Greed kills: Why smartphone lock-in will fail and open source win

In a previous post, How smartphones will disrupt PCs, I explained how and why I think small, ultra-portable, general-purpose computers that we’ll think of and use as “smartphones” are going to displace the PC. I promised then to explain why the software of these devices will be open source.

Go read Androids Will Challenge the iPad. It isn’t about smartphones, but the logic that will break the iPhone business model is clearly set out in it for anyone who’s paying attention. What we’re about to see in the smartphone and tablet markets is a repeat of the way the IBM PC shouldered aside the Apple II after 1980. Google’s deliberately slow-balled launch of Android via the G1 was just prelude; it’s with the Motorola Droid, the unlocked Nexus One and the generic Android tablets that the game begins in earnest.

After 1981, IBM seized the lead in the personal computer market by exploiting two advantages over Apple. The first was marketing and sheer size: IBM’s brand had a lot of power, and IBM’s run rate allowed it to fund product development and reap economies of manufacturing scale on a scale Apple couldn’t match. Twenty-nine years later Google has at least as cool a brand as Apple, more financial mass, and more engineers. And that’s all I need or want to say about the biz-journalism end of things.

IBM’s second advantage was openness. The PC was designed to be kit-bashed; it became the hardware platform that launched a thousand hardware startups and, effectively, the entire PC industry as we know it. The manuals included a BIOS listing, the bus specification was public; anyone could plug in, and did. IBM’s own attempt to close the platform a few years later, the PS/2, was a failure that sank almost without trace.

Fast forward three decades. The commoditization of hardware that the PC pioneered has succeeded so completely that all smartphones are built by anonymous OEMs on the Pacific Rim and the real competition has shifted from hardware to software. Forget details like smartphone vs. tablet form factors and which handset manufacturer is the belle du jour: the real competition is the OS X ecology vs. the Linux/Android ecology.

And isn’t it entertaining, boys and girls, how thoroughly Unix won? Both OS X and Android are Unix underneath. Windows Mobile is hemhorraging market share and even the most notoriously Microsoft-gullible elements in the technology press can see it’s a no-hoper.

On its way down, Windows Mobile gives us an object lesson that allows us to predict how the OS X/Android war will end. It’s the same lesson that the Apple II vs. PC war taught, and it’s heightened by the way that Microsoft has (just barely) managed to hang onto a dominant position in desktops — by allowing lots of third-party developers to make money from that dominance. (Update: I wrote “just barely” because Microsoft has had to give away most of its profit margin to maintain share.)

IBM won its battle for ubiquity over the Apple II because it was willing to give up control, to let third parties (including Microsoft and the peripheral-card industry) make most of the money and content itself with a tiny sliver of a rapidly expanding pie. Microsoft kept Windows viable on the PC desktop by yelling “developers, developers, developers!” and conceding third parties a huge share of that pie.

And now? Google is willing to let handset makers, telecoms providers, and third-party developers capture most of the overt value of the Android market. Google can give all that prompt revenue away because everything it’s doing in this space is actually funded the same way its search-engine business is; by the volume of consumer attention Android devices will bring to its advertising. Apple, on the other hand, acts as a very controlling gatekeeper of its products — requiring (and insisting) that it’s going to capture most of the profit margin for itself.

Apple’s problem now is the same as the Apple II’s problem in 1981: in markets reliant on a vigorous ecology of allies to add value to a product, greed kills. Gatekeepers lock out potential allies; walls limit the garden’s growth. Ask any strategic planner at a telecoms provider or handset maker why Windows Mobile failed and you’ll hear the same thing: they saw what happened to IBM and swore they’d never let Microsoft talk them into being its butt monkeys. Windows lost out to Linux in the medium and high-server market, because third-party developers are much less important there; customers tend to be writing their own bespoke software, so server Windows isn’t pinned in place by a huge collection of allies. There’s a harsh tradeoff between control and ubiquity; the original IBM PC and desktop Windows got on the right side of it, but Windows mobile got on the wrong one.

The competitive dynamic between Linux/Android and OS X can be understood in the same way. OS X is playing a control game and Android a ubiquity one. We can expect the outcome to be the same: when the bazaar meets the walled garden, the walls will eventually come down, crushing the life out of the garden.

This is why Symbian is now open-source in spite of having no inheritance from Unix-land; its backers have figured out that a control strategy collects short-term gains over a ubiquity strategy but simply cannot compete in the longer term against open-source Android and open-source Maemo. Apple will learn this, to its cost, too. Because Steve Ballmer may be an evil maniac, but when he yelled “developers, developers, developers!”, he was right. In the war for market-share, allies are better for your long-term prospects than walls, and ubiquity will always eventually triumph over control.

UPDATE: I wish I had read Where Android beats the iPhone before I wrote this. It bolsters the argument pretty effectively.

UPDATE: And here’s Gylnn Moody dispelling some anti-Android FUD.

125 comments

  1. I prefer superior UI, which open source projects sadly seem to be completely incapable of producing.

    Windows Mobile’s problem was not non-ubiquity, it was horrible interface.

    I wouldn’t apply a model that works great for backend server and programmer stuff (open source hack whatever you want; programmers can make their own tools the way they want) to a radically different market (end-users who just want it to god-damn work, and work right, the first time).

    Compare Android’s existence of a task-killer or scheduled-reboot app, with the iPhone. I think you’re making the classical “geek” mistake. The market has actually been happily rejecting the bazaar as long as the walled garden provides superior experience. (I speak of the iPod, primarily, but note that Apple’s computer sales are on a continual upswing, while linux on the desktop… well, there’s always next year, right? Just like every year since 1996 or so.)

    And given what the bazaar’s typically been producing in terms of user experience, I wouldn’t be selling my Apple stock, if I owned any.

  2. (And I should note that I’ve run linux for about 15 years now, quite happily, for fun and as a server solution, and run OSX and Windows every day, so I don’t have some fanboy dog in this fight.)

  3. Wait what? Between iPhone OS and Android, how is the Windows Mobile platform (the one where you can install anything without having to go through an “appstore” the CLOSED one?!

    I am honestly trying to understand your logic here and it seems to me that HTC+WM is the IBM PC of old: install anything, run anything, totally open. Still completely losing to iPhone, RIM and now Android. But why?

  4. >it’s heightened by the way that Microsoft has (just barely) managed to hang onto a dominant position in desktops

    I find this phrase bizarre, seeing as there has been no credible threat to Windows on the desktop within the last decade. Linux is not even in the game and never has been.

  5. Microsoft might have actually struck the perfect balance between control and ubiquity once again with Windows Phone 7 Series. They are not allowing anyone to make different skins for it and are requiring very specific hardware specs to ensure a consistent experience. OEMs can compete on making the specified hardware as cheaply as possible. So it undercuts the iPhone on price and might outdo Android on user experience.

  6. Sigivald I think you are missing the point a bit. This is not open source vs propietary. Android is backed by a company, and a very big one. Its interface is already quite slick and will only improve. This is about corporate strategies, ones favouring a bazaar approach vs others favouring control.

  7. IBM won its battle for ubiquity over the Apple II because it was willing to give up control, to let third parties (including Microsoft and the peripheral-card industry) make most of the money and content itself with a tiny sliver of a rapidly expanding pie.

    They weren’t willing to give it up. They simply neglected to assert it. They figured the personal-computer market was going to be minuscule compared to their mainframe business, but they needed to act fast to get their foot in the door of that space, and so they basically cobbled something together from off-the-shelf components. They copyrighted the BIOS, hoping that would save them, but as soon as that was clean-roomed the barn doors were open and everybody and their dog was churning out PC klones.

    I think they were hoping that Micro Channel would be their do-over. We all know what a flop it turned out to be under the staggering massive weight of ubiquitous cheap-ass PC compatible hardware.

    Regrettably, Sigivald is right: remember, unlike its competitors such as Commodore and Atari which also offered superior UI and superior engineering prowess to commodity PCs, Apple is still alive and still operating in that space. Virtually all of the research into how humans actually think and what is the path of least resistance in the use of a particular device was done at Apple. With that kind of advantage they will be selling world-class products and be the primary authors of the grand narrative of how we interact with digital technology for years to come.

    Oh, and in a world of uncertainty, malware, and good old information overload, a vendor-controlled restricted App Store has yet to be trumped in effectiveness as a quality and consistency filter by any open solution.

  8. Jeff you seem to be ignoring the fact that Apple lost the battle for personal computers even though they had an early lead, because an open solution just overwhelmed the possibilities of a single company. They have an early lead now in the mobile space. But they are basically using the same technique so chances are will follow the same path, which is not dissapearing, but losing market dominance.

    Of course they will continue to be a player, have a niche and as you say produce world class products with superior design and be a center of innovation. But the bulk of the market not belonging to them is what is being under discussion here, I believe.

  9. Jeff you seem to be ignoring the fact that Apple lost the battle for personal computers even though they had an early lead, because an open solution just overwhelmed the possibilities of a single company.

    Linux — the really open solution — had an early lead against Mac OS X and a worldwide community of developers and supporters, both individual and corporate. But Linux on the desktop is losing ground to the Mac, even amongst hackers. Why? UI and UX.

  10. Android is backed by a company, and a very big one.

    It is backed by sixty-five companies of varying size, hoping to combine to form a Voltron that will slay the proprietary handset platform robeasts, with Google forming the head.

  11. We agree on that. It’s just that I think that’s not the point. We are not arguing about open source vs propietary. We are arguing about one strategy followed by a bunch of corporations led by google which opens the platform for many to do business in VS another strategy which is centralized in a single company (Apple).

    There’s parallels with the open source world but it’s not the same thing. Nobody spends enough resources and effort in improving Linux on the desktop (you could say Ubuntu but it’s clearly not enough). On the server though it’s a different matter, there’s investment there and it pays off. But here we are talking about google spending on the interface too. And from what I’ve seen, Android has nothing to envy in MacOSX.

  12. >And from what I’ve seen, Android has nothing to envy in MacOSX.

    That’s not quite true. The Mac OS X interface is a better work of pure art than Androids’s is. Functionally, though, not a lot of difference.

  13. >I am honestly trying to understand your logic here and it seems to me that HTC+WM is the IBM PC of old: install anything, run anything, totally open. Still completely losing to iPhone, RIM and now Android. But why?

