year ago (palm) there was an app contest and I noticed there were no barcode readers and I’ve been writing some superefficient QR code stuff. Then I found out how painful it would be to actually do a scan without some direct camera API. I think it was completely impossible at the time I inquired.

    Games might work well, but you need full access to the hardware.

    (I have an android tablet – the Toshiba Thrive and I find it wonderful – there are no limits; I can use Bluetooth GPS, or OBD2 devices, plug in USB peripherals, and generally manipulate anything I would ever want to).

    As far as Apple, I wonder how well iCloud will work with throttled or capped data plans.

    Google’s iterations might be something higher than linear. The permanent crossover over iOS might happen sooner than later.

    It is also bad form to try to ban and have recalled your competitor’s product when they are also your supplier.

  7. I remember being really excited to hear that HP were developing a linux based smart phone (at the research labs here in Melbourne, Australia – at the time I was looking for work and it sounded like the sort of thing I’d love to do if only I have any experience with that sort of thing…) and about 5 minutes later hearing that the pending merger with Compaq meant the entire research group was shutting down…

    meh…. live the dream…

  8. At Linuxcon this morning an HP executive was touting WEBOS like gangbusters. An hour or two later, this. I wonder.

  9. “Unless these tablets had a return rate upwards of 50%, that’s still pretty threatening sales growth. And they didn’t, because that would certainly have made the news.”

    If Google had the numbers to say that they’ve gotten to 20% tablets sold in six months, that would’ve made the news as well ;)

    Estimating from honeycomb activations, the market share looks closer to 5% but that doesn’t count nooks, etc…

  10. Marc Andreesen’s vision of the web being the entire environment has already come true, regardless of whether any particular operating environment has been written that way. Even Google has implemented some elements of that, for example in Chrome 10 and higher much of the browser’s own configuration framework takes the form of a mini web site hosted by the browser itself.

    You know what? Netscape has won the browser war. It was a Pyrrhic victory to be sure, but they won. Because the browser war was never about Netscape’s browser vs. Microsoft’s browser. It was about web applications vs. legacy desktop applications. In that light, Netscape won, and won big. Anyone developing a standalone Windows application today, unless it absolutely needs a ton of local power (video editing etc) would be laughed at.

    Even Microsoft has effectively admitted that the desktop is over, by introducing Office 365. Netscape won the war. Google finished what Netscape started, and is reaping the benefits.

    And, to the delight of all of us, they did so using Linux.

  11. Well, Eric – I have come back to say you were right. HP screwed the pooch.This is the OS/2 story all over again. The inferior user experience wins out because the company behind the superior product couldn’t be bothered.

    webOS was a beautiful architecture. Palm didn’t have the oomph to get webOS where it needed to be, so everyone was overjoyed that HP bought them. Then Mark Hurd got railroaded out of power and they brought on the guy (Apotheker) that sank SAP. I wonder if we can blame the political forces that scared the board enough to force Hurd out over false claims for the death of webOS?

    So now what? I don’t want to live in Apple’s walled garden, and I don’t want to give Google access to every detail of my life to be mined for advertising revenue. WP7 is dead, and even if it wasn’t, it sucks. And BlackBerry is a closed ecosystem that’s only good for email.

    Thanks for killing webOS HP. Thanks for killing the dream.

  12. It may not have been truly open, but WebOS was and will probably be the only OS I ever wrote patches for myself in Eclipse. It was too easy, and pretty fun. That’s what open should be like.

  13. Not so fast (on the Web Taking Over talk). I think instead the failure of WebOS may show that the vision of Marc Andreesen is ultimately doomed, or at least limited. While I never inspected the WebOS API myself, I quite believe tz when he says it was impossible to do anything real with. Like Java, there are things HTML (and attendant technologies) are good at, and things it isn’t — especially things that touch the hardware.

    Further, if everything is a browser app, then it’s all dependent on One Big Application in reality, that of the browser software itself. Maybe I’m old school, but I like at least *trying* to keep to the Unix philosophy of each application having it’s own purpose. Sometimes you have to combine for efficiency, but putting them *all* in one boat? Like the Java Revolution, I just don’t think it’s going to ever happen all the way.

  14. Bummed. WebOS was the right direction towards a more democratized (eg less expensive to develop) mobile app culture. Oh well.

  15. Speaking as a developer, it’s possible that “untyped, bandwidth-limited, bloated, and browser-dependent” qualifies as “more democratized”, but I’m pretty sure that “less expensive to develop” is just flat wrong.

  16. @esr Is there any technical reason that Android can’t gradually become programmable from the web, if the browser (or via app store) becomes more programmable? Excuse my ignorance, but why do we assume that the limitations of the current browser declarative languages (e.g. HTML, CSS) are the future webOS?

    If it is security that is so limiting the browser platform, the solution is a general purpose (and open source) declarative language[1], i.e. code that is sand boxed inherently by the referential transparency of the language. The user then has to give access rights to limited resources, but these are well defined because I/O (i.e. comonadic, coalegebraic, coinductive, ephemeral types) is never referentially transparent (i.e. they are always imperative, not declarative). But I wouldn’t vote for Haskell filling that role, as it is too difficult to learn.

    [1] Verifiable Functional Purity in Java, section 2.3 Untrusted code execution.

  17. I ask because as you may know I am designing such a language, and I want to know what I don’t know, which you may know. Thanks.

  18. I really don’t understand the passion you people seem to feel about this particular type of consumer appliance. Personally, I want three features in a phone – the “on” feature, the “call” feature, and the “off” feature. The “text” feature is occasionally useful, as long as predictive text doesn’t stop me from saying what I want to say. Mostly I use a real phone.

  19. I’m not so sure why you think there only has to be one big winner. In the video console space there are three large players (Sony, Microsoft and Nintendo). In that space, there are three thriving platforms, with game makers often writing for all three. I don’t see why the smartphones can’t settle into that kind of equilibrium, rather than a winner-take-all one.

    1. >I’m not so sure why you think there only has to be one big winner.

      History. Also, the smartphone market seems to be headed that way fast – WebOS just gone, RIM crashing, Microsoft at barely above statistical level, and Android about to blow through 50% market share going up. The major unknown is whether Apple will be able to retain more than 10% market share when the dust settles.

  20. >>The 20% market share for Android tables is based on shipments, not end user sales. Apple reports sales, not the number of tablets sold into the channel.

    >Unless these tablets had a return rate upwards of 50%, that’s still pretty threatening sales growth. And they didn’t, because that would certainly have made the news.

    The difference isn’t just the return rate, it’s also the rate at which units in stores translate into sales versus just sitting around taking up shelf space. I don’t think we know that. But as a couple points of comparison:

    Galaxy Tab’s return rate is 15%:
    http://m.nypost.com/p/news/business/galaxy_tab_dim_bulb_KbD4K6OUjTC99SQn2efrsJ

    The HP TouchPad’s sell through *was* bad enough to be newsworthy. Before being cancelled, it had in some places a reported sell-through rate of less than 10%. Best Buy had 270k units of inventory of which it managed to sell fewer than 25k:

    http://allthingsd.com/20110816/ouchpad-best-buy-sitting-on-a-pile-of-unsold-hp-tablets/

    So Android probably isn’t *that* bad but given HP as competition, selling even 50% of what you ship to stores starts to look pretty good. Discounting by half might be a decent guess here.

  21. @Albert:

    In the video console space there are three large players (Sony, Microsoft and Nintendo). In that space, there are three thriving platforms, with game makers often writing for all three.

    Maybe four soon … If google gets imaginative and turns that Mot set-top box business in a gaming console business.

  22. Until Google screws up and blows a hole in that. Or Apple. The way this industry seems headed is that everyone is doomed to get their heart broken.

  23. Would there be any reason for Google to purchase the Palm/WebOS patents at this point, or would that be pointless given the Motorola purchase?

  24. @uma, agreed that is my point, compiling a general purpose declarative (i.e. referentially transparent lambda calculus) language to LLVM or JVM is a way to safely run untrusted general purpose software on any platform. My question is why do we (or do we) assume that this won’t be the web3.0, instead of limiting to HTML5 (or any other non-Turing complete DSL)? Technical or market reasons it isn’t likely? Thanks.

    Btw, apparently Clojure is not abstraction safe, nor pure functional (just optional immutable data structures), nor is it web scale composition, because it is only unityped. Also I am thinking of something that must be much, much easier to learn and use.

  25. @Shelby:

    > Is there any technical reason that Android can’t gradually become programmable from the web,

    Obviously not. I expect that google’s chrome native client will eventually migrate to Android.

    > I ask because as you may know I am designing such a language,

    While designing languages, and designing new industry-standard javascript database storage DOM APIs are sexy tasks, I am not sure the world needs a new language specifically for this task, but I am sure that the first browser that provides a built-in sensible way to associate website, or a subset of its web pages (and associated javascript) with a chrooted, size-limited directory in the user filesystem (that the user himself can manipulate outside the browser, examine, back up, restore, etc.) will rapidly start ramping marketshare until the other browser manufacturers realize the simple, forgotten utility of the lowly text file.

  26. If your website could pop up a dialog with a suggested directory name and size limitation (with the browser automagically warning you if the directory exists or if there is not much room on the filesystem, etc.) and the user could give your website local storage, there are lots of interesting programs that could be written in all sorts of languages, because there are Python to javascript compilers, Java to javascript compilers, etc. available.

    Something like Google’s native client implementation might be necessary for really, really high performance code needs, but for everything else, you should just assume that continued competition in javascript engines will make javascript supplant C as the new assembly language for fast CPUs with lots of memory.

  27. @esr: WP7 should already have been terminated for extreme failure (Samsung’s own-brand Bada OS is actually outselling it), but it isn’t a normal product bet – it’s Microsoft’s forlorn and only hope of staying relevant in the smartphone-centric future of computing.

    You are underestimating and overestimating Microsoft’s plight at the same time. I remember your World Domination 201: it looked logical at the time, but today it’s obvious that it was not entirely correct. Microsoft was one of the leaders in all stages of PC era: it produced BASIC for the Altair, it produced DOS for both 8bit (MSX DOS) and 16bit (MS DOS) systems, later it introduced Windows and made it 32bit, today it switches to 64bit without losing any momentum.

    Yet mighty Dec failed and disappeared in this time because minicomputers were disrupted and replaced with PC (also note that it successfully survived that “supercritical” transition from 16-bit PDP-11 to 32-bit VAX). Why? Smaller, initially incompatible systems come – and ate bigger systems without remorse. So what we need is not 32bit-to-64bit transition, but transition to the new, initially incompatible, platform. And for a long time situation looked hopeless: PC is already truly personal (and often moveable) computer, where do you move from it, to a wristwatch? Yet apparently such platform exist: it’s smartphones and tablets. Smartphones require totally different and initially incompatible OS and tablets may attack PC turf in time. Early clumsy attempts (like ASUS Transformer) look endearing, but they can not replace PC today: smartphones and tablets may be as powerful today as PC was 10 years ago, but they are not powerful enough to replace today’s PC. Not yet. Yet the direction is clear: eventually smartphones/tables descendants will kill the PC descendants. How fast? Well, the first popular personal computer was introduced [ironically enough] by Apple: that’s 1977. In about 10-12 years Dec (leader of the “world above”) started faltering and disappeared by 1998. As you can see it took roughly 20 years – and for 10 years it looked like Dec will not be affected at all.

    “Smart tables” revolution was started, again, by Apple when it presented iPhone in 2007 (PDA existed long before but they played entirely different role: noone ever tried to replace “PC work” with “Palm V mork”, it just made no sense; PDAs were for totally different tasks). This means Microsoft will not be really hurting till 2017 or so. But if it’ll not muscle in “smart tablets” world then by 2025-2030 it’ll be history: either it’ll disappear completely or it’ll go the way of IBM and will switch to some kind of consulting business.

