Why Android matters

I’ve posted a couple of times about how kewl I think my Android G1 is. But I’m not jazzed about a mere gadget; the really exciting thing about Android is going to be the second- and third-order effects of the software, and how these tie into Google’s strategic interests and the future of open source.

I’m going to start with the relatively far future, like five or even possibly ten years out, because I’m pretty sure my projections for it are very similar to Sergei and Larry’s and that they are what is actually driving Google’s corporate strategy.

Cellphone descendants are going to eat the PC. There will come a day when you carry your primary computer with you in your pocket as a matter of course — but not your primary display/keyboard/mouse. Those will be Bluetooth or son-of-Bluetooth devices (don’t hassle me about video bandwidth, okay? — I know what the issues are) that people leave lying around like we leave notepads and pencils; you will network with them by walking up to one, at which point your cell will do a crypto exchange and attach the device. When you’re done, you walk away. Of course, Internet will be ubiquitous. These devices will never be off-net unless you tell them to be.

Now. You are Google. You make your money by selling ads on the most successful search engine in the world. One of your strategic imperatives is therefore this: you cannot allow anyone to operate a technological or regulatory chokepoint between you and people doing searches, otherwise they’ll stunt your earnings growth and siphon off your revenues. That’s why you ran a politico-financial hack on the Federal auction of radio spectrum to ensure a certain minimum level of openness. And that’s why you are, very quietly, the single most determined and effective advocate of network neutrality.

Now, combine these two visions and you’ll understand why Google is doing Android. Their goal is to create the business conditions that will maximize their ad revenue not just two years out but ten years out. Those business conditions are, basically, an Internet that is as friction-free, cheap, and difficult to lock down as the underlying technology can make it.

Under this strategy, Android wins in multiple ways. In the longer term, it gives Google a strong shot at defining the next generation of dominant computing platforms in such a way that nothing but customer demand will be able to control those platforms.

In the shorter term, it outflanks the Baby Bells. As web traffic shifts to Googlephones (and things like them), telco efforts to double-dip carriage charges by extracting quality-of-service fees from Google and other content providers will become both technologically more difficult and politically impossible. By depriving them of the ability to lock in customers to gated and proprietary services, Android will hammer both the wire-line and wireless telcos into being nothing but low-margin bit-haulage providers, exactly where Google wants them. (A leading indicator will be the collapse of the blatant absurdity that is the ring-tones market, doomed when anyone can hook MP3s of their choosing to phone events.)

As bad as this sounds for the telcos, Microsoft gets outflanked and screwed far worse. As web traffic shifts to Googlephones (and things like them), IE7 and Windows will be about as effective at market control as a buggy whip at a Formula One race — how does it matter that you have most of the PC browser market when a steadily increasing proportion of browsing is done from phones? Microsoft’s efforts to move to a cloud-centric software-as-service strategy will also be severely and probably terminally damaged, because they’ll face a very painful choice. They can stay PC-only and hemhorrage users defecting to cellphone-based services, or they can deploy on Android only to face increasing pressure from business customers who will, quite reasonably, wonder why the service clients and protocols aren’t as open as the rest of the Android platform. (Another leading indicator is what pervasive Gmail will do to Outlook.)

One of the coolest things about this chain of dominoes is that Google itself doesn’t have to win or end up with control of anything for the future to play out as described. It’s not even necessary that Android itself be the eventual dominant cellphone platform. All they have to do is force the competitive conditions so that whatever does end up dominating is as open as Android is. Given that one of the largest handset makers is already being forced to open source their stack for other reasons (Nokia figured out that they can’t afford to hire enough developers to do all their device ports in-house) this outcome seems certain.

For the open-source community, it’s all good. The things Google needs to do with Android for selfish business-strategic reasons are exactly what we want, too. This isn’t an accident, because we’re both pulling in the direction of reducing the effects of market friction, transaction costs, and asymmetries of power and information. If Google didn’t exist, the open-source community would need to invent it.