    Because Microsoft had a history, and every time it thought it might have the leverage to enforce running only signed binaries it started muttering about doing that…and, as I reported, the telecoms execs all decided they didn’t want to have Microsoft do to them what it did to IBM.

  14. While I agree with the gist of esr’s piece, the part about Unix winning out and Windows Mobile losing rings hollow. I think some form of Unix will win out eventually, but it won’t be iPhoneOS or Android. As for open vs closed, Android is a little more nuanced than that, they explicitly chose the Apache license because they wanted to allow handset vendors to take parts of the platform closed if they wanted to. I have never seen a good writeup of how prevalent closed parts of Android handsets are, but Android is hardly a resounding endorsement of pure open source for this reason. As for Symbian going open source, that strikes me as their backers just throwing in the towel on that platform and moving to other linux-based platforms, as Nokia and others seem to be doing, particularly because they chose the badly-written Eclipse License for Symbian.

    The real solution for malware is a decentralized application/developer reputation system, not the central control of the Apple App store that is rife for abuse and incompetence. I was reading the Android SDK section on distributing apps recently and was pleasantly surprised with the way Android provides for signed apps and open distribution, which a reputation system can be built around. Such app/developer reputation systems have always been the solution to malware/security issues: it is a running joke that nobody like Microsoft or Symantec has implemented anything like this, though perhaps there are patents holding it back that I’m unaware of. ;)

  15. >That’s not quite true. The Mac OS X interface is a better work of pure art than Androids’s is. Functionally, though, not a lot of difference.

    We are talking only the portion of MaxOSX that runs in the iphone/ipad of course. Still, Apple’s is older and more mature and Android is still quite good and improving. Good enough for companies to run it in their phones.

  16. >I find [(just barely)] bizarre, seeing as there has been no credible threat to Windows on the desktop within the last decade.

    Yeah, now look at one of their quarterlies and figure out how much margin they’ve had to give away to maintain market share. I did this exercise nine months ago with a friend who’s better at reading financials than I am. The monopoly is still in place, but it’s paper-thin.

  17. Well Eric how do you account for the fact that Apple has been doing so well, not just in phones but in desktop computers as well?

    Apple’s Macs are even more closed than Windows machines, but have been gaining significant market share.

    As a holder of a large position in Apple stock for the past year, I can only say giggity giggity giggity!!

    Yours, Darren

    1. >Well Eric how do you account for the fact that Apple has been doing so well, not just in phones but in desktop computers as well?

      I never said the control strategy couldn’t capture short-term gains.

  18. >[IBM] weren’t willing to give it up. They simply neglected to assert it.

    That’s a little complicated. The original IBM PC development group had an active open-it-up agenda; if they hadn’t, they wouldn’t have published the BIOS source and the schematics.

    Later, IBM senior management decided the inmates had been running the asylum and air-dropped in some managers whose main goal was to prevent the PC from climbing the power curve and cannibalizing sales of a now-forgotten machine called the Datamaster. But they didn’t take direct aim at the open strategy, probably because they didn’t grasp its implications; it took nearly four years for IBM to decide that had been a bad idea and launch the PS/2 line.

    So what IBM was willing to do depends on what group inside IBM you’re talking about.

  19. >>> I never said the control strategy couldn’t capture short-term gains.

    But there seems to be a pattern here. The closed source world comes up with something, and it leads the way.

    Then at some point the open source world copies it and puts it out of business.

    But when does the open source world innovate something?

    BTW, as everyone is probably aware, IBM is no longer even in the PC business. They sold the PC business to Chinese company Lenovo.

    What good is setting a standard if you are shut out of it?

    1. >But when does the open source world innovate something?

      It’s hilarious to me when people ask this question, because I remember when the knock on hackers was that we could innovate but never deploy. OK, we didn’t call ourselves “open source” back then, but it was the same culture doing the same sorts of things.

      Dude, we imagined and designed the Internet. And the World Wide Web. And yes I mean “we”: I had a tiny but not completely insignificant part in that back in ’83 when the design of DNS was being finalized. Unix, too; I wasn’t in on that early enough to claim a piece, alas.

      Geez. That ought to be enough innovation to banish this pig-ignorant question for, oh, at least the next century.

  20. I’m a paying customer of Apple, via my second iPhone (or at least the company that pays for mine is). I was extremely excited to have in one device an audio player, phone, and most important of all, an untethered surfboard for the web. I love to go on long walks and read.

    I would have purchased an iPad, regardless of price, if I had been able to surf and make phone calls. If the android pad will do that I will buy one. Otherwise I won’t.

    Why? I love to read, and I my health won’t deteriorate if I have a screen large enough to read comfortably while walking. I don’t have to skip very many doctor visits to justify the price of an iPad. In fact, if medical costs are factored in, an OctoCore Xeon Extreme Edition OLED iPad with 22 Kg of Li-Ion batteries still looks cheap, if it will keep me moving and outdoors.

    But the current iPad has no phone access, so I passed.

  21. @esr — At a component level, I agree: the open-source compiler, web browser, etc will beat the closed one. The problem that open source doesn’t solve is integration, which is what makes the Apple experience compelling. There’s really nothing special about MacOS, or Safari, or iPhoto, or the Mail reader. It’s the fact that every Apple thing I own Just Works with everything else. It’s not possible to maintain that level of integration without having control over the components — otherwise something you depend on may go off in a direction not compatible with your own.

    The question is whether Apple can maintain a sufficient level of innovation across an ever growing portfolio of components to keep up with the open source marketplace. The value of integration is high enough that their components don’t all have to be best-of-breed, but not so high that they can fall arbitrarily far behind. So far they seem to be pulling it off.

    1. >The question is whether Apple can maintain a sufficient level of innovation across an ever growing portfolio of components to keep up with the open source marketplace.

      And the answer is “no”. They’re fighting a control problem the difficulty of which rises with the square of the number of applications. Increasing revenue will give their organization at best a roughly linear capability curve. This is the elevator effect rearing its head yet again.

  22. About walled gardens…

    What about services such as the Steam client that lets you download video games directly to your PC? They seem to be at least cousins to Apple’s App Store and I’m wondering what you think about their long-term viability.

    1. >What about services such as the Steam client that lets you download video games directly to your PC? They seem to be at least cousins to Apple’s App Store and I’m wondering what you think about their long-term viability.

      I don’t see the same kind of quadratically-scaling problem there. Games don’t generally have to integrate with each other, after all. Also games have a relative short product lifetime. These both help make the app store model work better and sustain longer.

  23. > OS X is playing a control game and Android a ubiquity one.

    I note that Apple is controlling their iPhone/iPad hardware market by (almost) freeing the attached software market. (Almost any app can be sold thru the App Store, although there are horror stories about Apple arbitrarily rejecting apps.)

    One side effect is that iPhone/iPad software development is easier because the hardware is more standardized. More importantly, Apple gets a 30% proportion of the software pie giving it a shot in the arm that IBM and Microsoft never saw. Apple also has a design quality advantage at least in the short term. Not sure how significant these effects will be in the war vs Android.

  24. I don’t think the parallel between then and now quite works, for several reasons. One is that the battle of the 1980s was a battle to sell computers, where openness is valuable to many people. This is a battle to sell appliances, and I doubt the average iPad consumer will ever want to know about or access the guts of the device.

    The other main difference is that there is probably a very large first-mover advantage in terms of the app stores (remembering that the Android app store is just as closed as Apple’s one). Ultimately it will be the apps that lock users in to a given piece of hardware, and if the apps on Apple’s service outnumber the android apps by ten to one, then it won’t make a lot of sense for either developers or users to go elsewhere. There was never any real analogue of this scenario in the 1980s, except in a short-term sense involving things like visicalc, because there was neither the developer base nor the user base to support hundreds of millions of pieces of software.

    1. >(remembering that the Android app store is just as closed as Apple’s one)

      There are already multiple Android app channels. I haven’t used any of the non-Google ones myself, so I make no representations about quality. But they do exist. This is something Apple is plain not allowing.

  25. Ok, but I think the open app channels will remain irrelevant, for psychological reasons. The killer app of these appliances is purchase-convenience. Unless someone has the resources to offer an open, free app store with the same kinds of servers and databases as the closed ones, they’ll be of interest only to us geeks. And it’s not mainly geeks who’ll be buying these devices.

  26. Software portability is another modern effect that doesn’t match the 1980’s. I bet if we compared porting even the most complex mobile app from iPhone to Android it would be trivial compared to porting Lotus 1-2-3 from PC to Apple II.

    1. >I bet if we compared porting even the most complex mobile app from iPhone to Android it would be trivial compared to porting Lotus 1-2-3 from PC to Apple II.

      I’m pretty sure this isn’t true. Android apps are written in Dalvik-machine Java (a slightly nonstandard dialect), whereas OS X programs tend to be written in C-family languages. There’s also the issue of very different native GUI toolkits.

  27. > I’m pretty sure this isn’t true.

    Yes, on reflection I agree with you. Portability can be far easier today, but tricks like basing an app on cross-platform libraries and cross-platform languages aren’t always available, and also today’s apps dwarf those of the 1980’s in sheer size and complexity.

    You can run C/C++ on Android, but only for critical functions, not for a full app. Supposedly the iPhone prefers Objective C but allows direct C/C++.

  28. But regarding the long-term market effects, it’s likely that Android will eventually support full C++. The same network effects supporting your premise will also push for wide support of all popular languages. Including Python, I hope.

    ESR says: Agreed about the network effects.

  29. I’m pretty sure this isn’t true. Android apps are written in Dalvik-machine Java (a slightly nonstandard dialect), whereas OS X programs tend to be written in C-family languages. There’s also the issue of very different native GUI toolkits.

    iPhone apps are written in Cocoa, a framework that’s so awesome people are replicating it (and Objective-C) in other languages. I wouldn’t be surprised if some enterprising company built up a framework based on Cocoa that let them get at least 90% of the way toward machine-translating their iPhone apps to Android.

  30. But regarding the long-term market effects, it’s likely that Android will eventually support full C++. The same network effects supporting your premise will also push for wide support of all popular languages. Including Python, I hope.

    It’s entirely possible that Dalvik will become an LLVM machine target, meaning Android apps can be written and compiled with clang, unladen-swallow, etc.