    That’s why Microsoft has a lot of time (today we are in 2011, not in 2017) but problems are significantly more acute as well: if in the next 5-6 years it’ll not catch the “smart tablets” world then it’s entire future is bleak. Microsoft knows it – that’s why it’s ready to spend countless billions on WPx while it still have them.

    @twilightomni It may not have been truly open, but WebOS was and will probably be the only OS I ever wrote patches for myself in Eclipse. It was too easy, and pretty fun. That’s what open should be like.

    Well, it’s nice dream, but in reality you need other kind of openness: you need separate vendor for OS and for the hardware platform. Windows is as closed as they come but it’s licensed pretty liberaly. WebOS may be nice technically but it still was tied to the hardware of a single vendor. Closed platforms can survive as minnows (Apple survived for years in the Wintel world, right?), but they can never be a whale.

    @Albert: In the video console space there are three large players (Sony, Microsoft and Nintendo). In that space, there are three thriving platforms, with game makers often writing for all three.

    Suprisingly enough video consoles don’t have a network effect! Gamers are intensely loyal to franchises (there are countless game series published), but they will happily play Soul Edge on Playstation (SONY), then later Soulcalibur on Dreamcast (SEGA), then continue with SoulcaliburII on Gamecube (Nintendo), then SoulcaliburIII on Playstation2 (SONY again) and finally SoulcaliburIV on XBox360 (Microsoft). Sure, there are some retention effects (that’s why all consoles of the last generation support some kind of backward-compatibility), but it’s limited: people are switching to newer games, they don’t expect to play the same games again and again (and a few ones who do expect that just keep old consoles around).

  28. @Patrick Maupin, agreed JavaScript is another ubiquitous VM to compile to, in addition to LLVM and JVM.

    I am thinking that all of those mentioned languages won’t scale to inter-website compositionality because they don’t enforce pure functions (per the reference I cited). If we talking about a webOS, then code from different sources probably needs to interopt. Whereas, capability-based security has a least common denominator of no interoperability (as the browser makers are sliding into, if they could only get rid of the legacy script hack). There only ones that enforce pure functions that I know of currently are Haskell, Clean, Mercury, and JoeE. (not including the total type dependent theorem prover languages Adga, Coq, and Epigraph). Imho, Haskell (and probably the others too) is not a viable candidate for mass adoption, and afaik they all have at least one critical weakness in terms of type safe compositionality (they conflate interface and implementation, i.e. violate SPOT). The details are covered in the links I provided.

    Any way, it is good to hear that this general thrust is what people want. I want to hear what others are thinking about this.

  29. @khim
    “This means Microsoft will not be really hurting till 2017 or so. But if it’ll not muscle in “smart tablets” world then by 2025-2030 it’ll be history: either it’ll disappear completely or it’ll go the way of IBM and will switch to some kind of consulting business.”

    I think MS will start to fall apart next year (2012).

    There are many reasons why MS plight will be different than DEC’s. Too many to list here.

    Most important of them is that MS has always lived of siphoning off the profits from productivity increases that came with moving new tasks to the computer. That stalled after XP sp3 and Sharepoint.

    Neither Vista, nor WP7 solved any problems and if anything, they lowered productivity. Hence no one wanted to pay for them. Since Vista (or earlier), MS are faltering. No new products that would entice people to shell out money, no new markets, and increased competition. The result was already visible before the release of Vista, a hemorrhaging of developers. When the OLPC XO1 came out, MS was unable to write hardware drivers for it for a very long time. They simply had lost the know-how.

    Now smartphones and tablets are a new market with unprecedented growth potential. This market is all about connection and service: Social media, Search, and adds. All areas where MS are weak, and all on a platform they have never taken seriously: resource limited ARM. MS are all about wasting computing resources, bad security, and i86. Hence their WP7 Phone 7 is a dog in every respect except the UI shine.

    So, in the fall of 2011, MS have no way out of a fixed market: PC based enterprise computing. They have a second (third?) rate OS that neither scales up to HPC nor scales down to mobile devices, nor can break out of the traditional Intel architecture with many GB of RAM.

    I think this fall the investors will see MS’ future clearly: Milk the desktop PC users (enterprise) for every penny, and stop all adventures outside of the PC Desktop/laptop. The rational thing would then be to stop all money losing enterprises (Bing, Phone) and sell off the ones that can be made profitable (XBox, MSN, Add network). And simply milk the really profitable parts (Office/Sharepoint, OS).

    So I expect Ballmer to start spending more time with his family next spring. And MS starting to “spin-off” losing businesses.

    1. >So I expect Ballmer to start spending more time with his family next spring. And MS starting to “spin-off” losing businesses.

      Alas, I think you are too optimistic. If Microsoft’s shareholders had any clue at all, Ballmer – as incompetent a buffoon as we’ve ever seen in charge of a Fortune 500 company – would already be out on his ass. That he isn’t means he won’t go until the rot becomes far more obvious. In particular, this means WP7 won’t be shitcanned before it wastes a lot more time and money.

  30. @Winter I think this fall the investors will see MS’ future clearly: Milk the desktop PC users (enterprise) for every penny, and stop all adventures outside of the PC Desktop/laptop. The rational thing would then be to stop all money losing enterprises (Bing, Phone) and sell off the ones that can be made profitable (XBox, MSN, Add network). And simply milk the really profitable parts (Office/Sharepoint, OS).

    This will be the rational choice. They can survive for a long time selling Office to govenment agents (which they need because nothing else can open these “standard-based” documents). But this will mean that Microsoft basically accepts it’s fate and slowly goes “IBM way”. Microsoft have way too many ambitions and way too large cash stash to abondon bigger ambitions just yet. Note that OS is profitability may evaporate very fast: the only reason people are even using this pig is the fact that it runs Office and a lot of other programs. Office can be tied to the OS easily, but other programs from third-party developers will be ported when/if they’ll feel that Microsoft is no longer unstopabble.

  31. @esr
    “If Microsoft’s shareholders had any clue at all, Ballmer – as incompetent a buffoon as we’ve ever seen in charge of a Fortune 500 company – would already be out on his ass.”

    What I heard was that Ballmer was bad, but no one knows anyone to replace him. All other candidates were seen to be even worse.

    @khim
    “Microsoft have way too many ambitions and way too large cash stash to abondon bigger ambitions just yet.”

    That was my point. The cash is basically supporting the stock price. Wasting it directly affects the balance sheat of the pension funds and other shareholders. The shareholders “rational” policy would be to stop MS wasting their cash reserves. But I think they are too late already, with $8B wasted on Skype and assorted billions on patents. Maybe the shareholders simply have no idea where to put their billions instead, and so let it rot in MS.

    And the short term view seems to be to invest heavily in patent trolls. Hence another argument for MS to switch business plan to become a full time patent troll.

  32. If I were Apple, I would be fairly pleased that a 15 month old phone was successfully maintaining market share while competing with new phones. This result is awfully unexpected from where I’m sitting — I’ve read Boyd and I believe in tight OODA loops too. So I dunno.

    Holiday sales ought to be interesting, which is not a code word for “I think Apple will rock it.” I’m honestly curious. Since I don’t know if Apple will be on a new Chinese carrier and I don’t know if there’ll be a low cost option, I’m a long way away from having enough info to make a prediction.

  33. @Bryant
    “This result is awfully unexpected from where I’m sitting ”

    Unexpected, but not new for Apple. Their 2nd/3rd generation Macs were so good compared to the PC stuff, that people were able to run them much longer than common PCs and still do useful work.

    However, this is a target market I am not very interested in. The $80 phone from Huawei is what I find captivating. Together with this very old, but little known “news” from 2008 Mobile Phone Payment System Blossoms in Kenya.

    A capable information device plus a working (micro-)payment system in one. The sky is the limit.

  34. Would there be any reason for Google to purchase the Palm/WebOS patents at this point, or would that be pointless given the Motorola purchase?

    Were I Google, I’d have somebody make a phone call, at least. More ammunition couldn’t hurt.

  35. @Bryant:

    The vendors with the Android OODA loop are increasing their own shipments at a much higher CAGR than Apple is, so if you plot the graphs Apple might not really be doing all that well in terms of growth.

    Having said that, the effects of a tight OODA loop are probably more easily visible in a static market. In a market growing this fast, it makes sense that sticking to a very SKUs, and leveraging economies of scale everywhere throughout the design chain, as Apple has done, might actually be a net benefit compared to juggling lots of SKUs at lower volumes — certainly for margin, and possibly for maximizing their own unit shipment growth rate.

    As a counterexample, Samsung still hasn’t brought the phone that is apparently the fastest selling Android phone in the world out in the US yet.

    But Samsung’s smartphone shipments were in the same ballpark as Apple’s last quarter, and Samsung certainly has manufacturing muscle, so it will be very interesting to see what happens the next couple of quarters.

  36. Yet mighty Dec failed and disappeared in this time because minicomputers were disrupted and replaced with PC (also note that it successfully survived that “supercritical” transition from 16-bit PDP-11 to 32-bit VAX).

    Having worked for a company (Honeywell Aerospace) that was at one time heavily invested in DEC’s minicomputers (PDP-11 and VAX), I can say that your analysis is a little short-sighted.

    PCs didn’t do what they appeared to do from the end-user’s point-of-view. From an IT and systems engineering point of view, corporate IT — and that’s what we’re talking about here because nobody but +Jay Maynard has minicomputers or mainframes in their house ;) — we went from centralized systems to decentralized systems. First thing that happened is that we went to a two-tier client/server model.

    We left some horsepower on the back end in the form of servers running Unix — and to some extent other OSes like Windows, VMS, z/OS, etc. Inititally, we started using PCs as fancy terminals and then over time we moved more and more presentation logic to the PC, while keeping the databases, storage and business logic on the back end. As time marched on, we scaled out and migrated from a two-tier to an N-tier system — servers being clients for other servers.

    Today, through Web-based technologies, we’ve moved more of the presentation logic back into the server room. Add in Web 2.0-technologies like AJAX and we move a bit more back to the PC in the form of JavaScript, HTML 5 and CSS 3. But the core of the system — databases, storage and business logic, still runs in the server room.

    The key difference is that in the server room, we went away from proprietary, closed stacks to open protocol and OS stacks that include a mix of proprietary and open source code. We moved from proprietary vendor-locked-in data formats to more open data formats. And we moved from proprietary, closed hardware to open, commodity hardware — and even to no hardware at all (virtualization/cloud infrastructure).

    Smartphones and tablets simply add yet another front-end technology into the mix. It currently allows us to divide between data producers and data consumers, but eventually data producers may even be doing so on their smartphones and tablets. To some extent, they already are.

    But that server room still isn’t going away. Sure, it may be “in the cloud”, or it may be in a “private” cloud, but it’s still there, it’s just more hidden now than it ever has been.

    IOW, DEC wasn’t really displaced by PCs, it was just displaced by other technologies that we were cheaper and better suited to the hidden, tucked-away server room that was needed to drive the PCs.

    BTW– those same applications that were running on the minicomputers? They’re running on the servers now, with fat fancy presentation logic on the PC clients. More of that is moving into a “private cloud” with all the servers virtualized and in the same global data center. Eventually, I suspect that the global data centers will move to third-party providers — or even the global data centers spun off as separate companies unto themselves — but not until corporate IT is ready to plunge into that pool.

  37. I’ve been a Palm guy since the 90s. I seriously like my Pre Plus, with reservations — the must-have HanDBase app was never ported over to the WebOS system, a major fly in the ointment. Yet, I had already made the decision to change to Android when I upgrade later this year. I had hopes when HP bought Palm that brighter days were on the way, but unfortunately not. Looks like my decision to switch was prescient — no sense in buying a Pre 2 or anything similar, and then watching the support dwindle away. Well, it’s been a long run, but so it goes. I’m looking forward to jumping into the Android world (where HanDBase is flourishing). The only problem is which of the many offerings to choose.