Oh. Wait. We did invent them. Where do you suppose Sergei and Larry came from? Why do you suppose they’ve been running Summer of Code and hiring a noticeable fraction of the most capable open-source developers on the planet? Well, here’s a flare-lit clue: before those two guys were famous, they sent me fan mail once.

That’s why I think those two know exactly what they’re doing. And that, if it’s true that their business strategy requires them to be open source’s ally, I think I can be allowed a guess that they chose their business strategy so that would be true. “Don’t be evil”; they’re not angels, but they’re trying.

And, from where I sit? All I can say is this: Bwahahaha. The sinister master plan for world domination – it is working!

31 comments

  1. This right here is what I think was the best outcome of “open source” ideology superceding “free software” ideology. By enabling, and indeed encouraging, entrepreneurs to get FILTHY STINKING RICH by aligning their interests with ours (rather than the old “it’s our way or the highway…100% commitment to our every whim and allowing no thought whatsoever for your own self-interest, or else you might as well be Satan” attitude of the FSF types), we get what any rational observer would have expected to get. Filthy stinking rich (and therefore extraordinarily powerful) folks whose interests are (wonder of wonders!) aligned with ours.

    Google isn’t perfect. The good news is that they don’t need to be. As long as they’re strong, and they’re rowing in the same direction as we are, we can call them allies and still challenge them to do better when they fail.

    And if rowing in the direction we want the boat to go happens (coincidentally or not) to make them richer, I say more power to them! :)

  2. >Filthy stinking rich (and therefore extraordinarily powerful) folks whose interests are (wonder of wonders!) aligned with ours.

    That was in fact, my sinister master plan for world domination, and why the community got really lucky by finding a hard-core libertarian to shape its propaganda. I focused first and hardest on changing the demand pattern of the Fortune 500, and addressed the business media more than the technology mags. This worked. Exhibit A was IBM joining up; Android is exhibit B, and in my judgment the most consequential turn of corporate strategy since then.

  3. I agree that the smart phone is likely to become the dominate platform category for certain computing functions, mostly related to communications (that includes most internet usage today). In that role it could also be the main revenue source for Google in the long term.
    Business computing, by which I mean functions that require people to sit in front of a computer display for hours at a time, is not going to follow the same path though. Desktop equipment will always be able to provide a significantly better experience for this type of work. Tools will still be the main revenue generator. Microsoft largely controls the latter market today, and will seek to control it in the future. Whether they succeed or not is debatable, but Android isn’t going to be a major factor (Google probably will be though).

  4. Hi, it’s me from #ubuntu, remember? OK!

    As much as I respect you, I think your vision is too google-optimistic; you’re just overwhelmed with Android.

    I mean, come on! Let’s think about it: Google dominates the world because they invented an open source phone.

    You not only assumed that there will be such advanced phones (I think that might happen, too) but you assumed that it’ll be coming from Google, why? Google is new in the phone market.

    IMO, if there’s any company that’s gonna dominate the world it’ll be one that’s like FB but in real world (you point your device at someone and you can view their public profile, you can interact with people in a virtual world (like lively but much more sophisticated), etc).

    Sorry if I have insulted you or anything, I just think that just like how Google popped out of nowhere, some other company will pop out and compete as a yet-another-computing-giant company.

  5. Maybe I’m hopelessly old-fashioned – that means: in favour of ergonomical user interfaces: big screen, big keyboard, you know, something that fits the biological realities of our bodies – but I just don’t get this mobile revolution at all. I didn’t get it 4 years ago when people started to trade desktops for notebooks. It has all sorts of ergonomical failures: small screen, screen is lower than eye level, small keyboard, touchpad (to a Logitech trackball-junkie, touchpads are definitely stone age) and so on. Notepads were designed to carry, not to use. Does easy carry outweigh the benefits of easy use? Depends on how big a nerd you are, of course, how much you move around etc. But notebooks can be seen as the acceptable compromise, if you define “acceptable” low enough.

    But these mobile crap, with their tiny screens and horrible keyboards… why the hell does anyone want to use them? Anywhere you are, in a café, on a train, why don’t you just use your notebook?