  31. You can drop in C++ or plain C to an iPhone app, though, as long as your I/O wrapper (which can be really simple) is written in Cocoa.

  32. Adding support for new languages to the Android is surely not insurmountable problem, but it’s a tricky one. The Dalvik VM is heavily specialized not only around Java, but around a slightly quirky house style of Java OO design that minimizes reliance on manifest typing. Other languages that target the JVM, like Jython and Clojure, have already been ported to Dalvik, but their performance is piss-poor. You could of course have other runtimes running alongside Dalvik, but then you’re dealing with the memory overhead that entails as well as the need to maintain FFIs for everything.

  33. I think it’s very foolish to make a comparison between the PC market and the Mobile market.

    You’re right about the PC market (although you really do need to acknowledge the role – no matter how much it may hurt – that Microsoft played in driving the success of the PC). But you’re absolutely wrong in trying to apply this lesson to the mobile device market, and as a result the whole premise of this piece dies on its feet.

    The problem that those of us who passionately believe in OSS and open ecosystems have is that consumers do not give a flying fig about the intrinsic moral goodness of “open”, they want a hella awesome device to play with. They don’t want to “tinker” with it, they just want it to work – This is a world away from the nightmare of device drivers, reboots, unstable software and general agony that the PC represented in the 80’s and 90’s.

    Apple was wrong in in the early days of the PC market, but you’re a long way from convincing consumers that they’re wrong about the modile device market in the 10’s

  34. “You can drop in C++ or plain C to an iPhone app, though, as long as your I/O wrapper (which can be really simple) is written in Cocoa.”

    You can drop in C++ or plain C to an Android app, though, as long as your I/O wrapper (which can be really simple) is written in Java.

  35. >>But when does the open source world innovate something?

    >It’s hilarious to me when people ask this question, because I remember when the knock on hackers was that we could innovate but never deploy. OK, we >didn’t call ourselves “open source” back then, but it was the same culture doing the same sorts of things.

    >Dude, we imagined and designed the Internet. And the World Wide Web. And yes I mean “we”: I had a tiny but not completely insignificant part in that back in >‘83 when the design of DNS was being finalized. Unix, too; I wasn’t in on that early enough to claim a piece, alas.

    >Geez. That ought to be enough innovation to banish this pig-ignorant question for, oh, at least the next century.

    Actually this is a really really fair question, and to simply say “we invented the internet” simply proves that open ecosystems are really really good at delivering totally non differentiating, horizontal function. Fair enough! But what about innovation in… say… office productivity?

    1. >But what about innovation in… say… office productivity?

      *boggle*

      One of us invented the spreadsheet.

      (I didn’t have anything to do with this one, either. But I know the guy who did it slightly.)

      Now, would you like to try moving the goalposts yet again, or are we done here?

  36. @Daniel Franke:

    Adding support for new languages to the Android is surely not insurmountable problem, but it’s a tricky one.

    I think that things have change a bit with regard to Android languages since you last looked at it. Have a look at the Android Scripting Environment, which gives you very simplified access to nearly the whole Android API using Python, Perl, Lua, BeanShell or JRuby.

    Of course, I’m rather partial to Python, myself.

  37. @esr

    >One of us invented the spreadsheet.

    I raise your *boggle* and add a *spits coffee*

    Good to know that the only two claims to OSS innovation you can point to date back three decades… you’re really selling that OSS innovation man!

    1. >Good to know that the only two claims to OSS innovation you can point to date back three decades

      Silly person, we’re still inventing the Internet. Every single day. Well, every time we publish an RFC, anyway.

      You’ve just learned to take year-over-year innovations for granted because ours don’t have marketing departments.

      Corporate “innovations” are the stuff that didn’t require enough bravery or imagination to need the hackers to do it.

  38. >> Now, would you like to try moving the goalposts yet again

    > Good to know that the only two claims to OSS innovation you can point to date back three decades

    I guess we can take that a yes…

  39. Dangerdad: The NotionInk Adam is an iPad killer if it can come to market at $300-$500.

    Sure, in the same sense that any ThinkPad T-series is a Macbook killer.

    esr: I’m pretty sure this isn’t true. Android apps are written in Dalvik-machine Java (a slightly nonstandard dialect), whereas OS X programs tend to be written in C-family languages. There’s also the issue of very different native GUI toolkits.

    While porting already-written apps between Android and iPhone is no doubt very difficult, there are tools available to reduce the effort of duplicate coding required. It’s apparently not an intractable problem.

  40. @esr: Spreadsheets? What open source projects is/was Dan Bricklin involved in? Or is his claim on inventing the spreadsheet dubious?

    @Gary Barnett: You Microsoft astroturfers/fanboys always point to “office/productivity” innovations as if Microsoft ever made even one — all of Microsoft Office’s innovations — except the one everyone hates, “Clippy” — came straight out of Xerox PARC, and were invented by then-PARC hackers Charles Simonyi and Butler Lampson before they ever worked for Microsoft. Everything else “new” in Microsoft Office is simple evolution. But I digress. Bayesian spam filtering is one example of a more recent office/productivity innovation that you probably take for granted. It was invented by hacker Paul Graham.

  41. > Corporate “innovations” are the stuff that didn’t require enough bravery or imagination to need the hackers to do it.

    Oh, please:

    Unix: AT&T
    Database: IBM
    Window system: Xerox PARC
    C: AT&T
    Java, NFS, ZFS: Sun Microsystems
    Darpanet: DoD
    iPhone: Apple

    Linux, MySQL, Gnome, gcc, and Android are all knock-offs of corporate innovation. There’s plenty of open-source innovation too, but the really *big* innovations — the ones that require bravery and imagination — are often corporate.

  42. Xerox PARC never produced a single successful commercial product for Xerox. Giving AT&T credit for Unix and C is also a dubious claim at best, once you know that it started as a skunkworks project. Calling the Department of Defense a corporation is also a dubious claim at best.

    ZFS is A) open source and B) not an innovation. There is nothing in ZFS that wasn’t already implemented by someone else.

    Databases, okay, yeah, it was developed by IBM R&D.

    Actually, in your list of knock-offs, you’re missing a few things:

    MySQL was the first RDBMS to be optimized specifically for Web applications
    GCC isn’t just a C compiler — it’s architecture allows for all sorts of languages to be plugged into it: GNU’s Java, Fortran, Pascal, C++, and so forth all are just plugins for GCC.
    GNOME combines some of the best features of Windows, BeOS, Mac OS, CDE, etc. That’s what makes it unique.
    Android isn’t a knock off of the iPhone. And Apple didn’t invent anything new with the iPhone: the basic concept, once again, emerges from Xerox PARC.

  43. @morgan — You’re right that Xerox PARC failed commercially. That does nothing to diminish their innovation.

    It’s true that many of these projects began as skunkworks. All innovative companies support such activity for precisely this reason — big ideas often start small but need sustained funding with mimimal distraction to bear fruit.

    I invented ZFS, so I’m biased, but I’ll point you to a neutral party regarding ZFS innovations:

    http://www.infoworld.com/t/platforms/sun-zfs-breaks-all-rules-512

    Gnome, I agree with you: I like it better than any of the things it borrows from. I just wouldn’t call it innovative because the differences between Gnome and its predecessors are pretty minor. Again, that doesn’t mean it’s not good engineering; it just doesn’t fundamentally change the way you think about window systems.

    To say that Android isn’t a knock-off of the iPhone… well, that’s just silly.

  44. Android isn’t a knock off of the iPhone.

    Android itself, maybe not, but Android phones as we currently see them? Hell to the yes.

  45. It’s true that many of these projects began as skunkworks. All innovative companies support such activity for precisely this reason — big ideas often start small but need sustained funding with mimimal distraction to bear fruit.

    I would say that many of the list that you refer to as being invented by corporations were, in fact, invented by hackers funded by corporations. The two groups are not mutually exclusive.

  46. Corporate “innovations” are the stuff that didn’t require enough bravery or imagination to need the hackers to do it.

    Not strictly true, but true in the sense that corporations who are interested in high yields of innovation will hire hackers and let them do what hackers do.

    The hackers tend to be well remunerated for their efforts. Correspondingly, the company tends to keep the source closed, so they can make that money back and then some.

    Good to know that the only two claims to OSS innovation you can point to date back three decades… you’re really selling that OSS innovation man!

    How about this: the OLPC XO-1. A literal game-changer. Negroponte’s foundation literally pulled a rabbit out of a hat, defying even my doubts about a project with such an ill-conceived and ill-organized business model. The XO-1 is rough around the edges and doesn’t fulfill anything close to all its campaign promises, sure, but it is a semi-rugged real laptop with a miraculous display, long battery life, wireless networking and an easy-to-use UI for $200. It also bootstrapped the netbook industry into existence (though, between powerful smartphones and the iPad the longevity of that market segment is now in doubt). The Sugar UI in particular is a marvel replacing the “desktop metaphor” with something far more manageable and less clutter-prone.

  47. “After 1981, IBM seized the lead in the personal computer market … Nineteen years later Google has at least as cool a brand as Apple, more financial mass, and more engineers.”

    You mean, “Twenty-nine years later”, don’t you? 2010 – 1981 = 29. (And yes, it’s not easy to remember how long ago 1981 really was, if one was alive and even a very young adult then…)

    ESR says: Will correct, thanks.

  48. Jeff, you nailed it here, but you missed one important supporting point. The iPad and its current-and-future ilk may very well kill netbooks, but without netbooks, the iPad would have never happened. The iPad is Apple’s answer to its own admitted inability to compete in the netbook market.

  49. @Jeff Bonwick: I didn’t realize you were the Jeff Bonwick. :) ZFS is pretty cool, but a lot of its coolest “innovations”, like snapshotting and variable block size for example, were already implemented by storage vendors like EMC. What ZFS brings to the table is combining all these cool technologies together in an open and open source filesystem.

    That and, um, what Tom Dickson-Hunt said.

  50. @Daniel Franke: The iPad will fail for precisely the reasons esr outlined above. It’s a walled-garden, and even more of a walled-garden than the Mac and Apple II were: your software must come from or through Apple. I tihnk it was Paul Graham that mentioned it not too long ago on his blog: the entire concept of software publishing was an epic fail in the late 1970s and its an epic fail today.