  38. I have thought that it was pretty brilliant of IBM to get out of the PC business as a way of breaking any influence or control MSFT had over their corporate direction. Now they can pursue Linux, or any other direction they choose in a way they could not back when OS/2 was new. I was cheered by the news that HP planned to make all their Windows computers dual boot with WebOS (i.e. Linux), and at the time thought only that it was a signal of how much power MSFT had lost that they could not kill that project for HP. Now I kind of wonder what has been going on behind the curtains, what pressures MSFT may have wrought to kill this experiment before it ever really began. Maybe nothing, but I am pretty sure both IBM and Google are happy they are not beholden to MSFT for anything they do internally.

    The whole Carly Fiorina and subsequent CEO shuffle is another whole saga. Really sad, wish the calculator division had been spun out as part of Agilent.

  39. > What I heard was that Ballmer was bad, but no one knows anyone to replace him. All other candidates were seen to be even worse.

    May I suggest Stephen Elop?

  40. Mobile dead bodies: Windows CE, Windows Mobile, Windows Mobile 7, Meego, Maemo, Maemo (linux) + proprietary layer (N9), WebOS (linux + proprietary layer), RIM OS (whatever that’s), RIM QNX, Symbian, Palm OS, Psion.

    In this rapidly evolving world does it make any sense to buy/own anything instead of renting?

    Some of WebOS products were release less than 6 months ago, weren’t they? Will it hardware support, probably not, software support/updates, probably not, support from external developers, NO! It was a very bad bet!

    And going forward every consumer/developer will be very suspicious of any new platform. It’s Android vs. iOS. And the worst position is for Nokia: fiasco after fiasco: maemo, meego, symbian, Qt, all dead ends. Who would be foolish enough to pick any of them, no one. N9 is still maemo plus proprietary interface.

    The third horse will be inevitably Microsoft Windows Mobile 8. We can’t discard them, they are stubborn and like to play dirt. WM8 will be their, what 5th, 6th try in the mobile space?

    In their favor they have: platform, access to content, developers, search, advertising, maps with Nokia, app store (their own or Nokia’s), and lots of money to burn.

    How about Google?

    Strong position but they have some things to alienate almost everyone: didn’t release Android v3 and buying Motorola.

    Not releasing the code was bad PR especially among their strongest supporters, geeks and open source communities. The explanation they gave can equally apply to future releases: will they release the source of Android v4?

    Buying Motorola puts them in hardware business and in clear conflict with their hardware partners. Partners that despite the canned responses in support of the deal are very uncomfortable. Anyone but Samsung is very bad spot and the cheap hardware makers are in a very bad spot if Google doesn’t release the source code to Android v4.

    Google has the platform (the web/Android/ChromeOS), are making strong inroads to content and now the hardware part, they control everything just like Apple, why would the give up on that?

    Hardware makers will lose either way. Stay with Android and are completely at the mercy of Google, turn to Microsoft and be at their mercy (a fate worse than death) or develop their own platform and face a short and quick death.

    Apple is stable. They know what to do and it’s working. Nothing more to say or there’s?

    Amazon. Will Amazon start selling their own tablets. It makes sense since they have lots of content and a tablet is the right vehicle to reach that. Follows the pattern of Kindle.

    Sony. They have access to lots of content: games, music film and tv, a platform: PlayStation, but no path to evolve. they are dependent of Google or Microsoft. Their mobile hardware is not desirable and their partnership is risk of collapsing.

    It’s a 2 horse race, with possibly a third, Microsoft, no matter how remote that might appear to be.

    It also appears that future, disappointingly, will more closed than it’s now.

    Apple planned everything almost to perfection, Google’s counter response to stop Apple having a majority of marketshare was realized almost to perfection as well, and their next seems to follow the Apple footsteps in closing everything and control every step and taxing everything that passes trough them.

    It’s all about platform and content. Since no one else controls the two the way Google, Apple, Microsoft controls it, the rest will be limited to producing hardware. I predict that the chinese will absorb every other small hardware player.

  41. Apple is the only real player in tablets. That 20% figure quotes shipments, while Apple reports actual sales.

    Best Buy is sending back their entire inventory of HP tablets.

    There really is no market for “tablets.” There is a market for Ipads.

  42. @esr
    “If Microsoft’s shareholders had any clue at all, Ballmer – as incompetent a buffoon as we’ve ever seen in charge of a Fortune 500 company – would already be out on his ass.”

    @winter
    “What I heard was that Ballmer was bad, but no one knows anyone to replace him. All other candidates were seen to be even worse.”

    All other candidates in corporate technology firms? I find that hard to believe. It’s desparately obvious that Microsoft’s next CEO needs to come in from outside Microsoft, someone who can change the culture and reinvent the company before it’s too late. They need a Gerstner, not an Akers.

  43. > I am sure that the first browser that provides a built-in sensible way to associate website, or a subset of its web pages (and associated javascript) with a chrooted, size-limited directory in the user filesystem (that the user himself can manipulate outside the browser, examine, back up, restore, etc.) will rapidly start ramping marketshare until the other browser manufacturers realize the simple, forgotten utility of the lowly text file.

    Yes, yes, yes.

  44. ” What I heard was that Ballmer was bad, but no one knows anyone to replace him. All other candidates were seen to be even worse.”

    May I suggest Stephen Elop?

    And a loud cheer from Finland echoes through the ‘nets.

  45. >> May I suggest Stephen Elop?

    > And a loud cheer from Finland echoes through the ‘nets.

    (At the risk of stating the obvious, I’m writing from here and that’s my real first name.)

  46. Blackberry’s image is changing fast from “boring corporate phone” to “phone used by rioters” here in the UK. BBM has made huge inroads into the youth Market, and might have some real network lock-in effects.

  47. I am sure that the first browser that provides a built-in sensible way to associate website, or a subset of its web pages (and associated javascript) with a chrooted, size-limited directory in the user filesystem (that the user himself can manipulate outside the browser, examine, back up, restore, etc.) will rapidly start ramping marketshare until the other browser manufacturers realize the simple, forgotten utility of the lowly text file.

    While I do appreciate the flexibility and low overhead of plain text, I don’t see what’s so bad about using sqlite to do pretty much this–in fact, it’s precisely what webOS uses for per-app storage, and I think it’s both Firefox and Chrome use for HTML5 local storage. I’ve actually twiddled some app settings by hand, and while it’s not a text file, I think manipulating it’s about a wash.

  48. @Christopher Smith:

    > I think manipulating it’s about a wash.

    The average user, who doesn’t even know how to use diff, doesn’t know what an sqlite database is, and barely knows enough to use wordpad instead of word — is going to somehow open up two versions of a database file and merge them by hand?

  49. The average user, who doesn’t even know how to use diff, doesn’t know what an sqlite database is, and barely knows enough to use wordpad instead of word — is going to somehow open up two versions of a database file and merge them by hand?

    The “average user” isn’t going to even be aware of the per-site directory setup. I think the gap between the sets doesn’t-know-anything and knows-about-the-schema-but-can’t-use-SQL is much wider between the latter and can-use-basic-SQL.

  50. +Christopher Smith:

    1) Perhaps you missed where I said the browser would present the user with a dialog asking where he wanted his files?

    2) What makes you think one of the files the application emits into the top directory can’t be README.TXT?

    3) What makes you think all programs that need local storage need particularly difficult schema?

    4) What makes you think that one size really fits all, or that I think that one size fits all? I’m not arguing against the ability to use a database if its appropriate. But it’s a really big hammer for a lot of potential programs.

  51. What I find far more significant than HP pulling the plug on WebOS is the fact that they’re looking to get out of the PC hardware business. They’re the #1 vendor in unit volume, but their margin is only around 5%. This confirms my impression that I’ve had for many years, that being a Microsoft OEM must really, really suck.

  52. > being a Microsoft OEM must really, really suck.

    And a lot of the Android OEMs will have similarly sucky experiences. But so will Nokia and RIM. And a few MS and Android OEMs will differentiate and do really well. Then they will get complacent and get their butts kicked.

    Those who eat fugu soup are stupid. But those who don’t eat fugu soup are also stupid.

  53. Perhaps you missed where I said the browser would present the user with a dialog asking where he wanted his files?

    I did manage to overlook the following comment. That approach, interestingly enough, fits in quite well with a capability-based OS model that I’ve been toying around with for some years now that could make the functionality available to any program, not just browsers, but implementing the jail in the browser would be an interesting first step. IE, of course, would obligingly support “..”…

  54. I am going to question the fundamental premise of “smart” phones as a future computing platform for serious users. From what I’ve seen, it can never replace a full fledged desktop computer system, or even laptops, and even for the casual purposes like using Word Processors. Why? Because the touch screen centric input model used by a majority of these smart phones can never ever replace a full fledged keyboard or even a keypad. Touch screen can get irritating on the finger fast and even the soft touch of capacitive models is not the solution to the problem of finger irritation.

    I’ve noted that even web browsing is quite limited on the Android platform. Leave aside the technical issues with regard to resolution etc. (which can be fixed somewhat by the zooming/scaling done by these devices), even searching within a web page is a pain using the default Android browser. And apps are still limited to the computing power of these devices making only the simplest apps worthwhile. More importantly when it comes to serious computing, I don’t think these smart devices are quite good enough in terms of battery life. The one biggest advantage a smart device like a phone or tablet should have is superior battery life, but because of the increased computing power, the battery life inevitably takes a beating.

    All these aspects convince me that we will move back to the traditional computing platforms after the smart phone fad has died. In fact, the so-called “dumb phone” might even make a revival since the actual purpose of these devices is primarily to make phone calls than anything else and people do value a battery that stays charged longer increasing the talk-time also.

    Yes, smart phones and tablet devices will continue to stay in the market, but I don’t see them dominating computing in any way, now or in the future unless somebody discovers a battery that is capable of delivering longer charge cycles even with increased usage.

    1. >Because the touch screen centric input model used by a majority of these smart phones can never ever replace a full fledged keyboard or even a keypad.

      You haven’t been paying attention. Those of us who envision a smartphone-centric world agree with you about this part. We think that these ergonomic problems will be solved by hardware like the Redfly that is a peripheral docking station for smartphones. Want to use a full-sized screen and keyboard? Sure….walk up to one and plug in.

  55. @Hari
    @esr
    RE: Smartphones not taking over the world because they are touch screen focused.

    Doesn’t have to be. A hammer has to be one thing, not another. As does a screwdriver. A box with storage, a processor and connectors with sufficient bandwidth (both “local” and “net”) doesn’t have to be one thing or another. It can be both at different times or the same time.

    @Patrick
    RE: Filesystems and Databases. At a certain level of abstraction a filesystem IS a database. It’s the cost of the abstraction that matters. On a cellphone/tablet that cost is higher than it maybe should be, unless SQLite has gone back to raw tables on a partition.

  56. @esr I realize that the problem of input could be handled in other ways but then the main USP of smart devices is lost if the mobility aspect is sacrificed even if only temporarily. As a matter of fact, I’m typing out this comment on my viewsonic tablet, but I am using a USB keyboard (attached to the leathercase I purchased) because typing on the main screen is painfully slow literally. But having this leathercase-cum-keyboard does make the device bulkier and reduces the convenience of carrying it around everywhere, even if only a little bit heavier.

    @William I still think that smaller devices will always have this input method issue even with voice recognition, leave alone traditional input methods because of design constraints but I agree that it doesn’t have to be touch-centric. Leaving that aside, I think the bigger issue will be the battery life as devices get more advanced in terms of OS and processing power.