    We should be progressing to exactly the other direction: not small, uncomfortable, and hard to use, but rather putting all our computing stuff on the big screen, TV, so that browsing the web or watching a YouTube video is just as nice, easy and comfortable as watching TV. Not smaller: bigger, more spectacular, more Dolby, more fun.

    On a brighter note, notebooks are OK when put into a docking station, with proper screen, keyboard etc. If I get it right, ESR is talking about the same thing with mobile devices. OK – one day we will get there, but as of now I just can’t get why would anyone want to read PajamasMedia or use GMail on a mobile device. It really, really sucks, because of all those biological realities.

  6. Quoting: http://code.google.com/android/kb/licensingandoss.html#apache2

    Why are you releasing the code under the Apache License instead of GPLv2?
    One of the best explanations for the reasoning behind releasing code under Apache2 can be found in a ArsTechnica article by Ryan Paul.

    Quoting that article

    ermissive licenses like the ASL and BSD license are preferred by many companies because such licenses make it possible to use open-source software code without having to turn proprietary enhancements back over to the open source software community. These licenses encourage commercial adoption of open-source software because they make it possible for companies to profit from investing in enhancements made to existing open-source software solutions. That potential for proprietary investment on top of an open stack is most likely what inspired Google to adopt the Apache Software License for its mobile platform. Availability of Android under the ASL will ensure that a broader number of companies will be able to adopt the platform and build on top of it without having to expose the inner workings of proprietary technologies that give them a competitive advantage.

    Are you still willing to state that you’re unaware that, while the Android stack from Google is ‘open’, its not necessarily what you get with the G1 (or, for that matter, any other Android phone)? Google is *quite clearly* endorsing that the handset makers and carriers in OHA can make proprietary changes to the codebase, prevent the owner from changing these, and not give the source code to those changes back to the community (except for those portions which are covered by GPLv2/3, of course.)

    And even then, I challenge you to find an example of anyone changing the (GPL-licensed) linux kernel on an Android phone.

    Android is a success for Open Source as a marketing term, but the net effect of Androids source availability and licensing remains to be seen. I predict that it will fall somewhat short of your prediction.

  7. Perhaps it’s time to move beyond the enforced back-contribution of the GPL and see if voluntary back-contribution works or not. There is nothing inherently wrong with the GPL, it’s a fair contract, quid pro quo etc. but it would be really cool if back-contribution worked without any contractual enforcement, if it could be driven by economic self-interest and thus could become entirely voluntary. In the longer run, those contracts tend to work that say X gives this thing to Y today, Y pays N amount of money in a week. These are easy to enforce. Very general contracts like GPL or any other kind of licencing/IP contract are very hard to enforce, perhaps they are OK in the short run but it would be nice if in the long run they would prove unnecessary.

  8. A bit more explanation. The core idea about contracts is that you KNOW when they are broken. We contract that I ship you some stuff and you pay me some money: if you don’t get the stuff or I don’t get the money we KNOW the contract was broken, and can prove it at court.

    General contracts like every kind of IP and licencing are a much harder, and I’d say clumsy case. Just like Big Media is fighting an uphill battle against Torrent and will lose it sooner or later, Groklaw is fighting an uphill battle in enforcing the GPL. I could take a code base that took 10,000 hours to write and change it in 200 hours so that you would never prove that a compiled commercial product resulting from it was hijacked from that.

    Such general contracts are perhaps necessary as a temporary measure to protect a fledgling industry, but in the longer run, they are a clumsy method at best and it would be really better if things could be arranged so that back-contribution is driven by economic self-interest rather than a contract, and all the legalese could be dropped.

  9. Such general contracts are perhaps necessary as a temporary measure to protect a fledgling industry, but in the longer run, they are a clumsy method at best and it would be really better if things could be arranged so that back-contribution is driven by economic self-interest rather than a contract, and all the legalese could be dropped.

    Congratulations, you’ve just invented BSD.

  10. >As much as I respect you, I think your vision is too google-optimistic; you’re just overwhelmed with Android.