  51. morgan:

    Yes, it was Paul Graham who said that. But at the same time, Y Combinator’s most recent Request For Startups is a call for innovative applications for the iPad. Even if the iPad dies, and I hope and expect that it will, its carcass will produce fertile ground for similar ideas. Netbooks suck. The iPad sucks too, for different reasons. But somewhere equidistant from the two, there’s a new paradigm waiting to happen.

  52. The market seems to think the iPad is going to be a win. Apple announced a ship date today, and the market for Apple stock jumped nearly 4%.

    Just in today’s trading, Apple’s market capitalization is up over $7.5 billion dollars, at around $198B total; compare to Google’s $180B, HP’s $123B, and IBM’s $165B. Microsoft still stands at $250.75B

    Where is your god now?

  53. Maybe the most honest thing to say about innovation is that both the closed and open source worlds do it.

    Open source did it with the internet, with a little help from DoD.

    But closed source has done a lot, like the Mac and Windows GUI. Gnome and KDE came later. Xerox never capitalized on PARC innovations because they didn’t have the vision to see how a point and click interface would help them sell more copiers.

    I’m not so sure about the Ipad failing; the IPhone is also closed source, a walled garden, and it has done great. Developers can write stuff for the thing, it just has to be approved by Apple.

    What about promotion? People use Windows 7 because MS runs commercials for it.
    Who will run commericals for Linux? How do you expect ordinary people to use Linux if they don’t even know what it is?

    Hackers can hack, but sometimes you need a smart businessman to promote a new thing and commercialize it.

  54. Who will run commericals for Linux? How do you expect ordinary people to use Linux if they don’t even know what it is?

    “Ordinary people” don’t even know what a browser is:
    http://www.youtube.com/watch?v=W3vv0_RNTM8
    …but they still use one.

    Ultimately ordinary people don’t have much choice regarding technology. They’ll use the things created by hackers and, generally speaking, ordinary people will think the availability of multiple vendors/brands/manufacturers is “choice.”

    Another nice Paul Graham quote: “When it comes to computers, what hackers are doing now, everyone will be doing in ten years.” He was writing in 2005 about Mac OSX increasing popularity: http://www.paulgraham.com/mac.html

    So the short answer is: In the future, ordinary people will use Linux more often, but they still won’t know they’re doing it.

  55. > I didn’t realize you were the Jeff Bonwick

    I didn’t realize I had become the Jeff Bonwick. ;-)

    I agree with you and Tom — hacker is what you do, not who pays you. Getting a regular paycheck does not somehow preclude innovation, nor does poverty (or being a trust-fund baby) somehow ensure it.

    As a programmer, I like open source because it means my work will outlive my employer and I will enjoy microscopic fame. As a consumer, I appreciate innovation regardless of the source, because it benefits all of us in the end.

  56. In the war for market-share, allies are better for your long-term prospects than walls

    ESR didn’t mention how Apple’s strategy is unusual. For most intents and purposes the App Store is “open” — a free market. Yes, when you purchase an app for the iPhone, there is an Apple tax, and the bits come thru Apple servers, but the typical users don’t think or care about that. Yes, Apple reserves the right to reject any app, but they normally don’t exercise that right.

    Apple has paired an approximately “free” software market with a strictly closed hardware market, using the openness of the former to energize and lock down the latter. Apple’s obsessive control gives them some power to provide a better-integrated experience vs Android, which has potential compatibility problems between manufacturers and no centralized authority to require a minimum quality standard for apps.

    How long will Apple’s quality control and first mover advantage allow them to hold off a whole world of competitors united behind Android? Probably several years at least, although if Google/Android ever produces a phone or tablet as well-designed as Apple’s, that will shake things up a bit.

  57. @tentech: The supposed “compatibility problems” that plagued the PC industry went away when manufacturers got together and created standards like Plug-N-Pray, Addled Power Mangling, etc. Some led by Intel and Microsoft, some led by consortiums like the Video Electronics Strangling Association, etc. I suspect that some defacto standards already exist, mostly around cheap Taiwanese chipsets, and I suspect that more standards are likely to come. Centralized authority never did totally materialize — today’s PC is patchwork quilt of standards.

    How long will Apple’s quality control and first mover advantage allow them to hold off a whole world of competitors united behind Android?

    How long did Apple’s quality control and first mover advantage allow them to hold off a whole world of competitors united behind MS-DOS and Windows?

  58. BTW — the wildcard in this whole scenario is, once again, Steve Jobs’ infamous Reality Distortion Field, this time with Eric Schmidt playing Bill Gates’ part. That’s one thing that ensured the Macintosh a place in the market, even despite his absence from Apple. Say what you will about the Black Turtlenecked One, but one thing he’s always been is a damned good marketeer.

  59. >>> So the short answer is: In the future, ordinary people will use Linux more often, but they still won’t know they’re doing it.

    Well that’s all well and good. Why aren’t they already? Why is next year the year of Linux on the desktop….. and always will be?

    Why is it when a person goes to Best Buy there is not a single machine there with Linux running on it?

    Those people don’t know the difference between a browser and a search engine…. but they still use them both. Their machine came with a browser when they bought it, and they quickly learned to use Google or Yahoo.

    I wonder, if you did a man on the street type interview, and asked “Do you know what Linux is?” what percentage would have even heard of it.

  60. Posting this on my OLPC laptop tethered through my Android phone. (Open source totally rules.)

    (The coffeeshops in the Cambridge area are starting to clamp down on WiFi freeloaders.)

    Darrencardinal, Craig didn’t necessarily mean the desktop. And it won’t be called Linux, it’ll be called Android, Maemo, WebOS.

    I see sexy women of the kind who used to pick on us geeks in high school carrying around Android phones now.

    The situation for Linux is better than it was, but it’s just not happening on the desktop. And Apple still offers the more compelling user-experience scenario.

  61. It was my understanding that OLPC was a failure.

    And the sexy women I see are all using IPhones. :)

  62. It was my understanding that OLPC was a failure.

    It was not quite the success that everybody expected. Like I said, it didn’t fulfill many of its campaign promises. Yet it’s out there and governments are adopting it.

    And the sexy women I see are all using IPhones. :)

    There are probably more iPhones in the hottie demographic than Androids, but it’s not completely one-sided. :)

  63. I think Mr. Raymond makes a whole set of fundamental assumption errors:

    1. That open-source/closed-source or open-system / walled garden matters. It does not.
    2. That Business not consumers drives phone sales. It does not.
    3. That technology is anything other than a low-cost commodity.

    All these assumption errors lead him to ignore the biggest issues driving the relative success and failure of phone sales, profitability, and market dominance.

    It is that technology is a (relatively) cheap commodity sold on top of “service” to consumers for mostly entertainment. Apple’s profits come from the sales of Ipods/Iphones, approximately 1 billion downloads off Itunes and more than 100 million Ipods. This is because the market for consumer electronics is far larger, and more profitable, than the business market.

    In the 1980’s, this was the opposite. But the money is NOT in business specified appliances, like the excellent Palm phones that could sync with all sorts of calendars, and so on, back in the 1990’s. That’s now just a commodity. Rather, what consumers WANT (and will pay a premium for, in far larger numbers that DWARF business purchases), is entertainment on the go. Games, music, and increasingly movies/video.

    And they want it all just to work, without any hassle. Because they are technologically illiterate themselves. Its an appliance that they push a button on. It just works.

    This means the money is in systems integration for a near 100% reliable, consumer entertainment appliance. One that works with the existing model of service (Itunes, Facebook, maybe YouTube and Hulu). Consumers don’t care about running shell scripts to query a database. Or how “open” Android is which provides a ton of more apps.

    They want a phone that can browse Itunes and download a song, a TV episode, a movie, and play it. They want to browse/listen to free/streaming audio/video ala Lala.com. They want to watch a movie on their phone, or if at home, stream the movie stored on their phone to their TV. They want apps that do this seamlessly, without much hassle on their part.

    There is an opportunity for Android here. But it is not an ecosystem of alliances that provide all sorts of business (or even consumer) apps. Rather, it is doing the basics of consumer entertainment and working with services that exist BETTER than Apple does now.

    Itunes will always be iffy, if for no other reason than its Apple’s service. Currently the Iphone and Ipad don’t run flash. This makes streaming video from Hulu and other sites a consumer issue (they expect to be able to just use it). Whoever races to the finish line FIRST to provide a seamless consumer entertainment device that works with services (and TVs) will “win” in the marketplace.

    BUT … the potential exists for the marketplace to be bifurcated. One section of low-cost (Android/Linux/Symbian open source no/low-cost licence OS) phones that do the bare minimum of basics and are sold at very low cost, in both developing nations and developed markets for budget consumers. The other, high-end latte sippers who demand the best of mobile entertainment, competing with existing consumer entertainment folks. This will generally require tight systems integration (software-hardware limits, testing, etc.)

    The handset people, many of whom are stressed with suddenly reduced unit sales (Nokia, Samsung, Sony/Ericcson) might go the low-cost way because it plays to their strengths: manufacturing lots of electronics at low margins. No need to innovate the brilliant next consumer entertainment feature (better voice recognition, integration with car audio systems and GPS, etc). On the other end you could see Apple in battle with not Microsoft or Google but with Sony’s “next Gen” Playstation Portable, or a new Nintendo device, that also makes phone calls. Each with “gotta have it” games like Sonic or Halo or Call of Duty.

  64. After I check out their rack, of course.

    If you hang out with them long enough they do make a phone call. :)

  65. Hmm. Apple has a powerful innovative entry into an important new market.

    But Steve Jobs insists on a closed architecture for the product.

    I seem to recall seeing this happen before. And Apple had to fire Steve Jobs to change the policy and survive.

    Will Apple have to fire Jobs again?

  66. Will Apple have to fire Jobs again?

    No. I think it’s likely that Jobs will retire by the year’s end due to his failing health. Not trying to be rude about it, just stating what I believe to be to true.

  67. >>> And Apple had to fire Steve Jobs to change the policy and survive.

    Did Apple really have to fire him? As I recall that was perhaps the biggest mistake they ever made. They damn near went under without him. They should have fired Scully, not Jobs.

    Surely you’ve noticed since they brought him back, they have been kicking butt all over the place, Ipod, IPhone, resurgent market share for the Mac, etc. The Apple stores were his idea and were a big risk, and have proven to be a huge success.