    In any case. I am eagerly awaiting future software developments in the smart devices field.

    1. >I realize that the problem of input could be handled in other ways but then the main USP of smart devices is lost if the mobility aspect is sacrificed even if only temporarily.

      Who says it has to be sacrificed? I think where we’re heading is for these docking stations to be common as dirt. You have one at home, another at your office, and Internet cafes are full of them.

  57. >a lot of the Android OEMs will have similarly sucky experiences.

    At least they won’t be in the situation where their OS vendor is making the lion’s share of the profit per unit. They will still have the problem of not being able to offer significant differentiation to allow them to maintain a workable gross margin, though.

  58. “Docking station” will be the wrong term.

    I saw a movie where someone from Google talked about a bluetooth like system to connect a phone wireless to keyboard, mouse, and screen.

    Just stand next to the station and it works.

    Think of a cafe or lobby with cheap stations where you just sit down and surf as a step up from free wifi.

    1. >Just stand next to the station and it works.

      I used to like this idea. Then I thought about the issue of how your smartphone will know which station to bond to in a public space where there might be dozens of them. There are security issues as well. I think this is a case where cable would be best.

      >Think of a cafe or lobby with cheap stations where you just sit down and surf as a step up from free wifi.

      Exactly. All this needs is for somebody with the clout of, say, Google, to publish an open connector spec that all Android phones would support. The micro-USB connectors on recent Android phones would only require a bit of firmware on either end to talk to a USB monitor, and keyboards are of course trivial.

  59. I would argue that given the market has been reacting to Apple moves since 2007 that Apple is still inside the OODA loop for the industry. That there are more models of Andriod phones released than iPhones is not an indicator that Apple competitors are executing a strategy more tightly than apple management.

    Who is spotting and exploiting market conditions more rapidly? Who is establishing strategic control over the battlespace through preemptive component purchases and investment in supply chain? Who has a tight focus on optimizing both revenue and profit? I would argue Apple. Having the superior OODA loop means you’re skating to where the puck will be and forcing the enemy to react to your moves more often than you are reacting to theirs and you can react more quickly.

    Who’s supply constrained on 10″ LCDs, flash RAM and milled aluminum laptop cases? Who’s scrambling and who’s not? RIM scrambling. Nokia scrambling. Dell scrambling. Sony scrambling. Microsoft scrambling Moto dead. HP concedes. Even when these companies forsaw Apple’s moves they couldn’t preempt them. The iPad was telegraphed but attempts to effectively beat Apple to market failed…failing the Act portion of the OODA loop.

    Even Googles acquisition of Moto is reactive as Apple foresaw that IP would be a strategic battleground whether you agree it should be or not. Google has been able to react to Apple moves reasonably effectively by changing android based on iOS, releasing honeycomb after the iPad, etc. But if google was inside Apple’s OODA loop the it would have had honeycomb out before the iPad.

  60. There’s no need for using smartphones and public docking stations unless the expectation is that there is no Internet connectivity at these stations. Given that most would reasonably expect that places that would provide public docking are also providing wifi that’s a hard assumption to make

    Given Internet connectivity then simply accessing your data on the cloud is superior. No smartphone pairing needed. No smartphone needed at all except, perhaps, as part of two factor authentication.

    Auto syncing of data between phone, desktop, pmp, set top and cloud would be a highly useful product. Who recently announced a product like that? OODA loop…

  61. A side point…the cost incremental from keyboard + mouse + monitor to keyboard +mouse + monitor + Internet appliance is no more than $99 retail given the aTV is only $99.

    And who has a strategic partnership with Starbucks already in place?

  62. You may find it interesting to note that, after announcing that the WebOS hardware was to be killed off, HP also announced deep liquidation price cuts. For Touchpads the liquidation prices are $99 and $149 for 16GB and 32GB models respectively, and there is a fairly frightening feeding frenzy in effect. For a dead, orphaned platform, whose only practical hope for long-term ongoing support is community-developed Android ROMs.

    And yes at that price I want one, now. But I can’t FIND one.

  63. @esr
    Security problems with wireless docking stations are easy to solve. Do an ssh handshake while pointing your camera to its screen. The screen displays the fingerprint as a bar code. Or other relevant information.

    Be the first to create that app ;-)

  64. @Christopher Smith:

    IE, of course, would obligingly support “..”…

    Heh. They probably would, at that.

    @Hari:

    Yes, smart phones and tablet devices will continue to stay in the market, but I don’t see them dominating computing in any way, now or in the future unless somebody discovers a battery that is capable of delivering longer charge cycles even with increased usage.

    People are working on the power density problem.

    @William O. B’Livion:

    At a certain level of abstraction a filesystem IS a database. It’s the cost of the abstraction that matters. On a cellphone/tablet that cost is higher than it maybe should be, unless SQLite has gone back to raw tables on a partition.

    I don’t disagree at all. But (to tie this post in with Eric’s next post about protocols), the FAT file system is an abstraction that is already fully understood by the smartphone and any computer you might plug the thing into. In fact, the SQL database abstraction is probably already installed inside the FAT abstraction.

    @Some Guy Says:

    At least [the Android OEMs, unlike the MS OEMs] won’t be in the situation where their OS vendor is making the lion’s share of the profit per unit.

    No question the MS OEM has it worse.

    They will still have the problem of not being able to offer significant differentiation to allow them to maintain a workable gross margin, though.

    But there are some high-end MS computer manufacturers that do OK. Not as well as Apple, obviously, but I think there is going to be room inside the Android ecosystem for niche high-end players.

    @esr:

    I used to like this idea. Then I thought about the issue of how your smartphone will know which station to bond to in a public space where there might be dozens of them. There are security issues as well. I think this is a case where cable would be best.

    Some sort of nearfield communication might suffice. Might be able to charge the battery at the same time.

    All this needs is for somebody with the clout of, say, Google, to publish an open connector spec that all Android phones would support.

    Maybe google could use its new patent portfolio to bludgeon Apple into allowing others to use magnetic connectors (I assume there are some patents there, because I haven’t seen those elsewhere).

    @Nigel:

    I would argue that given the market has been reacting to Apple moves since 2007 that Apple is still inside the OODA loop for the industry.

    It’s a bit more complicated than that, I think. Even Apple has been reacting to Appl moves since 2007.

    Who is establishing strategic control over the battlespace through preemptive component purchases and investment in supply chain?

    It’s not preemptive when you realize belatedly that the industry really isn’t ready for your model of work for a long time to get everything exactly right, and then release and sell a gazillion in the first two months. The “preemptive” component purchases are merely examples of planning that Apple should have done before they did, and even now isn’t executing as well as some other companies. In one sense, these are good problems to have, but in another sense, every dollar left on the table for your competitor to pick up is bad business.

    Who has a tight focus on optimizing both revenue and profit? I would argue Apple. Having the superior OODA loop means you’re skating to where the puck will be and forcing the enemy to react to your moves more often than you are reacting to theirs and you can react more quickly.

    There is no question that, by deciding to take a lot of time and build the “perfect” phone, and then take a lot of time and decide to build the “perfect” tablet, that Apple sped up the industry. But those were actually long-term projects, and once they showed the money available in these product categories, their competitors have ramped up to competitive levels with many competing series of shorter-term product categories. For example, this is supposedly Samsung’s product roadmap. The question is, if Samsung stumbles on a phone and Apple stumbles on a phone, who is worse off? In reality, Apple is quite unlikely to stumble too badly on the next phone, but you can’t claim that’s because of a tight OODA loop — quite the opposite. A tight OODA loop practically implies that you’re going to start off shipping shit, and then rapidly improve. Is this how Apple works? Lately, Apple has worked well by finding a product category where the status quo is not very competitive, and they can bring one big idea to bear and dominate the category. They did this with MP3 players and enough cash to buy the RIAA. They did it for awhile with smartphones and enough financial clout to convince the carriers things had to change. They are currently doing this with tablets, based considerably on repurposing technology they developed for MP3 players and smartphones.

    This is really smart, but is it not really a tight OODA loop, unless you compare it to the entrenched phone players who had been slow-waltzing with the carriers for so many years they had no idea how to jitterbug.

    Who’s supply constrained on 10? LCDs, flash RAM and milled aluminum laptop cases?

    There’s a glut of LCDs of all sizes right now, there’s no such thing as flash RAM, and other vendors use fiberglass cases.

    Who’s scrambling and who’s not? RIM scrambling. Nokia scrambling. Dell scrambling. Sony scrambling. Microsoft scrambling Moto dead. HP concedes.

    This is how the free market works. You’re conveniently leaving out all the Android players who are doing really well. Not a good way to argue.

    Even when these companies forsaw Apple’s moves they couldn’t preempt them. The iPad was telegraphed but attempts to effectively beat Apple to market failed…failing the Act portion of the OODA loop.

    No other manufacturer thought that the demand would be that high. Now they know. You see a tight OODA loop. I see a really good high-stakes gambler.

    Even Googles acquisition of Moto is reactive as Apple foresaw that IP would be a strategic battleground whether you agree it should be or not.

    Yes. The police force at Columbine was reactive, because the killers foresaw that guns would be discharged on campus that day. I agree that what Apple is doing is apparently legal, but it’s not ethical.

    Google has been able to react to Apple moves reasonably effectively by changing android based on iOS, releasing honeycomb after the iPad, etc. But if google was inside Apple’s OODA loop the it would have had honeycomb out before the iPad.

    I don’t agree. The dust hasn’t cleared yet, but tablets aren’t phones, and google was still really just getting started in phones when Apple was getting started on the iPad. I think that if Apple was inside google’s OODA loop, the growth percentages in phones would be reversed.

    There’s no need for using smartphones and public docking stations unless the expectation is that there is no Internet connectivity at these stations. Given that most would reasonably expect that places that would provide public docking are also providing wifi that’s a hard assumption to make

    A docking station that could provide a standardized, secure internet and KVM connection, as well as charging, would probably be golden. No reason to use a third-party’s computer that others might have infected. No reason for those nearby to be able to sniff your packets. Granted, other advancements, like SSL-everywhere, would make the second one of thos irrelevent.

    @Winter:

    Security problems with wireless docking stations are easy to solve. Do an ssh handshake while pointing your camera to its screen. The screen displays
    the fingerprint as a bar code. Or other relevant information.

    I saw something like this a few months ago. Can’t find it right now.

  65. @esr: >>Just stand next to the station and it works.

    >I used to like this idea. Then I thought about the issue of how your smartphone will know which station to bond to in a public space where there might be dozens of >them. There are security issues as well. I think this is a case where cable would be best.

    Ironically this issue has been solved. By the Pre3 and TouchPad, although we’ll never see it in action as the Pre3 died before shipping. They’re using data over the TouchStone induction charging system to control bonding. That gets you down to a range of a few centimeters and solves the which station issue.

  66. I used to like this idea. Then I thought about the issue of how your smartphone will know which station to bond to in a public space where there might be dozens of them.

    That part is really simple. The station simply has a QC code of its connection URL either on screen as stated above, or even on an ID sticker. The QC code would be more secure, of course, as it makes MitM really difficult if not impossible.

    My concern is more with the insane amount of bandwidth that will be needed for all those phones to talk to all those displays at once. (It’s one thing to wirelessly export your laptop/tablet/smartphone screen to your TV at home to watch a movie, but if you think it’ll still work when two dozen other people are trying to do the same thing?) For that, the HDMI port on the cell phone would be perfect. (It can also carry some other two-way protocols between the devices if need be.) I’m not a bit worried about using bluetooth for keyboard/mouse though. Keystrokes and mouse events are orders of magnitude less bandwidth than display updates in the megapixels-at-60-Hz league.

  67. It’s not preemptive when you realize belatedly that the industry really isn’t ready for your model of work for a long time to get everything exactly right, and then release and sell a gazillion in the first two months. The “preemptive” component purchases are merely examples of planning that Apple should have done before they did, and even now isn’t executing as well as some other companies.