    I know I pointed out that Google itself doesn’t have to win for the future to play out the way we want. Nokia’s discovery that they have to open-source because of scaling issues demonstrates that there’s a non-Android path forward. On that path, though, there’d be a lot more of the futile thrashing as various players resist the inevitable conclusion.

  11. >Business computing, by which I mean functions that require people to sit in front of a computer display for hours at a time, is not going to follow the same path though. Desktop equipment will always be able to provide a significantly better experience for this type of work.

    If you mean a desktop display and keyboard, I agree with you. A key piece of the scenario, though, is breaking the physical coupling between those and the thing with your files on it.

    Yes, corporate IT departments will want to keep everything physically tied for security and complexity-control reasons. They’ll fail even more completely than they have at preventing laptops from making inroads.

  12. >Are you still willing to state that you’re unaware that, while the Android stack from Google is ‘open’, its not necessarily what you get with the G1 (or, for that matter, any other Android phone)?

    You’re right; it’s theoretically possible that T-Mobile hired a team of programmers and embedded some cyst of closed-source evilness in my phone. Do I think this has happened? No. Do I think this is a likely outcome for a lot of Android phones? Again, no. The reasons are legion, but a few will do as representative:

    1. Cost of goods. T-mobile’s prompt advantage from adopting Android is all the software they don’t have to write, lowering their bill of materials and increasing their profit margins. Every line of proprietary code they stir in forfeits some of that advantage.

    2. Time to market. Even supposing T-mobile has a troop of acquiescent code monkeys who’ll work for bananas, writing code takes time. That is time, and market advantage, you are ceding to your competitors who write less of it.

    3. When there are a dozen Android phones on the market, your cyst of evil will make you stand out – and not in a good way.

    Increasingly, the question is going to be: “Why bother?” What business advantage is there? Time was that the answer would have been video codecs and DRM, because you wanted to get in bed with Big Media. Both those reasons are fading, perhaps gone.

  13. >Such general contracts are perhaps necessary as a temporary measure to protect a fledgling industry, but in the longer run, they are a clumsy method at best and it would be really better if things could be arranged so that back-contribution is driven by economic self-interest rather than a contract, and all the legalese could be dropped.

    I’ve been maintaining for several years that the GPL is a confession of fear and weakness that we no longer need to feel. Most of the community has not caught up with me yet.

  14. 1. Cost of goods. T-mobile’s prompt advantage from adopting Android is all the software they don’t have to write, lowering their bill of materials and increasing their profit margins. Every line of proprietary code they stir in forfeits some of that advantage.

    Russia, China, India, … work cheap. There is a recession on, and the software labor pool is both shrinking and getting less expensive to hire and retain. Or if you’re a carrier, you may be hiring Nossler or WindRiver to do the customization for you, eliminating all the direct expensive of an in-house group.

    Not a big deal.

    And the economics of the cellular market (in the US, anyway) are not about the manufactured BOM of the phone .vs what it is ‘sold’ for at retail. T-mobile has a huge, huge problem: churn. They have the largest churn rate of any carrier in the US. If Android acts to reduce churn, then its a win for them, due in the large to reduced customer acquisition costs.

    Point of reference: there are good-faith estimates that the pre-manufactured BOM of the G1 are circa $145. http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=212001813. This figure doesn’t include software, shipping, handling, the actual manufacturing or accessories, documentation and packaging. You paid T-mobile $199 plus a 2-year contract.

    2. Time to market. Even supposing T-mobile has a troop of acquiescent code monkeys who’ll work for bananas, writing code takes time. That is time, and market advantage, you are ceding to your competitors who write less of it.

    Except
    1) T-MObile got a huge, huge head-start on the first one.
    2) they’re unlikely to have a troop of code monkeys, they’ll hire experts (above) who know how to push the buttons on the configuration GUI such that (for example) tethering doesn’t work.

    3. When there are a dozen Android phones on the market, your cyst of evil will make you stand out – and not in a good way.