    God I hope his health holds up.

  68. Whiskey is right. Individual apps and a free ecosystem are nothing; systems integration is everything. Apple does this better than anybody and this makes for a more compelling UX. How do they deal with the scaling issues? By vastly constraining the problem space; to wit, centralizing platform resources. Yesterday it was the onerous cost of developer docs and type/creator code registration; today it is the vendor-approval-required app store. Hackers still can do what they do best, provided they accept Apple’s terms. For most apps, this is a fair trade.

  69. Darrencardinal: Jobs had a vision for the Macintosh that absolutely excluded any sort of hardware add-ons. It was supposed to be an appliance, like a toaster. And Jobs refused to make any changes in that vision. IIRC, this including rejecting any features that would require a cooling fan (such as an internal hard drive). Ultimately, he had to go.

  70. Whiskey & others,

    the whole “entertainment on the move” idea actually depends on something that may seem quite remote, but yet it depends on them 1) on the assumption that urban sprawl will (globally) go on and even increase and 1-hour commute times will be the norm 2) people will generally do that with public transport, because entertainment on the move in the current forms just doesn’t work while driving a car.

    If the world of the near future will be one of masses and masses in every city spending significant time at sitting at a suburban train / light rain and being bored, it will be a hit. If, however, the car makes a comeback or bicycles become more popular or (or what I’d consider the most enjoyable idea in warm climates) people stop being afraid of motorbikes, or somehow commute much less (reversed urban flight?) then the whole “entertainment on the move” idea will fade into disinterest.

  71. Hmm, interesting point, Shenpen, I broadly agree that total time commuting to work is going to go down a lot as telecommuting is going to explode. However, you’re always going to spend time visiting other people and you’re going to want not to be bored or unmoored from your personal digital information while doing so, that’s where mobile computing comes in. Since there is not going to be a significant drop in people leaving their houses to visit each other, there will always be demand for mobile computing devices to make up for the lack of your home desktop when you’re away from home.

  72. Darrencardinal: Jobs had a vision for the Macintosh that absolutely excluded any sort of hardware add-ons. It was supposed to be an appliance, like a toaster. And Jobs refused to make any changes in that vision. IIRC, this including rejecting any features that would require a cooling fan (such as an internal hard drive). Ultimately, he had to go.

    Actually, before he was driven out of Apple, he contemplated the idea of creating a high-powered scientific workstation, an idea which was realized with NeXT. In fact, this idea consumed him so much he abandoned the original Mac, and it was his inability to move unsold inventory that got him driven out, not unmovable determination to an unlimited platform.

    Oh, and need I remind you that the NeXT platform was used to implement the first Web server and browser. So Jobs deserves indirect credit for those, too. :)

    Yes, he really is that insanely great, and he and his Macintosh team deserve credit for much of what computing is today.

  73. not unmovable determination to an unlimited platform.

    Er, strike this and add:

    …not unmovable determination to a limited platform. Jobs learned his lesson. :)

  74. Two observations:

    1. Is the comparison of PC consumers roughly equal to smartphone customers? I’ve always assumed that people would rather work on their PC, but the smartphone is appealing due to its convenience. If my assumption is true, then the number of things one would want to do is limited, and third party support is less interesting… I’ve observed that people have an “apps phase” when they first get an iPhone/iPod Touch, they can’t get enough of them, at first. They scale back after getting sick of the amount of garbage on their device, but I suspect there is a long-term souring effect on extraneous apps.

    2. Nothing makes one pine for developing in Windows than developing on Windows Mobile. Which I’ve always contended is the biggest barrier to third-party support on Mobile.

    ESR: Could you make a related post concerning the sentence in TAOUP: “Never bet against the cheap plastic solution.”

    1. >According to the EFF, the iPhone App store is very, very far from the open market some have described above. See Digital Rights Group Blasts iPhone’s Faustian Developer Deal.

      Um, to whom does this come as a surprise? Takes a starry-eyed Apple fanboy not to notice how controlling they are. And this is precisely the sort of thing I meant when I said “Ubiquity or control. Choose exactly one.”

  75. > Um, to whom does this come as a surprise?

    To me, a little. I’m no fanboy, but I thought the large number of apps was pretty good evidence that Apple was letting almost everyone play.

    Yours,
    Tom

  76. >To me, a little. I’m no fanboy, but I thought the large number of apps was pretty good evidence that Apple was letting almost everyone play.

    The thing is, really big players can’t. Firefox, for instance. And in general, anything that competes with Apple’s monopoly on control is frozen out. No interpreters; no file managers; nothing that duplicates Apple’s functionality. One of the things I use my iPod Touch for these days is quick file-grabbing in situations where getting out an actual computer is impractical; to use it in that way, I need it to be jailbroken, because Apple won’t let me have proper file transfer from iPod to computer, or a file manager to delete the files once I’ve got them on the computer, or a browser that will download files, or SSH/SSHD. All these things come with the jailbreaking hack, called so for a reason.

    1. >All these things come with the jailbreaking hack, called so for a reason.

      Notice the second-order consequences. Apple is going to collect the worst of both worlds – it will lose control of the app ecology and look overcontrolling and petty. Two lossages for the price of one!

      Next stage in the game is when somebody releases a must-have iPhone app that fights Apple’s business plan. Apple bans it. Then jailbreaking gets really popular. Hilarity ensues.

  77. In light of this, isn’t it rather fitting that HTC elected to name one of their Android smartphones the Droid Eris? :)

  78. An interesting article on “Old World” vs. “New World” computing. Roughly Old World computers are general purpose, open, and tinkerer-friendly; New World computers are purpose-specific, closed, and Aunt-Tillie-friendly. I think Eric is, like me, fundamentally afraid of what the New World portends. But Steve Jobs realized back in 1980 or so that the New World was the future. (That’s why they call him a visionary, folks.)

    I think deep down we want the best of both worlds (heh). Android almost gets there… almost. When I use my Android phone sometimes it feels less like a phone than it does my Linux laptop… things get sluggish and unresponsive when many apps are spawned in the background. (Steve nixed multitasking on the iPhone for a reason, though it may appear in an upcoming OS update… yet another usability compromise.)

    1. >An interesting article on “Old World” vs. “New World” computing.

      From somewhere, a hollow voice says: “Thiiiiiin clients! PDAAAAAAAAs! They never leeeaaaaarn!”

  79. Do you actually have examples of these “New World” computers other than the iPhone, which has negligible market share, and the iPad, which hasn’t even launched yet? It strikes me that these are merely examples of appliance electronics, the mp3 player and cell phone, the latter spawned from the peculiarly closed world of cellular carriers. Now that those appliances are becoming capable of actual computing, they’re being opened up: Symbian going open source and the popularity of Android means that more than half of all smartphones are now nominally open source. The actual trend is the opposite of what you claim, Jeff, despite a relatively small blip in iPhoneOS phones being sold. Hilarious that you think Steve was right in 1980 despite the ensuing 30 years repeatedly proving him wrong. :D

    I just read those links that esr added, I’ve always said the real problem with Android is the horrible Java/XML tech they chose. It might be better than the iPhone but it’s still bad. It’s still early times in mobile; given all the mistakes being made, it’s ripe for someone else to come in take away share, perhaps Windows Mobile 7.

  80. The actual trend is the opposite of what you claim, Jeff, despite a relatively small blip in iPhoneOS phones being sold.

    And in 1998 you could have claimed that the actual trend is towards more PCs with floppies and serial and parallel ports. Without the iMac to make the user-experience case for a PC with only high-capacity removable storage and USB/FireWire ports, you’d be right to this day.

    Apple takes point for the entire industry. Their failures are massive boondoggles, but their successes are copied by everyone.

  81. Jeff, What? Saying open source mobile OS’s now have 50+% market share is like claiming that some ephemeral hardware accessories would stick around forever? You need to get a grip on your analogies. Btw, every desktop motherboard I’ve seen since 1998 still has serial and parallel ports, though I have never used either of those ports, so they’ve hardly disappeared. ;) The iMac didn’t make the case for squat, those are standards that several companies implemented long before Apple did. All Apple does is bring these already developed innovations to their poseur fans, who will overpay for any slop they serve up.

    As for Windows going the app store route, I’ve long said that Microsoft’s M.O. of simply copying everybody else doesn’t work quite as well for them now that they dominate the market, as they now copy everybody else’s stupid ideas too. I see no other reason for them to copy Java with .Net and use so much XML in their recent tech, groupthink that unfortunately infects Apple and Google too. Good luck to Apple trying to maintain their meager 5% smartphone market share, with an increasingly deranged Jobs believing he owns multitouch, perhaps because of cheerleading from ignorant acolytes such as yourself. It appears he’s finally turned the Reality Distortion Field on himself. :D

  82. You’re right, of course. But I have a theory on when consumers will say ‘meh.’ . It’s not always about the best tech (remember VHS vs. Beta), but it’s about the best tech that’s good enough. Good enough means that it does the job that the consumer wants it to do.

    The two biggest uses of VCRs were purchased movies and time-shifting TV: people wanted to be able to watch what they want, when they wanted. Since then, time-shifting moved to PVRs, movies moved to DVDs and later BluRay Discs. Now those are converging to media center PCs and media center appliances. Why did things happen this way? Time-shifting went to the PVR because it was tons easier. Movies went to DVDs because of the better quality necessary to take advantage of home-theatre hardware. In the end, the two are converging because really DVD movies and home movies in general were always about one thing: the ability to watch what, where and when one wants to. That’s why the proliferation of video-capable portable media players.

    By the same token, laptops were about computing where and when one wants to. Laptops got ever more smaller until we got PDAs. But PDAs never hit big until they started merging the capabilities of the PDA and the mobile phone. Computers, for most people, are about communications, information search and retrieval, and data creation. Smartphones bring that to the table.

    The thing is that smartphones aren’t good for data creation. They’re good for data consuming. So, in the end, people will meet that need in the future in these four ways:

    – Desktop PCs at home
    – Laptop PCs on the go
    – Better smartphones, perhaps with esr’s better displays and input devices
    – Something like the iPad

    If esr is right, Desktops and Laptops might well go the way of the dodo if they do not retain any significant advantage over his model. That remains to be seen, but he may right.