    Dropping half a billion on components before your competitors are even aware those components are about to be really important next year is sure as heck preemptive and is not an example of bad planning and hardly “belated”. Effective supply chain management in business is like effective logistic management in warfare. There’s a saying about strategy vs logistics.

    So name some companies executing better than Apple in supply chain management.

    For example, this is supposedly Samsung’s product roadmap.

    “It looks like the highly anticipated Android is back to its holidays launch date instead of the recently rumored October release.”

    Okay, so the superior Samsung and Google OODA cycle is following iOS in terms of having a single cohesive OS across tablet and phone…something that Apple strove for from the beginning.

    Why thrash around with Honeycomb if Ice Cream Sandwich is the desired end state? Which is reactive?

    A tight OODA loop practically implies that you’re going to start off shipping shit, and then rapidly improve.

    Boyd is rolling over in his grave. No, this is not what a tight OODA loop implies. You achieve a tight OODA loop through superior training, practice and planning. The foundation of modern operational planning is not starting off with some shit and making things up as you go along. If you start with a shit plan (or product), you’ll simply evolve a shit plan (or product).

    The key to a effective OODA loop is IMHO primarily in the Orientation phase. The ability to identify those observations consistent with achieving your desired strategic goal and ones that are detrimental to your objectives. It’s not simply acting faster but acting effectively. This also implies that you have clear strategic goals. Clear strategic goals is not “kill the bad guys” or “maximize profits this quarter”. Here’s an interesting article (if slightly fluffy) on the topic:

    http://www.bloomberg.com/news/2011-08-19/want-to-be-like-apple-lose-the-bafflegab-commentary-by-virginia-postrel.html

    The military is obsessed with articulating a clear and decisive strategic goal and conveying this to all the troops (doubly so among the Marines). That way, everyone know what the desired end state should be and more importantly (for the Marines anyway) WHY that end state is desired. Not starting with some shit product/plan/objective and improve it as you go.

    You don’t win just because you’re executing the OODA loop faster. A tight OODA loop isn’t JUST fast.

    So you are correct that Apple does not ship a shit product and improve it as they go along. They create a product in line with their strategic objectives and make moves toward those objectives. Whatever Apple’s faults, lacking a keen sense of identity and strategic objectives is not one of them. They do Orientation better than anyone.

    Their best of breed products, market cap and massive growth is the result of their currently superior OODA execution toward a clear and internalized corporate strategy. They operate inside the effective decision loop of their competitors because the OODA loop isn’t a simple linear loop.

  68. @The Monster
    “My concern is more with the insane amount of bandwidth that will be needed for all those phones to talk to all those displays at once. (It’s one thing to wirelessly export your laptop/tablet/smartphone screen to your TV at home to watch a movie, but if you think it’ll still work when two dozen other people are trying to do the same thing?) For that, the HDMI port on the cell phone would be perfect. (It can also carry some other two-way protocols between the devices if need be.) I’m not a bit worried about using bluetooth for keyboard/mouse though. Keystrokes and mouse events are orders of magnitude less bandwidth than display updates in the megapixels-at-60-Hz league.”

    First, this would be a short range channel, like bluetooth. That would be ~5m (class 3) and around 24Mb/s (theoretically).
    https://secure.wikimedia.org/wikipedia/en/wiki/Bluetooth#Uses

    I would expect the docking stations to use some kind of high level graphical language io bitmaps. Theoretically, you could drive a terminal using as an HTML5 display. Taht would be low-bandwidth.

  69. @nigel So you are correct that Apple does not ship a shit product and improve it as they go along. They create a product in line with their strategic objectives and make moves toward those objectives. Whatever Apple’s faults, lacking a keen sense of identity and strategic objectives is not one of them. They do Orientation better than anyone.

    Not really. Their observation is second to none. Sadly their orientation superb but slow and their “decision” and “act” parts are quite sub-par. Their iOS4 and iOS5 mostly inroduce things lacking which were present in Android from the beginning, for example. It took 2-3 years to bring them to iPhone. Android makers typically need 3-6 months to do the same.

    This basically means they usually have an early lead but their only hope for the win is to act quick enough to prevent any kind of coherent response. It worked with iPod, but iPhone and, most probably, iPad are different.

    @nigel Why thrash around with Honeycomb if Ice Cream Sandwich is the desired end state? Which is reactive?

    There are no “desired end state”, sorry. This is not a military operation. You may have “a vision” in your head, but you never know if the customers will share it or not. Thus the importance of OODA loop. Honeycomb was important because it gave something to sell before iPad lead become unbreakable.

    @nigel The key to a effective OODA loop is IMHO primarily in the Orientation phase.

    Only if you have fast enough OODA loop. If you don’t… orientation can only buy you so much. Apple is currently the biggest smartphone maker but I doubt it’ll be one for long: Samsung will overcome it either in Q4 or early next year. That’s the price of slow OODA loop: even suberb orientation is not enough if your cycle is slow.

    Apple’s cycle is faster then, for example, Nokia, but much, much slower then Samsung or HTC. There are only so much your excellent observation and orientation skills can buy if you do one cycle per year vs 3-4 (or even more) per year.

  70. In an unintentionally comical article, Helsingin Sanomat (the largest newspaper in Finland) writes about Ari Jaaksi, who led Nokia’s Maemo/MeeGo development until last September, and his comments about Nokia’s woes. Jaaksi says that Symbian had turned into a religion inside Nokia and that the management’s only concern about the release of the N900 was how it was going to affect the sales of Symbian products, not how the thing itself was received. Jaaksi was subsequently hired by HP to work on WebOS, and he goes on to praise HP, saying that they wanted him in that position more than Nokia ever did when he was working on Maemo. The article was apparently approved at Helsingin Sanomat before HP made the announcement, since there is not a word about HP killing WebOS. Somewhat mysteriously, this was corrected in the print edition but not yet on the web. Anyways, looks like Jaaksi either has pretty bad luck or an anti-Midas touch that kills operating systems. His comments about Nokia’s management should probably be taken with a dose of salt. He did head Maemo development and it seems he’s a little eager to blame the management for what happened.

  71. First, this would be a short range channel, like bluetooth. That would be ~5m (class 3) and around 24Mb/s (theoretically).

    In an environment filled with these docking stations (libraries, coffee houses, commuter trains?), I can easily imagine a 5m range including a dozen or two people close enough I’d have to share bandwidth with them. Even if they’re not close enough to reliably communicate with my docking station, they’re close enough to interfere with that communication. I know Bluetooth is supposed to be able to avoid contention because of the frequency-hopping technique it uses, but I’d really like to see some real-world testing to see if tightly packed high-bandwidth communications hit a saturation point where it all unravels.

  72. Not really. Their observation is second to none. Sadly their orientation superb but slow and their “decision” and “act” parts are quite sub-par. Their iOS4 and iOS5 mostly inroduce things lacking which were present in Android from the beginning, for example. It took 2-3 years to bring them to iPhone. Android makers typically need 3-6 months to do the same.

    The speed at which specific features are added don’t indicate decision cycle at all. The speed at which specific features are added are a function of development and QA timelines. The priority of these features that took 2-3 years was considered either low or that is should be fixed as part of a major subsystem refactor.

    There are no “desired end state”, sorry. This is not a military operation. You may have “a vision” in your head, but you never know if the customers will share it or not. Thus the importance of OODA loop. Honeycomb was important because it gave something to sell before iPad lead become unbreakable.

    Except there is a desired goal (aka end state) and Apple’s was to have coherent user experience across phone and tablet. Therefore features important to this goal was prioritized and features secondary to this weren’t. What you assert are slow responses from Apple are actually due to decisions Apple has made in terms of which ecosystem aspects are most important. If you don’t know your desired end states it’s very difficult to get there.

    Honeycomb hasn’t been all that effective as a stopgap IMHO. There are probably more 2.2 or 2.3 tablets out there than 3.0 tablets and probably more useful to their users than the honeycomb tablets are to theirs. The nook is a fantastic example of an android tablet (even if it is an e-reader) that serves the needs of their users very well. I recall reading that the Nook Color had the highest customer satisfaction among e-readers.

    Only if you have fast enough OODA loop. If you don’t… orientation can only buy you so much. Apple is currently the biggest smartphone maker but I doubt it’ll be one for long: Samsung will overcome it either in Q4 or early next year. That’s the price of slow OODA loop: even suberb orientation is not enough if your cycle is slow.

    Highly developed orientation skills buys you everything. It is what allows you to apparently move from observe to act without appearing to do the intervening orientation and decision steps. When you’ve internalized your objectives to that degree, response to stimulus becomes natural. That’s the key for rapid OODA loops for either fighter pilots or businesses.

    That Samsung will eventually end up shipping more units across the entire smartphone spectrum against a company that only produces high end smartphones is not an indicator of speed…either OODA or deployment but the strategic decision on the part of Apple not to compete in the low end smartphone market. Neither are specific features like notifications or whatever.

  73. I agree with Nigel on the docking station…. what’s the point? The trend towards all your data residing on the cloud is clear. So no need to dock – all you need is a thin terminal with internet.

    The docking station idea seems based on 5-10 year old thinking.

    I think most will be happy just using their smart phones when on the go anyway. Or maybe a tablet.

  74. @nigel
    Apple might be the fastes develepor with errorless decisions and the shortest OODA loop. That does nor change the end result. Because for any real world system, blind evolution will find “good enough” paths to optimal solutions faster.

    You only need an exponentially growing market, competition between enough agents and a short generation time

    A single company with a few models cannot win this race against dozens of competitor fielding a hundred models..

  75. @Winter ” Because for any real world system, blind evolution will find “good enough” paths to optimal solutions faster.”

    I’ll bet on very smart people with good track records over blind evolution every time.

  76. @Nigel:

    Dropping half a billion on components before your competitors are even aware those components are about to be really important next year is sure as heck preemptive

    No. It would have been preemptive if that is how it started. At best, you can chalk it up to the “lessons that Apple might have learned” category — they’re doing a lot better, but not perfect.

    and is not an example of bad planning and hardly “belated”.

    At several points, they have been severely supply constrained. It was after one of those points that they started shoveling money at vendors like mad to lock up future supplies. Hence belated. The facts are out there. You can argue till you’re blue in the face about the interpretation, but we’ll have to agree to disagree about that.

    So name some companies executing better than Apple in supply chain management.

    WalMart, Costco, possibly Samsung — they ship a lot more phones than Apple and shortages are never in the news.

    Okay, so the superior Samsung and Google OODA cycle is following iOS in terms of having a single cohesive OS across tablet and phone…something that Apple strove for from the beginning.

    When you start from behind, it takes time to catch up. I think it’s pretty amazing that Google has caught up and surpassed in volume, and is now busy cementing the feature set to hold the lead.

    Why thrash around with Honeycomb if Ice Cream Sandwich is the desired end state? Which is reactive?

    Having something that is good enough to ship a lot of phones now, to deny those customers to Apple, seems proactive to me.

    A tight OODA loop practically implies that you’re going to start off shipping shit, and then rapidly improve.

    Boyd is rolling over in his grave. No, this is not what a tight OODA loop implies.

    So if your country is overrun by what appears to be a superior force, and you have some cheesy old munitions available now, but can build some better stuff in 6 months, Boyd would suggest you do nothing now? You’re right — he is rolling over in his grave. How fast can you make him spin?

    You achieve a tight OODA loop through superior training, practice and planning. The foundation of modern operational planning is not starting off with some shit and making things up as you go along. If you start with a shit plan (or product), you’ll simply evolve a shit plan (or product).

    Conflating plans and products is not very smart. Really good execution does not at all depend on starting off with the best product.