    This remains to be seen. Will AT&T launch an Android phone? Unlikely while the relationship with Apple/iPhone is still in good repair. That leaves Sprint/Nextel and Verizon in the US, and of these two, only Sprint/Nextel is a member of OHA.

    To this point, cellular customers have only been able to grumble about the occluded features of some phones. Since there will likely only be two carriers in the US market with Android phones, (in the near future) it would be straight-forward for them to collude.

    The point is (still) that, even though the phone is “open source”, you as a customer, are powerless to correct the situation. Therefore, the “open source” nature of the phone is hype, and here we come full-circle.

  15. If you mean a desktop display and keyboard, I agree with you. A key piece of the scenario, though, is breaking the physical coupling between those and the thing with your files on it.

    You want your files *with you*.

    Yes, corporate IT departments will want to keep everything physically tied for security and complexity-control reasons. They’ll fail even more completely than they have at preventing laptops from making inroads.

    The corporation is dead. Small is beautiful.

  16. You mentioned net neutrality. I’ve read about this, and the opposition to it. As far as I can tell, net neutrality is more supported by liberals/democrats, while the opposition is made up more of conservatives/republicans. But for the life of me I can’t figure out which is the the more libertarian position.

  17. Google combines idealism with making money. They’re one of the few companies who do this, along with Apple and Peet’s Coffee. This seems to be a uniquely Californian trait, especially in the Bay Area. Why would people do that here and nowhere else?

  18. >But for the life of me I can’t figure out which is the the more libertarian position.

    Your confusion is entirely reasonable. I’ll probably blog about this.

    And now, I have.

  19. @JimThompson: UWB? Is that the same thing that Intel just dropped? And some other developer of has one month worth of funding?

    The economics of the GPL are simple: you can’t afford not to share your improvements with the community. Thus the GPL is just a feel-good measure. It appeals to the socialist in people.

  20. @RussNelson: UWB? Is that the same thing that Intel just dropped?

    They disbanded their internal group, yes.

    > And some other developer of has one month worth of funding?

    Yes, at least two. Staccato and Olidata, both funded by Intel Capital (a VC firm internal to Intel.)

    Both have certified single-chip all-CMOS product too.

    > [The GPL] appeals to the socialist in people.

    I always saw it as leverage for a capitalist business plan.

  21. “A key piece of the scenario, though, is breaking the physical coupling between those and the thing with your files on it.”

    A 120 GB Passport Drive can do that quite well. Or a big enough thumbdrive. This is a good direction, what I’m baffled about is why did totally unergonomic peripherials – tiny screens and keyboards – suddenly became a hot, trendy idea…

  22. >This is a good direction, what I’m baffled about is why did totally unergonomic peripherials – tiny screens and keyboards – suddenly became a hot, trendy idea…

    A friend gave me a Palm Pilot years ago; I never found a use for it. I had a Samsung 660 on which I briefly had browser capability, and that sucked pretty badly. The G1 is above some screen-size threshold where the browser actually gets useful. That threshold, for me at least, is somewhere above a Blackberry and below the G1.

  23. >But Eric, you didn’t use your Compaq iPAQ either, and it’s screen iw quite large.

    It wasn’t a phone, and it didn’t have net. I never found a use for PDAs without these functions.

  24. > It wasn’t a phone, and it didn’t have net. I never found a use for PDAs without these functions.

    There have been other touch-screen phones with 3G + WiFi networking.

    There were any number of solutions for dial up networking over Bluetooth using cellular bandwidth.

    There were iPAQs with built-in WiFi. (I have a couple knocking around in a box.)

    What makes the G1 unique?

  25. One of the coolest things about this chain of dominoes is that Google itself doesn’t have to win or end up with control of anything for the future to play out as described. It’s not even necessary that Android itself be the eventual dominant cellphone platform. All they have to do is force the competitive conditions so that whatever does end up dominating is as open as Android is.

    Early signs are on the horizon that this is starting to happen. Meanwhile, AT&T is trying to force the issue the other way by mandating Symbian on all phones sold through them (the iPhone doesn’t count).

Leave a comment

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