    I don’t see a future for the iPad because it doesn’t enable better data creation. Smartphones are already good enough for most data consumption needs, with laptops and desktops (and servers and clusters) filling the gap in the data creation department.

    But the iPad being a walled-garden is exactly one of the reasons why it doesn’t enable better data creation. There are others as well, but the walled-garden is a big one.

  83. I don’t know.

    Pundits are notoriously bad at predicting large scale consumer moves. Anyone remember Rob Malda’s famous iPod dismissal?

    Android has problems of being dragged along for any ride it has to take. Coworkers of mine picked up some HTC Hero phones and there’s a bizarre bug where MMS pictures will send but not be received. People are naturally going to blame this on Android. I blame it on HTC, but what makes this embarrassing is that this isn’t HTC’s first Android phone, not by a long shot, and certainly not HTC’s first smart phone. But this speaks volumes about the Android ecosystem.

    What needs to happen for the iPad to be unseated is that someone needs to come out with a mobile OS where the OS licensing requires strict QA and enforces user experience. while this is a largely marketing decision, it is the market that will decide which product lives and which product dies. It’s also why the iPhone is doing so well. 3 phones by apple have 50% of the mobile web traffic. 3. Despite how many dozens or hundreds of phones with Android, Windows Mobile, Maemo/Symbian, Apple’s 3 phones make up a huge marketshare for mobile devices.

  84. taiki, you may be right from a pure marketing/branding perspective, but that’s sort of like when people blame Windows when Photoshop crashes. Android is an open platform, the last thing it needs is the kind of legal straitjackets Apple has. Apple phones may get a lot of web traffic but that’s actually a negative as far as the carriers are concerned. With flat rate pricing, they’d prefer if you used less bandwidth, which is yet another argument for metered pricing. As for iPhoneOS market share, I noted above that it’s less than 10%, not quite as big as its mindshare that leads you to erroneously guess higher.

  85. Ajay,

    Negative for telcos? Big deal. Telcos will either adapt or die.

    However, considering the blaming Windows when Photoshop crashes scenario. Why did Photoshop crash? Was it because it’s running on an OS with a horrible memory subsystem and horribly outdated disk filesystem? Was it an internal error? What’s the crash rate versus running on WINE or on OSX natively? I’m not talking about strictly legal straitjacket here, just a thorough licensing scheme that keeps the platform both open and stable. The platform would be open for all vendors to participate, and yet, remove the huge problems Android’s facing right now. Android’s not a bad OS, but, it’s being handled poorly by OEMs who are used to producing crappy phones year after year; it’s why the iPhone has the share for web traffic, mind share, mobile apps sales and even why it’s taken a bite out of the portable gaming market. Google doesn’t seem to be particularly interested in keeping all Android devices upgradable to the latest and greatest versions of the OS, for instance. The big question shouldn’t be, “Why does the HTC G1 not have 2.1?” but “Why isn’t that upgrade path required to be available for all devices?”

  86. Jeff, What? Saying open source mobile OS’s now have 50+% market share is like claiming that some ephemeral hardware accessories would stick around forever?

    When the leading open source mobile OS (Symbian) sucks and is only so prevalent because of inertia… yes, yes it is.

  87. taiki, when it comes to cellular bandwidth, it won’t be the telcos dying, I’ll tell you that, as much as I’d like them to go away. My point regarding Photoshop was that it’s not Android’s fault if some phone vendor sets it up badly, though consumers may perceive it that way, to the extent they’re even aware of Android. As for your licensing solution, I’m skeptical of doing such QA enforcement at the level of OS licensing. Better to have a decentralized certification scheme, where Open Handset Alliance members or some third parties could come up with various Android certifications that the device vendors would compete to garner for their handsets. I don’t think you can completely wish away some of the UI issues people have, as a lot of it is because we’re genuinely experimenting with what are the best interfaces for these new mobile form factors, but we can certainly do a better job at that.

    I don’t think the iPhone has the share it has because it’s not a crappy phone. I’ve never really owned a smartphone but some friends with iPhones tell me it’s not a great experience. If you think all the phones are bad, that’s a great opportunity for you or someone to step up and produce a better phone at similar cost. I don’t think it’s quite as easy as you imply. As for upgrade paths, a lot of it is just that all this is new territory for mobile (how many mobile devices offered guaranteed OS upgrades before?) so everyone’s still getting the hang of it, plus the traditional closed mentality of carriers persists to some degree. That’s changing with time, but replacing the carriers with an even more controlling Apple is certainly not the solution. Because of Android’s open source nature, there’s basically nothing you can do at the OS licensing level anyway, so the only solution is certification.

    Jeff, just like the floppy or parallel port in 1998, Symbian dominates today. Nobody said that would last forever, but the zip drive never got anywhere close to that share and neither will Apple.

  88. Ajay,

    Apple can do a lot of things other companies can’t because they’re Apple. Few companies have the magical combination of taste, engineering prowess, and relentless focus on the non-tech-savvy end user that Apple does. After the iPad becomes commonplace, keyboards and mice will seem less like essential peripherals and more like extraneous pieces of cruft we only tolerate on our PCs; speaking of which, why do we have to put up with viruses, malware, and install CDs on our PCs anyway? Why not just have a single centralized place to automatically download and install all the software we need, like the iPhone and iPad have? (To be fair, Debian got there before Apple did…)

    The problem is, you’re still thinking like you: a techie. You need to get into the mindset of those who can’t, or don’t want to, bear the cognitive load a PC imposes. That’s the mindset Apple markets to, and it’s the mindset of millions of people worldwide, vastly more than us. If computer companies want to stay in business, they will follow Apple’s lead. After all, that’s what they’ve been doing since the mid-seventies.

  89. As for smartphones, they too will become more iPhone-like as the remaining companies struggle to compete for attention with Apple’s massive mindshare-fu. Android has already gone that way…

  90. Ajay,

    My point about cellular bandwidth was that it’s a huge red herring and that telcos are either going to put up with it or watch their competition put out unlimited plans while they flounder charging per kb. Given that Sprint’s moving with it’s 4G service to a true unlimited plan… It’s not hard to imagine that’s where the future is in wireless broadband in America. Granted, it’s a US Centric view, but, oh well.

    What I’m implying is that Google really screwed the pooch letting any idiot OEM throw Android onto their handset, net top, tablet, netbook, refrigerator, or whatever miscellaneous device they’ve thrown it on to. The reason why the iPhone was such a game changer and why everyone’s chasing the iPhone form factor is that Apple didn’t have to play ball with crappy OEMs who can’t design usable hardware if it were to save their lives. Nexus One is a good phone, but it’s not a great phone. HTC isn’t that good at developing consumer friendly hardware. If Google imposed some level of quality control as part of the licensing of the Android OS for OEMs(but left the consumer side say, Apache or some other OSS licensing agreement that allowed them the freedom to restrict OEMs), then we wouldn’t have the problems where picking up any given Android phone is going to guarantee you the same baseline user experience as any other Android phone. The fact that Google lets HTC get away with shipping out the Sense UI is kind of a bad thing from a consistency point of view. Sense UI is better than the standard Android UI, but it’s still inconsistent.

    The Zip Drive never had the same dominance as the 3.5″ floppy did, but Iomega did themselves a huge favor selling buttloads of Zip Drives. No one else did what Iomega did, which was repackage a large capacity removable storage device into something as small and as portable as Zip Disks. By that same measure, Symbian may be dominant but no one’s getting rich making Symbian apps. Granted not many people are getting rich with iPhone/iPod Touch/iPad apps either, but let’s face it, I can make lunch money if I decided to put together some cheap, fun apps for the iPhone. There’s no way in hell with Symbian’s ecosystem could I do that, or WinMo, or WebOS, or Android. There’s no guarantees of UI consistency, and the app market places suck. Apple’s App Store might be restrictive but it works.

  91. Jeff, Haha, taste? You think that matters at all, other than with Apple’s niche
    group of poseur fans? Apple has a small niche that has dovetailed nicely with
    the move of entertainment online. However, Apple has no idea how to hang onto that content market, just like they had no idea how to hang onto the PC market 25 years ago. The Zip drive is a perfect comparison to Apple because it was a temporarily somewhat popular niche item that was superceded by superior decisions and tech. Touch computing of the form Apple is currently peddling will never be the most common input method. I expect voice recognition to be the end game, touch isn’t good enough to pass keyboards or to compete with voice. We put up with viruses and the like on PCs because most PCs were designed before the internet, when that wasn’t much of a problem. Microsoft is trying to redesign Windows for that and have had their problems doing so. A central store completely punts on the problem by shutting the desktop down, no chance that wins out.

    I’m not thinking like a techie because I have not offered any techie arguments: you’re thinking like a techie who’s trying to imagine what a non-techie thinks like and you’re failing. A non-techie doesn’t care what the tech is, they just want the most apps at the lowest price and hassle. Open platforms have always beaten closed platforms at that game, which is what the Android guys realize and Jobs doesn’t. How has Android gone the “iPhone way,” because of some silly UI choices? Please.

    taiki, cellular bandwidth is anything but a red herring. Metered pricing is the future of all online services, whether it’s fiber to the home or 4G or micropayments for services and software. Some people may prefer to have fixed quotas, like being allowed to download 50 GBs for $20/month or 100 GBs for $35/month, but many won’t. Yes, there are carriers who want to run scams of overpromising and then having the service issues I linked to, but the results speak for themselves. Anytime you try to provide a fantasy of a commons, whether it’s through “unlimited” bandwidth or medical “insurance”/prepayment, overconsumption and problems ensue. That’s why nobody will ever offer a “true unlimited” plan, all plans have quotas or metered thresholds baked into the contractual fine print someplace.

    As for Android, Google is following the Microsoft model of Windows everywhere, we all know how that worked out before. It’s Apple who is now scared of Android, with their patent lawsuit basically aimed at Android through HTC. The iPhone is not a great phone either, if you actually go around asking its users. Google cannot impose licensing restrictions because of the Apache license and I don’t see them changing that, as you seem to want. Certification imposes the same incentives while leaving it voluntary, a much better option. I’ll take variety and the resulting experimentation and innovation over consistency anyday. I chose the Zip drive analogy because it perfectly mirrors Apple’s success, a flashy niche product that never became mainstream. Considering the much bigger Symbian market, I bet those devs are making a lot more money off their apps than iPhone devs. The App store is a flash in the pan and has no legs, partly cuz Jobs is cutting its legs out from under it with his dumb patent lawsuit, which is naturally responded to in kind. Blowback’s a bitch. :)

  92. You think that matters at all, other than with Apple’s niche group of poseur fans?

    It matters, but Apple’s intended target audience doesn’t realize why it matters. They just know that their Mac is prettier, easier to use, and less prone to viruses or breaking as soon as it gets out of warranty than their friends’ PCs. And hey, genie effect!