    The key to a effective OODA loop is IMHO primarily in the Orientation phase. The ability to identify those observations consistent with achieving your desired strategic goal and ones that are detrimental to your objectives. It’s not simply acting faster but acting effectively. This also implies that you have clear strategic goals. Clear strategic goals is not “kill the bad guys” or “maximize profits this quarter”.

    Sounds like what google’s doing to me, but what do I know?

    Here’s an interesting article (if slightly fluffy) on the topic

    Yeah, I read that. Looks like what google’s doing to me too, but again, what do I know?

    The military is obsessed with articulating a clear and decisive strategic goal and conveying this to all the troops (doubly so among the Marines). That way, everyone know what the desired end state should be and more importantly (for the Marines anyway) WHY that end state is desired. Not starting with some shit product/plan/objective and improve it as you go.

    Again, I believe everybody at google knows the clear goals. Look at the google+ ramp.

    You don’t win just because you’re executing the OODA loop faster. A tight OODA loop isn’t JUST fast.

    No, and I don’t think google’s and samsung’s and HTC’s are JUST fast, either.

    So you are correct that Apple does not ship a shit product and improve it as they go along. They create a product in line with their strategic objectives and make moves toward those objectives. Whatever Apple’s faults, lacking a keen sense of identity and strategic objectives is not one of them. They do Orientation better than anyone.

    Which is absolutely fine until you take too long doing orientation. Which may or may not have happened for the iPhone 5. We’ll see in a few months.

    Their best of breed products, market cap and massive growth is the result of their currently superior OODA execution toward a clear and internalized corporate strategy.

    There is no question that they are currently executing well and are being rewarded handsomely for it. But I still think that that’s more a function of the vision of Steve Jobs about what customers really want than anything else, which is certainly a valuable subset of things that could be called “orient.” So far, where Apple has made their money is by finding a niche market, asking “What would it take to make this into a mass market?” and then doing that. Now, on the iPod, they managed to leverage the RIAA’s paranoia into an iTunes near-monopoly that helped them to completely take over the market. And with the cellphones, they managed to crack the power held by the carrier oligopoly where other vendors hadn’t. But in other markets they haven’t managed those kinds of tricks. Apple can’t seriously expect to hold anywhere near Android’s market share long-term, in either phones or tablets, so they need to keep doing what they’ve been doing — being the high-end cachet vendor, and using profits from one market to blow a different, sleepy one wide open. They always need to be on the prowl for fresh competitors.

    They operate inside the effective decision loop of their competitors because the OODA loop isn’t a simple linear loop.

    No, it’s not a simple linear loop, and google greatly complicated the loop and split it into a zillion sub-loops with Android.

    The speed at which specific features are added don’t indicate decision cycle at all. The speed at which specific features are added are a function of development and QA timelines. The priority of these features that took 2-3 years was considered either low or that is should be fixed as part of a major subsystem refactor.

    OooooKaaaaay. So when Google ships something that’s not 100%, it’s because their OODA loop is bad, but when Apple does it, it’s because their OODA loop is good? Got it.

    Except there is a desired goal (aka end state) and Apple’s was to have coherent user experience across phone and tablet.

    And this particular goal will be realized soon enough by multiple operating systems. So that can’t be the only goal. In fact, that’s not a goal at all — it’s a means to the goal, and one that the enemy capability will soon have as well.

    Therefore features important to this goal was prioritized and features secondary to this weren’t.

    Sure, but Google’s goal with Android (which is a real goal, not a method for achieving it) is to make sure that network effects are never put in place that allow Apple or anybody else to place a toll booth between them and the web public, and from their perspective, this could wait.

    Honeycomb hasn’t been all that effective as a stopgap IMHO. There are probably more 2.2 or 2.3 tablets out there than 3.0 tablets and probably more useful to their users than the honeycomb tablets are to theirs. The nook is a fantastic example of an android tablet (even if it is an e-reader) that serves the needs of their users very well. I recall reading that the Nook Color had the highest customer satisfaction among e-readers.

    Ah, but Honeycomb demonstrations probably slowed iPad takeup.

    Part of market share is selling. Another part of market share is keeping your competitor from selling. A lot of Android vendors have credible demos with Honeycomb, and even if people aren’t buying because it’s not done and the price is too high, it might put them into a wait-and-see mode. If Honeycomb did that with enough potential iPad customers it served its purpose admirably. If feedback for how honeycomb works, from developers and early adopters, makes Ice Cream Sandwich better, it’s a bonus.

    Highly developed orientation skills buys you everything. It is what allows you to apparently move from observe to act without appearing to do the intervening orientation and decision steps.

    And to the extent Apple has those, we’ll probably soon see some almost unrelated product come out of nowhere.

    That Samsung will eventually end up shipping more units across the entire smartphone spectrum against a company that only produces high end smartphones is not an indicator of speed…either OODA or deployment but the strategic decision on the part of Apple not to compete in the low end smartphone market.

    This I can agree with, to a certain extent. Having said that, I don’t think Apple was expecting the capabilities of the low-end smartphone market to rise, and the price to drop, quite as rapidly as happened. So I think the lawsuits are about Apple seeing that they are on the way to being relegated to niche status long before they were prepared for that fate.

  77. @khim:

    This basically means they usually have an early lead but their only hope for the win is to act quick enough to prevent any kind of coherent response. It worked with iPod, but iPhone and, most probably, iPad are different.

    Exactly. And one of the things they had going for them with iPod was the RIAA’s greed — after they saw how well Apple was doing, they didn’t want to ink any more deals until they were absolutely sure they were squeezing the golden goose within a millimeter of its life.

    There are no “desired end state”, sorry. This is not a military operation.

    I disagree. The desired end state was to not have Apple lock up the next computing platform. And google is seriously winning.

    You may have “a vision” in your head, but you never know if the customers will share it or not. Thus the importance of OODA loop. Honeycomb was important because it gave something to sell before iPad lead become unbreakable.

    Exactly. The network effects of the platform mean that you want to get customers now. The OODA loop doesn’t mean tha you stop until you have something perfect. It means you have a process in place to get you to your end goals, and if you have a compelling reason to ship something now, then you find something to ship now.

  78. @The Monster:

    In an environment filled with these docking stations (libraries, coffee houses, commuter trains?), I can easily imagine a 5m range including a dozen or two people close enough I’d have to share bandwidth with them.

    At that range, a phased antenna array with a lot of DSP processing could probably pick out individual transmitters pretty easily. There’s some work going on in this area. It seems physically plausible, so hopefully it’s not just 250 mpg style hype:

    http://venturebeat.com/2011/07/28/steve-perlman-unveils-dido-white-paper-explaining-impossible-wireless-data-rates/

    @phil:

    I agree with Nigel on the docking station…. what’s the point? The trend towards all your data residing on the cloud is clear. So no need to dock – all you need is a thin terminal with internet.

    The point is merely one of trust. People are much more likely to believe that you built an unhackable keyboard and mouse than that you built an unhackable terminal…

    I’ll bet on very smart people with good track records over blind evolution every time.

    Yeah, especially the people who are smart enough to use genetic algorithms to solve really hard problems. Oh, wait…

  79. @phil
    If you think design finds better solutions faster than a directed random walk (evolution) you have something to learn about search. Google seems to have some experience in the field.

  80. @Patrick Maupin
    I disagree. The desired end state was to not have Apple lock up the next computing platform. And google is seriously winning.

    Well, I partially agree. Google obviously wanted to keep anyone from becoming man-in-the-middle WRT to access to the consumer – but they started the game back in 2005, when Apple was not in game. And if you’ll consider how good Apple is at keeping it’s secrets…

    I doubt it was ploy against Apple – more likely they geared to fight Microsoft. But when it was found out that their foe is Apple, not Microsoft they changed the direction in hurry (talk about tight OODA loop, heh). In fact is was done so fast apparently even Google’s CEO was not really informed:
    http://www.businessinsider.com/the-first-time-steve-jobs-flipped-out-on-eric-schmidt-2010-3

    But yes, that was the goal, but… it’s almost like “make everyone happy” kinda “goal”, not really constructive…

  81. @khim:

    I doubt it was ploy against Apple – more likely they geared to fight Microsoft. But when it was found out that their foe is Apple, not Microsoft they changed the direction in hurry (talk about tight OODA loop, heh).

    In my explanation, “Apple” is certainly just a proxy for “those forces that want to intermediate themselves just to seek rent.” The Business Insider piece is interesting. It is plausible that Eric Schmidt knew before 2007 that Apple was working on a phone — they started in 2005[1] as well, and he was on the board. And it is hard to believe that Eric didn’t know that Google was doing anything at all in the cellphone space. So it may have been a reaction to Apple, or Microsoft, or just the fact that the carriers were, in a lot of cases, effectively controlling access to the web.

    [1] Kinda, sorta: See http://mobile-review.com/articles/2010/iphone-history3-en.shtml and then read parts 1 and 2 if you want to see how long Apple was kicking around the idea of a phone.

    But yes, that was the goal, but… it’s almost like “make everyone happy” kinda “goal”, not really constructive…

    No, but the internal marching orders given in furtherance of that goal, whatever they were, obviously worked really, really well.

  82. @Patrick Maupin
    “The point is merely one of trust. People are much more likely to believe that you built an unhackable keyboard and mouse than that you built an unhackable terminal…”

    Just don’t run windows on your terminals and you are 95% of the way there. I’m thinking just a browser and that’s it. No opportunities to download anything.

  83. @phil:

    Just don’t run windows on your terminals and you are 95% of the way there. I’m thinking just a browser and that’s it. No opportunities to download anything.

    More like 47.5% of the way there, since most of the browsers allow you to download plugins. Even a big sign telling people to logoff and back on before using the browser, combined with a startup script that killed the browser cache, would be no match against an adversary who logs into each terminal and modifies the startup script.

  84. @Patrick Maupin

    “More like 47.5% of the way there, since most of the browsers allow you to download plugins. Even a big sign telling people to logoff and back on before using the browser, combined with a startup script that killed the browser cache, would be no match against an adversary who logs into each terminal and modifies the startup script.”

    what do you think is easier/cheaper: doing a special Google Chrome build that disables plugins/etc or building a dock that works with mobile devices?

  85. @phil:

    what do you think is easier/cheaper: doing a special Google Chrome build that disables plugins/etc or building a dock that works with mobile devices?

    The answer to that question is obvious. And google-branded public internet terminals could certainly be cheap and trusted. But absent that kind of juggernaut entry into the market, which would you be more likely to trust — a specialized secure keyboard interface, or some white box computer running a random browser? Even if the guy at the internet cafe seems to know what he’s doing, do you trust his google customization?

    Or let me put it to you another way: what if Microsoft got into the market with a branded internet terminal? Trust or not trust?

  86. what do you think is easier/cheaper: doing a special Google Chrome build that disables plugins/etc or building a dock that works with mobile devices?

    How about ChromeOS in a VM that is always re-instantiated from a read-only image for each user? You could probably put your VM host and your Chrome image on an SD card, flip the write-protect tab, and put it on a card reader behind a locking door. Operator of the site where the terminals are installed gets a key and some spare SD cards to be used if there’s ever a problem booting.

    If you have any local writable storage, you could have an init script wipe the swap and browser cache on the HD at boot, do it again between users. Booting off the SD card while the HD spins up ought to be pretty fast, too.

  87. Yeah, you can design an architecture where you have a CPU-less docking station for a phone. But silicon isn’t expensive enough for that to make any sense as a design; such a platform would quickly evolve to have a in-station supplemental CPU-and-RAM so you aren’t constrained to the processing power you can carry in a pocket. And then it would be an obvious next step to allow the in-station CPU to operate even without a phone docked, since it’s there anyway. At which point the docking station has evolved back into a PC. So why would the PC ever go away to start with?

  88. @hari:
    “…since the actual purpose of these devices is primarily to make phone calls [more] than anything else…”

    Who says? I think you are letting the name “phone” overshadow the common uses of the device. Follow a typical Millennial around and see how often she is actually making phone calls, as opposed to texting, cruising the web, listening to music…

  89. “You may find it interesting to note that, after announcing that the WebOS hardware was to be killed off, HP also announced deep liquidation price cuts”

    I stopped in at Best Buy today to buy a peripheral. There was a sign in a prominent location the front of the store that said in so many words “We don’t have HP tablets, we aren’t going to get any more, and we can’t order one for you. go away.”

    Makes me wonder if they’ve had to field a lot of inquiries…

  90. Oh yes, they did. Quite a lot. This also gives you an answer to the question: is it possible to dethrone iPad? With $99 decent tablet? Sure.

    Of course this answer raises the other question: is it even possible to profitably create $99 decent tablet? The answer is: “probably not, but $199 is not out of the question”. We’ll see what’ll happen next…

  91. No. It would have been preemptive if that is how it started. At best, you can chalk it up to the “lessons that Apple might have learned” category — they’re doing a lot better, but not perfect.

    Apple consistently makes strategic purchases (NAND 2005, 2008, 2009, touch panels 2010, magnesium-aluminum milling 2011, high density IPS and LTPS displays 2011, etc) that constrained supplies for competitors and increased their costs. Sure, Apple competitors have found alternatives but at higher cost and lower function…rendering their products less competitive than Apple’s for that period.

    Characterizing this as some bizzaro weakness on Apple’s part is, well, bizzaro. Are you suggesting that Apple keeps relearning the same lesson over and over and THEN spends billions on component purchases that accidently constrains supplies for competitors?

    Samsung has an advantage as a component manufacturer but it doesn’t completely insulate them from this effect. Explain why Samsung went with Tegra 2 instead of their own silicon in some Galaxy S2s. You can try to argue that Tegra 2 is better or that Samsung is having yield problems on the Orion/Exynos…but having Apple consume so much 45nm capacity probably had a negative impact on Samsung.

    When you start from behind, it takes time to catch up. I think it’s pretty amazing that Google has caught up and surpassed in volume, and is now busy cementing the feature set to hold the lead.

    Copying is easier than initial development and Apple has never had volume production as the primary goal. What you guys keep insisting is an Apple failure (losing a volume lead it never had in Smartphones) isn’t.

    Having something that is good enough to ship a lot of phones now, to deny those customers to Apple, seems proactive to me.

    Not having a tablet OS ready to go when the iPad was highly telegraphed isn’t proactive. It was rushed to production for the Xoom and didn’t deny any customers to Apple given that Apple has sold every iPad 2 it has produced.

    So if your country is overrun by what appears to be a superior force, and you have some cheesy old munitions available now, but can build some better stuff in 6 months, Boyd would suggest you do nothing now? You’re right — he is rolling over in his grave. How fast can you make him spin?

    Boyd would likely suggest you failed your OODA cycle by getting caught flatfooted and not starting your new production 6 months ago. Scrambling is an indicator that the bad guy is inside your effective OODA cycle.

    Conflating plans and products is not very smart. Really good execution does not at all depend on starting off with the best product.

    Plans ARE the product in the case of military planning. This is not conflating plans and products. Really good execution does not depend on starting off with the BEST product but that’s NOT your characterization. Apple also doesn’t start with the BEST product and evolves them as they go. What they don’t do is start with a SHIT product.

    Yeah, I read that. Looks like what google’s doing to me too, but again, what do I know?

    Really? Because Google’s strategy at this point appears to me to be “throw a lot of stuff against the wall and see what sticks”.

    Look at the google+ ramp.

    Look at the Google Wave ramp. Oh wait…

    But in other markets they haven’t managed those kinds of tricks.

    What other markets? There aren’t many (any?) markets that Apple has entered and hasn’t been successful. Exactly what is the litmus test here? Nice that you characterize what Apple does as “tricks”.

    No, it’s not a simple linear loop, and google greatly complicated the loop and split it into a zillion sub-loops with Android.

    Somehow, I don’t think you actually understand the OODA loop concept.

    And this particular goal will be realized soon enough by multiple operating systems. So that can’t be the only goal. In fact, that’s not a goal at all — it’s a means to the goal, and one that the enemy capability will soon have as well.

    Of course this is an intermediate goal and not an ultimate goal. The argument was that there are no desired end states in business as opposed to military operations. I pointed one out.

    Sure, but Google’s goal with Android (which is a real goal, not a method for achieving it) is to make sure that network effects are never put in place that allow Apple or anybody else to place a toll booth between them and the web public, and from their perspective, this could wait.

    We’ll see if this goal is ultimately achieved. Arguably this is also an intermediate goal to retaining domination of search revenues.

    Personally, I view Android as another attempt for Google to find a second successful profit center. Apple has 3 and a half (computers, ipods, iOS and a half for iTunes Store and other random things like aTV).

    Ah, but Honeycomb demonstrations probably slowed iPad takeup.

    Part of market share is selling. Another part of market share is keeping your competitor from selling.

    If your competitor has sold every unit it can produce you haven’t slowed them up at all.

    And to the extent Apple has those, we’ll probably soon see some almost unrelated product come out of nowhere.

    You mean like the iPhone 5 and iPad 3?

    If very very lucky Apple MIGHT get to define a new market in a few years. Unfortunately, I doubt Jobs will be around long enough for them to get another major category changer.

    This I can agree with, to a certain extent. Having said that, I don’t think Apple was expecting the capabilities of the low-end smartphone market to rise, and the price to drop, quite as rapidly as happened. So I think the lawsuits are about Apple seeing that they are on the way to being relegated to niche status long before they were prepared for that fate.

    Apple couldn’t see that low end smartphones, which existed when the iPhone launched, would be cheaper than the iPhone? That their capabilities would increase as a result of the iPhone launch was a given.

    The lawsuits aren’t about Apple being “relegated to a niche status” because Apple has ALWAYS been in a “niche status” but a desire not to have folks freeload off Apple’s investments. Especially if they were serving on the Apple board.

    At what point did Apple have majority share of the smartphone market? Never. It has always been a “niche” player at the high end. The thing that will continue to hold it there is that Apple can’t scale large enough to ever dominate the phone business and is unlikely to try. There’s no reason to build a cheaper iPhone line to attempt to address the entire market when you’re already production and distribution limited on your high margin, high ASP product. Apple isn’t Nokia or Samsung.

    Tablets might be a different story. The volumes there are lower and much more computer/ipod scale. I can see Apple making the attempt here and scaling the iPod touch up to compete against $200 Android tablets and compete across the entire tablet market.

  92. @khim I like the Nook…and I just bought a TouchPad for $150. That $199 price point has a lot of appeal as an impulse buy/upper end birthday/christmas gift price.

    I can see a $299 7″ 16GB iPod touch compete well with $199 android tablet. Apple would likely have to drop the iPod touch base price down to $199 for their lowest model.

  93. “The lawsuits aren’t about Apple being “relegated to a niche status” because Apple has ALWAYS been in a “niche status” but a desire not to have folks freeload off Apple’s investments. Especially if they were serving on the Apple board.”

    Free markets are about people doing better what you do getting a chance of outselling you. Ideas cannot be patented. And a tablet computer is not an “invention”, but an idea. An idea that has evolved without interruption from Sumerian clay tablets.

    What you say is that it should be illegal to take an idea and implement it yourself. What is the difference with a full blown plan economy?

    Whatever eric wrote in “Homesteading the noosphere”, ideas are not land you can grab and fence in. Ideas live in heads and forbidding the execution of ideas because “I thought of it first” is the ultimate in censorship. A legal monopoly on ideas is to the detriment of everybody.

  94. @Patrick Mapuin “The answer to that question is obvious. And google-branded public internet terminals could certainly be cheap and trusted. But absent that kind of juggernaut entry into the market, which would you be more likely to trust — a specialized secure keyboard interface, or some white box computer running a random browser? Even if the guy at the internet cafe seems to know what he’s doing, do you trust his google customization?

    Or let me put it to you another way: what if Microsoft got into the market with a branded internet terminal? Trust or not trust?”

    Personally, I think this docking/internet terminal idea is a niche space as far as public cafes go. The docking idea sounds like a big mess to me. “this isn’t working with my device X” “This connector is broken” “My docking cable is missing” “This file should be available, but I can’t see it” etc

    A simplified version of what internet cafes do now makes so much more sense. Instead of a PC running Windows and a browser, you get a cheap, small piece of hardware with no accessible ports and a locked down version of Chrome OS that erases all user data for each session.

    But for either scenario (docking or not), some sort of setup/strategy has to be created and promoted – it doesn’t exist now. I’m just picking the one that’s easier/cheaper/more compatible.

  95. Free markets are about people doing better what you do getting a chance of outselling you. Ideas cannot be patented. And a tablet computer is not an “invention”, but an idea. An idea that has evolved without interruption from Sumerian clay tablets.

    What you say is that it should be illegal to take an idea and implement it yourself. What is the difference with a full blown plan economy?

    Apple didn’t patent tablets or ideas. So no, I didn’t say it should be illegal to take ideas and implement them yourself. Just the specific solutions to problems that Apple has been granted patents and you’re free to develop your OWN solutions for.

    Nice strawman.

  96. “[Best Buy’s demand for the discontinued HP tablet] gives you an answer to the question: is it possible to dethrone iPad? With $99 decent tablet? Sure. Of course this answer raises the other question: is it even possible to profitably create $99 decent tablet? The answer is: ‘probably not, but $199 is not out of the question’. We’ll see what’ll happen next…”

    Well, we now know that there is little demand for a $500 Android tablet, and tremendous pent-up demand for a $99 tablet. Clearly, somewhere between these two points the demand curve starts to create a sizable market. My gut says that $200 would still generate substantial sales, but we won’t know until someone announces one at that price. Given that the Nook sells for $250, I’d be very surprised if if wasn’t possible to produce a $200 version similar to the Nook (but shipping with full Android rather than just e-reader software) some time in the next few months.

    Christmas will be very interesting this year.

  97. I doubt B&N is making much money if anything on the Nook Color. The BOM is estimated to be around $200. I expect the business model is to make money on selling books.

    That there are and will be more $200 Android tablets is a given…but so far I haven’t seen one with as nice a screen and build quality. A $200 tablet on the same level as the Touchpad? Not for a while I don’t think.

  98. The big wildcard, I think, is Amazon. Unlike HP, they have the possibility of the razorblade model, selling an iPad-class tablet at a loss or cost to drive sales of digital media (ebooks, but also songs, video, and apps).

  99. @nigel
    “Just the specific solutions to problems that Apple has been granted patents and you’re free to develop your OWN solutions for.”

    I do not trust a patent system that honors patents on “swinging sidewards on a swing” and “combing hair over a bald spot“. It also rewards patents on mathematical formulas “when executed on a computer”, including the XOR.

    Apple has used patents on using several fingers on a touch screen and many other non-sense. Nothing illegal, but all designed to kill competition. Apple shifted towards “If you cannot innovate, litigate” mode some time ago.

    So yes, the patent office rewards patents on ideas “when execute on a computer” and Apple has patented many ideas. You might call them “solutions”, but they are still just ideas.

  100. @Winter ” Apple shifted towards “If you cannot innovate, litigate” mode some time ago.”

    They are agressive litigators, I agree. It sucks. Goole seems to be the only big company not aggressively defending their patents (possibly because they aren’t in a strong enough position to?). But Apple are also aggressive innovators. Ever heard of the iPad? iTunes match? Lion is quite innovative with Apple’s move to remove the need for consumers to directly deal with the file system. iOS 5 and OS X Lion’s Cloud enablement will push all storage to the cloud. etc.