    The thing about good taste is that if executed well, it becomes practically invisible. For example, go watch a recent Pixar movie. There’s probably something in every second of that movie that at least led to an ACM SIGGRAPH paper, or possibly even got someone their Ph.D. for some innovative new rendering technique. Yet all you see is a tender movie about a rat who wants to be a chef. The only people who are conscious of the artistry that goes into a tasteful execution are the aforementioned trendy artfag types.

    Counterpoint (in the interest of equal time): trendy tech-hipster author Cory Doctorow hews close to Eric’s argument. Like Eric and Cory I Want To Believe in the triumph of hackers over crypto-fascist artfag types, but we are losing the battle for hearts and minds and it is there, not in the realm of technical superiority, that the war will be won or lost. I’m just waiting for the “mandatory security update” that removes the ability to boot an unsigned kernel image on your Mac. Worse yet, on your HP Pavilion.

  93. Ajay,

    I work for a 4G wireless provider(no, it’s not Sprint either, but we may have some sort of relationship with them ;) ), trust me, metered plans aren’t coming back.

    I am an iPhone 3GS user. I had a choice between several Windows Mobile devices, several Android Phones, WebOS phones, the huge list of Blackberry devices and a Symbian phone. I opted for the iPhone. I’ll take a walled garden against an open, barbaric wilderness any day of the week(at least for my mobile devices; the second OSX starts heading towards iPhone like walled in status is the day I replace OSX with FreeBSD or Linux). Those “silly UI choices?” Those break mobile devices. Running multiple apps? Kills battery, makes apps unstable. Particularly when apps are left open and not closed from going from app to app to app. Multitouch? Means no fiddling around with tiny keyboards meant for Gnomes, Leprechauns and Dennis Kucinich. Multitouch isn’t going to replace a keyboard on the desktop anytime soon, but it does mean that the stylus is mostly dead for mobile devices(Drawing apps not withstanding), of course, but why do I need a stylus for browsing the internet on my phone? No plugins/flash? Means Safari isn’t crashing every five minutes because some idiot web designer decided that Flash was a perfectly reasonable alternative to a DHTML menu bar. These things make the iPhone better than most other phones I’ve ever used(although no Hulu on my phone is kind of sad, but oh well; I’ll deal).

    As far as the App store being a flash in the pan, let me remind you that Nokia did try to break into the mobile gaming market some six years ago with a device called the N-Gage. It was a lousy gaming device, and it was a lousy phone, on top of that, it was running Symbian. It lasted all of 2 hardware revisions and it was the laughing stock of the portable gaming market failing to even dent Nintendo’s marketshare. Meanwhile, the iPhone/iPod Touch represents more than 19 percent or so of all games sold in the mobile segment. Sony’s PSP? 11%. Granted, it’s also due to the fact that the device does more than just play games and thus, there are a lot more out there. Besides that, no other mobile device has shown up with those kinds of sales figures. Not Android, not Windows Mobile, and definitely not Symbian.

    As far as the Elan lawsuit goes, well, if Apple did violate patents, it’s not like apple’s hurting for money to properly license those patents. ;) Unlike some other mobile device makers, namely Palm, they’re not going broke.

    But this is getting too much into defending the iPhone and not actually discussing why the hacker mindset is wrong when it comes to the wider electronics market place. On the desktop, yes, walled off devices rarely do well, and those that do are gaming devices. Open devices rarely beat walled garden devices in the mobile market however. For instance the GP32, the GP Wiz, and the Pandora are never going to see the same sales figures as the Nintendo DSi or the Sony PSP. Even if they are open. Symbian may take the majority of the smart phone share world wide but how many of those phones are 3 or 4 year old devices? How many of those phones are the cheap subsidized phones that telcos give away as part of the signup package? It really has nothing to do with how open or closed they are, it often comes into question how much thought and usability has been put into these things. It’s not just the software that matters, it’s also the hardware. The iPhone’s success is also based on it’s formfactor. it has all of 4 buttons, and two of those are volume control. Not just that but the hardware itself is pretty stable. It has half the failure rate that Blackberry devices see. By comparison, Android devices at minimum must have something like, 6 buttons and a scrolling device, either a DPad or a trackball of some sort. It’s not user friendly when someone has to figure out which button does what. Which is something I’m often doing on smart phones. Extend this to the tablet market and I think we’re going to see more success than what’s been seen before in the tablet field.

  94. Jeff, the Mac is only less prone to viruses because it’s such a niche platform that crackers don’t bother going after it, it’s actually less secure otherwise. I admit the genie effect was nice, it had an effect on me when I first saw it 6-7 years ago. :) No doubt taste helps sell Macs and Pixar movies, but it’s a marginal effect that’s swamped by one main priority: how much can you get done. btw, Ratatouille and Up were great. :D I think you’re conflating my business/consumer arguments with the hacker argument about open hardware/software. I have never raised that concern because it’s irrelevant, hackers will hack no matter what obstacles you put in their way. If anything, the more roadblocks you put up, the more you goad the hackers into pushing them aside, and one of them will. My argument has always been the business one, that if you decide to filter who can publish on your platform, you will inevitably make mistakes, intentional or not, and leave good software out. If you make it a decentralized certification process, one certifier might make a mistake but it’s unlikely that all will, particularly since the user can just install without certification if wanted. I don’t think any battle is being lost, because these niche closed platforms might wax and wane but they have never won anything.

    taiki, I understand economics, trust me, metered plans are coming back, plus plans with explicit quotas. Sprint called their plan “unlimited” then put in a 5 GB limit, every carrier does some version of the same bait and switch. The broadband ISPs are the tip of the spear here, as they put in monthly quotas to try and limit bandwidth hogs. Every other utility, whether it’s electricity, water, gas, or phone/cellular voice minutes, uses mostly metered billing but sometimes various quota levels to conserve usage, bandwidth is no different. I don’t want to use a cellular network that is bogged down because some teenagers are video conferencing each other on their tablets: the best way to avoid that is through metered plans. I’m glad you like the closed environment of your iPhone, obviously it works for you. My point is that you’re in the minority and so far that’s borne out by the fact that iPhoneOS has less than 10% smartphone marketshare. Let’s see if that ever changes. I don’t deny that the iPhone kicking something like flash out has it’s benefits, I just don’t think those benefits outweigh the costs.

    As for gaming, eh, all it means is that mobile gaming devices are finally converging with smartphones somewhat. I’m skeptical that general-purpose mobile devices will ever replace specialized gaming handhelds though, because they don’t have d-pads and the like, the PSP and DS currently outnumber iPhoneOS devices by almost 4 to 1. I have to repeat that I’m not making the hacker argument for open hardware/software, merely the business/legal argument for open platforms that anyone can code for. The reason the Pandora doesn’t see the same sales as the DS or PSP has nothing to do with being open, it’s because they don’t have a giant corporation behind them pushing it. However, Android has that in Google, so the superior open approach will not be swamped by corporate might in the smartphone market. Of course good hardware and software design matter: my point is that design is more likely to come out of open competition from Android vendors than from Apple’s closed process. That open vs closed experiment has been run many times and the Apple almost always loses. I think the iPad will do well- I’ve been big on tablets for years- but it will be swamped by all the other tablets out there, just as it is in the smartphone market today.

  95. Jeff, I really think that there’s a huge difference between hacking Javascript and CSS and Apple II hacking.

    That being said though, if I had kids who were interested in hacking something, note that devkits exist for free on all platforms, and Apple’s prepackaged Xcode into OSX install media, with documentation about how the whole system works.

  96. Ajay,

    Sprint doesn’t have any actual cap on their 4G WiMAX service(I work for the company that provides their network, Clear; I guess I should be a little bit more upfront about it). Neither do we. That article you linked was from early 2008, before Sprint and Clear developed their partnership for WiMAX service. Bandwidth availability for mobile devices isn’t limited at the back haul, it’s limited at the last mile, from the phone to the tower. As tech goes forward to WiMAX and LTE, this ceases being a problem with providers being able to offer devices that connect upwards of 10 to 20 mbps with no huge strains on the network itself. EVDO and HSDPA both kind of suck in this regard.

    As far as the number of units sold, you do realize that 50 million is for iPod Touch alone right? Not all iPhone OS devices? That number is probably closer to 90 to 100 million units sold. Putitng it nearly 2-1 for the PSP and striking distance from the DS.

    Pardon the take on a George Carlin routine, but imagine the average OEM, and imagine that fifty percent of OEMs are worse than that. My argument against the Android model that Google is taking is also both a technical and business one. Having market share isn’t as important as having a stable and sane ecosystem. It’s what’s wrong with Windows right now. This means you’re going to have a lot of really horrible and useless devices out there. As the market place stands right now for Android, some apps are already incompatible with some phones because of various reasons. Either hardware isn’t sufficient, hardware doesn’t support 2.1 OS, etc. while right now it is true that there is software that is only compatible with the 3GS, particularly apps that rely on the magentometer and the better 3D chip, it’s much easier to outline what’s compatible with what. iPhone owners don’t have the problem with wondering whether or not their device will be supported in the future. Granted, iPhone OS 4.0 could cut compatibility with the original iPhone and first generation iPod Touch, but, I don’t really see that as likely.

    Also, Being number one for OSes sold doesn’t mean you have a superior product, it means you’ve got an OS that just about anyone can use. Sure there are going to be a lot of Android devices out there, but at what point are Android devices going to become the crappy generic phone that are given when someone gets a mobile handset? How many people are going to be enthusiastic about their device of choice? How many people are going to actively hit the market place? What if the OEM is ordered by the vendor not to include the default Android marketplace but instead force them to use their own?

    It’s really this reason why in pure market share, Apple may barely have double digit percentage of ownership, world wide, but in the long run, will continue to BE profitable and capture mind share and why the iPad’s going to make a crap load of money versus tablets that will barely break even.