  101. @phil No, Apple is not innovative…folks here have some other term for it which I forget. Optimize or evolve or something. The iPad isn’t revolutionary but an inevitable evolution of the tablet concept. Etc.

    And Samsung isn’t copying Apple. It’s just parallel evolution that all the buttons are in the same place as a result of functional design.

    After Microsoft taking the Mac UI I think that Jobs and Apple have this constant refrain: We welcome competition. Just do it your own way. Which is why Jobs announced in 2007 “we patented the hell out of it” regarding the iPhone.

    Are there bad patents? Sure. Are there good patents? Again, sure.

    But there are a lot of folks out there that don’t respect IP rights except when it favors them.

  102. Nigel,

    > No, Apple is not innovative…folks here have some other term for it which I forget. Optimize or evolve or something. The iPad isn’t revolutionary but an inevitable evolution of the tablet concept. Etc.

    I don’t think anyone is saying that. I think Apple is innovative in an evolutionary way. You can get back to me when Apple develops something like a transistor or a laser. Those are revolutionary innovations.

    > And Samsung isn’t copying Apple. It’s just parallel evolution that all the buttons are in the same place as a result of functional design.

    If you think the placement of less than ten buttons requires a significant (and patentable) amount of innovation, I understand why you think Apple is revolutionary, but by that criteria, so is everyone, and I can patent this sentence. Which is exactly why I don’t want you running the patent office.

    > After Microsoft taking the Mac UI I think that Jobs and Apple have this constant refrain: We welcome competition. Just do it your own way. Which is why Jobs announced in 2007 “we patented the hell out of it” regarding the iPhone.

    Do you mean the Xerox UI?

    > Are there bad patents? Sure. Are there good patents? Again, sure.

    No, they are all bad.

    > But there are a lot of folks out there that don’t respect IP rights except when it favors them.

    But most of them aren’t commenting here. You have wandered into a nest of people who don’t want patents for anybody, and who want copyright strictly limited. I actually respect Apple’s trade dress rights, but one of the problems of functional designs is that trade dress is precluded.

    Yours,
    Tom

  103. @nigel @phil
    Fashion designers are often revolutionary and if you overstretch the meaning some even call them “inovative”. But they can work wonders without patents and sueing competing desi
    Most “inovations” of Apple are of the fashion, pret a porter kind. And like those who cannot keep up with other fashion houses, Apple wants to use patents to avoid competition.

    The funy thing is that those who copy and “steal” liberally from everyone are always accusing FOSS, who avoid patents like the plague.

  104. Do you mean the Xerox UI?

    No. The Mac UI. Almost all of the tropes of the most common style of GUI are Apple inventions. As I said before — your favorite application has a File / Open menu item because of the 128K memory limit of the original Mac. (The original idea was to switch the user back to Finder when they wanted to open a new file; this broke down when it was determined that Finder and a running application couldn’t coexist in memory and was never fixed in subsequent System versions. due to inertia.)

    The history of the relationship between Xerox, Microsoft, and Apple is a bit more intricate than most geeks suppose but it boils down to that Apple licensed Xerox’s technology legitimately — and improved significantly on it. Microsoft simply copied the Mac interface wholesale, without making substantive improvements (indeed, they went backwards) and without licensing the technology.

    So no, Steve didn’t try to steal Xerox’s TV only to find that Bill stole it first. Steve chipped in for Xerox’s cable bill in exchange for getting the TV every other weekend — and Bill stole it.

    1. >Almost all of the tropes of the most common style of GUI are Apple inventions.

      That is not true. Most of them were already present on the Dorado, c. 1973.

  105. @Winter
    what qualifies as innovation then? Either you are in denial from Apple hatred or you have a very high bar.

    1. >WSJ says Sprint is getting the iPhone

      Yeah, if the Verizon rollout is any guide this might improve Apple’s market share by – wait for it – a whole 1%.

  106. As I said before — your favorite application has a File / Open menu item because of the 128K memory limit of the original Mac.

    I’m pretty sure I was using DOS apps with File / Open menus (or File / Retrieve, etc.) well before the Mac existed.

    /frmyfile.wks~

  107. > Yeah, if the Verizon rollout is any guide this might improve Apple’s market share by – wait for it – a whole 1%.

    Yeah, well, for Sprint, having Android wasn’t enough, apparently.

    In the second quarter, Sprint blamed a decline in its contract subscribers on more pronounced “competitive headwinds,” most prominently, “the first full quarter both major competitors offered the iPhone.”

    source (WSJ)

  108. phil, do you disagree with the actual measured market share over the period of Verizon adding the iPhone to its network or not?

  109. “phil, do you disagree with the actual measured market share over the period of Verizon adding the iPhone to its network or not?”

    esr didn’t put a time range on his prediction. So maybe he’s right – he can adjust his time range to whenever the 1% pickup happens and claim he is anyway…..

    But that’s a good question. I do think it’s a bit odd that Verizon sold 4.5 million iPhones in less that 2 quarters and comScore only shows maybe a 1-2 % increase in market share. I mean 4.5 million is 1.5% of the entire US population. And the US smartphone installed user base is much less than 307 million (the comscore script appears hosed, so I can’t see the latest report to get a count).

    I can think of a couple possible explanations:
    1) comscore is wrong
    2) a lot of iPhones were retired
    3) the 4.5 million # is BS

    1. >esr didn’t put a time range on his prediction.

      I did have a specific one in mind. Apple’s market share has hung within 1% of 25.0% from Dec 2009 (and possibly earlier) to March 2011. From February to May we’ve seen it climb to 26.6%, then level there in June. So, allowing for statistical fuzziness, we’re looking at a post-iPhone-4V gain between 0.6% and 1.6%, mean 0.8%. Of course Sprint doesn’t have as large a userbase, so at the same rate of conversions the share gain will be smaller. Topping out at 27.4% seems like a fair guess.

  110. I had the wrong link, here is the right one:
    http://www.catb.org/esr/comscore/

    So comscore is claiming the US iPhone user base in December was 15.8 and as of June it was 20.88. Could someone reconcile this with Verizon reporting 4.5 million sales (2.2m 1st quarter, 2.3m 2nd quarter) and AT&T claiming 7.2m (3.6m + 3.6m). If all these #s are true, that’s a lot of iPhones being retired.

  111. @Jeff Read
    “No. The Mac UI. Almost all of the tropes of the most common style of GUI are Apple inventions.”

    It is rather comical, but almost nothing in Apple’s GUI proved to be protected.

    Apple lost all claims in the Microsoft suit except for the ruling that the trash can icon and folder icons from Hewlett-Packard’s NewWave windows application were infringing.

    http://en.wikipedia.org/wiki/Apple_Computer,_Inc._v._Microsoft_Corporation

    See for more:
    http://lowendmac.com/orchard/06/apple-vs-microsoft.html

    @phil
    “what qualifies as innovation then? Either you are in denial from Apple hatred or you have a very high bar.”

    Innovation is from the stem “nove” new. It should be something NEW in capitals.

    The Apple II was undeniable NEW, as it was arguably the first personal computer. The GUI of the Mac was smashing, but not really new as everything in it was already designed by Xerox (the Star was NEW). The iPod touch was new in the fashion sense. It could have had technical innovations under the hood, but the rest was as new as this years Milan fashion week.

    Yes, I believe words should have a limited scope so they retain some meaning.

    In the end, both phil and Jeff Read claim Apple should be legally able to prevent others from implementing ideas. Ideas that Apple took from others in the first place. In the fashion world, Apple would try to close down all European fashion shows because all those designers “stole our precious IP”.

  112. @phil and Jeff Read
    Case in point that Apple gets ideas from others and them uses the legal system to block others from using the same ideas. Apple doesn’t invent, but simply markets other peoples ideas. And then sue others for copying the same ideas, “solutions”, or whatever.

    Samsung cites Kubrick’s ’2001? as legal defense in Apple patent battle
    http://venturebeat.com/2011/08/23/samsung-kubrick-2001-apple-patent-war/

    Attached hereto as Exhibit D is a true and correct copy of a still image taken from Stanley Kubrick’s 1968 film “2001: A Space Odyssey.” In a clip from that film lasting about one minute, two astronauts are eating and at the same time using personal tablet computers. The clip can be downloaded online. As with the design claimed by the D’889 Patent, the tablet disclosed in the clip has an overall rectangular shape with a dominant display screen, narrow borders, a predominately flat front surface, a flat back surface (which is evident because the tablets are lying flat on the table’s surface), and a thin form factor.

    Yes, Apple caught with their pants down.

  113. The full comment is awaiting moderation, but you can use a search engine yourself:

    Samsung cites Kubrick’s 2001 as legal defense in Apple patent battle

  114. @Winter

    “In the end, both phil and Jeff Read claim Apple should be legally able to prevent others from implementing ideas.”

    Really? News to me.

    Back to reality – I think software patents should be abolished.

  115. OK, now I’ve seen everything. From a WSJ opinion article:

    “Google has come from nowhere to build a dominant position in smartphone software based on tying a free Android to Google search advertising (which is fine) and, arguably, by helping itself to seminal Apple innovations that created today’s smartphone industry (not so fine).

    “Google’s approach implicitly assumes that nobody has a right to exclude Google from use of their intellectual property. At best, after litigation, they might have a right to be compensated by Google.

    “But this is not how the patent system is supposed to work. Let us understand that Google’s purchase of Motorola is the purchase of a bargaining asset; it does not automatically put Google in the right. Apple, as a patent holder, has every right to seek to preserve exclusive use of its inventions. In its eBay decision, the Supreme Court allowed that the possibility of ‘irreparable harm’ might justify banning an infringing product outright, equivalent to an ITC import exclusion.”

    I completely disagree with his interpretation of the law.

    http://online.wsj.com/article/SB10001424053111903327904576526130093390612.html?mod=WSJ_Opinion_LEADTop

  116. Jobs is throwing in the towel.

    End of an era. For my entire lifetime he was involved in the industry, actively breaking molds and setting trends — whether at Apple, NeXT, or the Voltron-like combination of the two.

    Best of luck, Steve.

  117. It’s sad, really. Battle for the future of the phone is just starting (till now iOS and Android killed all these also-run systems, they have not fought for supremacy directly). And now, even if Android will actually win everyone will just say “Oh, yeah, that’s because Steve’s brilliance is no longer with us”…

    In a sense he did what Bill Gates did: he left when his early decision produced maximum effect yet still have not produced coherent backlash. So now people think Microsoft is in trouble because Steve Ballmer is an idiot while he’s fighting a losing battle because Bill Gates managed to alienate most of the big partners. The same will probably happen with Apple – and again, it will be blamed on Tim Cook and the rest of the Apple while in fact it’ll be result of Steve’s decisions.

  118. Compared to Gates, Steve Ballmer is an idiot. He is famously loud, aggressive, and pushy, and doesn’t seem to know how to build the business rapports that Gates did. He also lacks the technical chops that Gates had; whether you like him or hate him Gates knew the business of computing better than almost any other CEO in the field — including Jobs — because he was a hacker himself until the early 1980s.

    That said, the best release of Windows ever — Windows 7 — occurred under Ballmer’s watch. The reputation of Windows had been so screwed by that point that even a release like 7 couldn’t help it regain ground against the Linux and Mac OS X encroachment.

  119. So the HP tablet sold off so well discounted, and because HP has parts in the supply chain that they’ll have to pay for anyway, HP will actually make more of the cancelled tablets.

    Yet another sign of incompetence.

  120. WebOS is really good, the problem is: a good part of the market is offline – in Italy for example nobody would buy something that constantly needs to be on line because broadband access is stricly limited, has poor quality and costs a lot.

Leave a Reply to Steven Ehrbar Cancel reply

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