  97. oh Ajay?

    2010 Pwn2Own showed that Charlie was wrong. ASLR and DEP, the big features of Windows that Charlie touted as being the key to making Windows more secure than OSX, didn’t actually help.

  98. 300,000 iPads and one million apps sold the first weekend… yep, it sure is fun watching that lock-in fail, eh?

    1. >300,000 iPads and one million apps sold the first weekend… yep, it sure is fun watching that lock-in fail, eh?

      In related news, it’s just the morning after and high-profile buyer’s remorse has already set in.

      I think the iPad is going to turn out to be another Newton, for reasons that don’t have a lot to do with the lock-in issue. Fancy, attractive technology but few use cases – early adopters will be all over it but it will fall just short along several dimensions of being the right thing.

      One reason I expect this is that the tablet form factor with a color screen has been tried before without catching fire.

  99. taiki, Sprint may not have caps on their not widely available 4G service (in 27 mostly small to mid-sized cities) yet, but they do still have a cap on their 3G service. They may talk up a scam of not having such limits on 4G now, but they will go metered or put caps in, just like they did with 3G. It’s not the wireless link alone that’s the problem for cellular data, most towers have meager bandwidth to the tower itself. Of course, long-haul backhaul is overprovisioned and available, but as that link notes, it’ll take time to get that fiber to the cellular towers. btw, Brough has some excellent posts at that blog, including a recent one about how wifi is where it’s at in wireless, partly because of open competition among wifi vendors of the sort I’ve been trumpeting with Android.

    Let me explain something to you: there will never cease to be a problem of how to share capacity, regardless of how much capacity there is. “Unlimited” plans are a scam, because you cannot really sell that and expect there not to be strains. As I said, I don’t want to use a cellular network overburdened by bandwidth hogs and neither do most consumers, though they may only realize that once they’ve bought in and the inevitable strains of such unlimited plans manifest themselves. Umm, Apple wishes they sold 50 million iPod Touches, more like 33 million, according to this estimate, which also has the latest number of iPhoneOS devices at 75 million. Apparently the number I quoted is from last summer, so the real sales number is between my older link and your overestimate. However many it is, a small fraction of those users will ever use it for the type of intensive gaming you see on a PSP or DS, so they have nothing to worry about from Apple.

    Whatever you may find wrong with Windows, they took the right approach of selling as widely as possible, which Android is following. You can bluster all you want about enjoy the straitjacket of Apple’s closed model, but Apple still hasn’t hit the double-digit market share you claim and my prediction is that they will never get very far. If an Android OEM is ordered by the vendor to use their own app store, they obviously think they’re good enough to compete, I have no problem with such competition. Oh, Apple is very profitable but that’s not the path to ubiquity, nor is such a closed model. Apple can have a comfortable niche making the Jaguar Cars of computing, while everybody else fights over the larger market that they can’t hack. 2010 Pwn2Own showed nothing of the sort, as Charlie’s comments were about 2009, when ASLR and DEP did help Windows last longer, before they found holes in time for this year’s contest. However, just like last year, Safari was the first to fall so whatever problems Windows has, the Mac is likely worse.

  100. So Atlanta, Chicago, Las Vegas, Baltimore, Philadelphia and Seattle are small to midsized cities(Soon, New York, Los Angeles and San Francisco)?

    This guy is really wrong when it comes to frequencies. Take a tweeter, and a subwoofer and place them against your nearest facing outside wall. turn on something that has heavy bass and also somehow hits very high tones and crank it. Step out side and listen to which frequency is more audible through the wall. :) he’s also wrong about Wifi. Wifi isn’t designed with this kind of use in mind. Sure, anyone can compete but also this means everyone is going to flood that spectrum with well, crap. It means my local wifi install may suffer. It means that devices that aren’t wifi are going to have problems. Unlicensed spectrum sucks from an SLA point of view. Which is key for an enterprise such as a wireless ISP. Beyond that when multiple ISPs are competing in the same spectrum, speeds *are* going to suffer. Range *is* going to suffer. The laws of RF and physics do not generally care about the market place. I hope to god that ISPs do not start selling through unlicensed spectrum bands. If they do, I’m going to start suing. I don’t want my cordless landline phone to suffer because someone decided to make a buck off of the commons.

    As far as Apple’s double digit share, you just said that Apple makes up 10% of the mobile smart phone market. Those are barely double digits. As far as the 50 million number comes from, that comes from the source you linked to, which showed that 50 million iPod Touch devices were shipped total. Another source, here shows that 41 million iPhones have been shipped. Which equaled somewhere between 90 and 100 million iphone devices sold. Even if it were the 75 million, you realize between the 4 models of PSP released(PSP 1000, 2000, 3000 and PSPGo), PSP Software accounts for about 11% of games software sold in 2009, iPhone OS accounts for *19%*. (Source: Here) Symbian did have a chance to make it’s mark in this territory, but couldn’t because the hardware was terrible and the OS was unusable. The N-Gage was a laughing stock for a really good reason and no matter how open the software platform is, if the hardware’s garbage and the OS UI is garbage, no one’s going to buy it, and if it’s bad enough they’ll also laugh at it too.

    (FYI, I love the psp as a gaming device. even more so than the iPhone.)

    What I find wrong with windows is that the entire ecosystem is entirely unstable. The next OS revision could change everything yet again. From XP to Vista to 7, everything keeps breaking because of poor machine design and even worse device drivers. This is what’s happening to Android. It’s going to have to support just huge swaths of horrible hardware and be tied to lousy OEMs who aren’t interested in putting together good devices. I’m not a hardware engineer, but I’m not inflicting my lack of a desire to put together good hardware on anyone. While there are good Android phones, that’s independent of Android as an OS. Droid’s a decent phone, and so was the HTC Dream/G1. But the Hero’s kind of a lousy phone and the HTC Dream falls into the crap-phone category because it’s not compatible with 2.1(I’m personally looking forward to the HTC Evo, but, I don’t have my hopes up too high). I haven’t gotten my hands on any other Android phone to play around with, but, as time goes on, there may be a time when Android may not be acceptable for licensing even if it is free because changes in the market may move OEMs elsewhere for a phone OS, even if they go back to developing them in house.

    BTW, Apple Safari? The core is open sourced. Javascript engine, html renderer, all of it available as WebKit. The iPhone OS and Mac OS X Kernel? Open source. Darwin. Grand Central Dispatch? Open sourced. It’s not like the entire ecosystem is locked down like in the Windows ecosystem. Sure quartz and quartz extreme are still closed source but, I don’t give a damn about Quartz. I really don’t care that Apple is in a niche market. Apple users tend to be more affluent. So why not target them? Besides, given the fact that no one’s targeting Apple users, shouldn’t it be a race to market to break in? The windows malware market’s completely saturated. There’s money to be made with compromised macs.

  101. taiki, Atlanta and Chicago are the few exceptions that caused me to use the phrase “mostly small to mid-sized cities,” so you contradicted nothing I said. as for the bigger cities that are “coming soon,” I’ll believe it when I see it: they’ve been talking about Boston happening any day now for some time. As for frequencies, your analogy is not so great because sound is a pressure wave and radio is an electromagnetic wave, two different phenomena with different physical reasons for interference, and Brough’s well aware of those effects if you read his comments and other posts. Your claim that he’s wrong about Wifi would have some credibility if you could actually articulate it. As for people setting up too many wifi routers, that’s a contention problem that’s easily solved in software and a nice problem to have at that. :) You’re really reaching when you start bringing up nonsense like unlicensed spectrum can’t have good service agreements, what? You won’t have a leg to stand on if you sue someone for their Wifi outside your house, which is where most commercial Wifi will be competing. I said Apple had less than 10% share and linked to that data further up, but looking at that same link again, I was remembering the 2008 number, as it’s up to almost 15% now. The Wikipedia page I linked to stated numbers for all iPhoneOS devices, which is clear from the table (they got their numbers from this pre-holiday source. I don’t know why you keep repeating that their numbers are only for Touch, then link to a lower iPhone estimate than the one I already linked to in my last comment. As I said, I linked to older pre-Xmas data before, while you are overestimating, apparently without looking at the data even when I link to it. iPhone may have outsold the PSP games by revenue, but your own link shows the same 4:1 ratio for DS/PSP to iPhone game software that I mentioned earlier. I think we’re agreed that Symbian is a joke, :) but I already said that in my first comment.

    I agree that supporting all that hardware like Windows and Android are trying to do is much harder, but that openness is why Windows has shipped and Android will ship many more units, if not with the same margins as Apple. The fact is many consumers are very price-sensitive and they’re willing to pay less for a device that mostly works, without the polish that Apple charges a large premium for and definitely without the App Store nannies keeping the apps they want away from them. Nobody is going back to developing mobile OS’s in house, it’s too big of an investment and there’s too many open source options available to use at minimal cost. Yes, I’m well aware that Safari is based on KHTML, that they grabbed from KDE, and that Darwin is an open source Mach kernel with a BSD userland. Apple has certainly been opportunistic by grabbing open source developed by others and then played nice by contributing back their own changes even when they didn’t have to. However, most of the Apple stack is horribly closed and they’re adding even more controls on top with the App Store Nazis and ridiculous patents. I have never downplayed Apple’s success with rich hipsters that will swallow anything they offer, but then I don’t pay much attention to Jaguar Cars either. ;) Why should script kiddies go after Apple security holes when they can make it up on volume with Windows? ;) I know you’re not an Apple fanboy so please stop drinking the Jobs kool-aid. It will not end well for Apple and their closed model, just like the last time around. :)

  102. Many thanks with regard to the information, I ‘m truly looking forward to the Samsung Epic 4G, although I have constantly been very curious in gizmos, for a long time I have already been researching for my 1st smartphone and I did not get it yet. The actual explanations are quite simple: each and every several of months there is a revolution in the marketplace (difficult to decide!) and prices here in Brazil are quite abusive. Maybe I would enter this world in wonderful style with a Samsung Epic 4G. I hope so.

  103. One reason I expect this is that the tablet form factor with a color screen has been tried before without catching fire.

    Microcomputers were unsuccessful, niche devices. Until Apple made one.

    MP3 players, smartphones. Same thing.

    It’s a year later and it looks like that’s happened for tablets too.

Leave a Reply to Jeff Bonwick Cancel reply

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