What Does Winning Look Like?

I’ve been thinking: what would the world look like if Mozilla won? Perhaps something like this:

Users at home, at work and on the move are using their choice from a number of available operating systems on a variety of hardware and, on top of them, a number of web renderers, according to their preference. All of these stacks have good support for modern web standards. It’s possible and commonplace to write a web application which runs pretty much everywhere. Web apps have most or all of the capabilities of native apps. No one company has control of the web or the standards – standards are collaboratively developed using an open process. Users have one or many online identities, real or pseudonymous, and associated profile data, which they own and control independent of any company or organization, and which cannot be taken away from them. Every site supports that identity mechanism for interacting with it, and users can take their data easily between systems.

I think this vision of winning is widely shared across the Mozilla project. Most of us would like everything above (and a pony too :-)). I suggest that any differences of opinion regarding our future strategy are not about what winning looks like, but about how we get there, and avoid losing.

So, if that’s what winning looks like, what does losing look like? Clearly, if you invert all of the above, with closed systems, no choice, lagging and incompatible HTML engines, single vendor control of identity and so on, that’s clearly a loss. But I doubt it’ll be that clear cut. If we ‘win’ on some points and lose on others, how do we decide whether we’ve won or lost?

Here’s one plausible future scenario which looks to me a lot like being on the way to losing:

There are a variety of platforms and operating systems, but control of the single web renderer on each platform belongs exclusively to each platform vendor. On minor platforms with less clout they see the value of the web as a competitive mechanism (few people write apps specifically for those OSes), but on the major platforms, vendors prefer people to buy (monetizable) apps through the platform store. Standards are collaboratively developed, but no-one can force vendors to implement them. So, as vendors innovate in their native app APIs instead, the capabilities of native apps and web apps on those platforms diverge. This forces developers who want to use new platform features to write native apps, which (via network effects) entrenches those existing dominant platform players and hinders choice. People can still write web apps, and they run on most platforms – but only with whatever capabilities the vendor allows the web platform to have on each system, and the integration and experience is poorer. And occasionally gets worse, in the name of ‘security’, or some other excuse. There are solutions for portable identity, but API incompleteness makes it hard to give users a smooth experience using them.

The broad outlines of this situation are actually familiar – we’ve been here before. Back in 2001, Microsoft was the dominant force, and IE 6 the majority browser. They much preferred people to write apps for Windows (or for proprietary extensions in the browser) than cross-platform apps – they thought they’d seen off that risk when they stabbed Netscape and Windows-ized Java. (To some extent, monopolies legislation prevented them turning the screw more than they did.) Web apps were sort of possible, but the platform had nothing like the capabilities of native code.

Mozilla fixed that – by bringing innovation back to the browser on the dominant platform, by showing people there was a better way – first in usability, and then in platform features. We’ve vastly improved the capabilities of the web platform – and Microsoft has had to follow. Without us and our allies, there would be no web innovation.

Could we end up back there again? In a future, it might not be Microsoft – it might not even be a single company. (In fact, if Windows Phone 7 doesn’t take off, Microsoft may see open web apps as their way to compete, and become a convert!) But right now, in the mobile space, some vendors strongly control their platforms and renderers – and there may be others in the future. With only their browser available, innovators can only build things on the web that the platform vendor allows you to build – and it’s in the vendor’s financial interest for native apps to be a better experience than web apps. The long-term future for the open web and web-based innovation on these platforms does not look good.

If that becomes the norm, it looks like losing to me. How do we avoid losing like this?

11 thoughts on “What Does Winning Look Like?

  1. I don’t think it’s as simple as winning vs. losing. It’s a valuable exercise, but I suspect Mozilla’s mission will continue to be needed — and evolve — as the netscape changes.

    Somewhat arguably, we “won” the challenge presented in 2001… The web is vibrant, alive, and rich with competition. But the dangers and risks have also shifted. Would anyone have considered mobile such an important area for the web in 2001?

    Perhaps the right question to ask is where we want to web to be in N years (for small values of N). Given the stakes and competing interests, I’m not sure there ever will be a winning state. But there are clearly a myriad of losing states.

    The challenges are huge, and the cost of failure is high. That’s why I love working for Mozilla.

  2. Dolske: Right; it’s not as simple as winning and losing a football game. But it is a useful exercise to take our current understanding of the mission and translate that into “winning”, in terms of the landscape as we see it now.

    And yes, there are multiple ways to fail. :-| The one I give above is just one. But it’s a quite plausible one, I think.

    BTW, “as the netscape changes”? Amusing slip… :-)

  3. Currently the biggest threat for Mozilla is to lose their community by bad communication and sometimes arrogant fuck you (right Asa?) statements. And the community is Mozilla’s biggest asset.

  4. @Pete: It’s not about the threat to Mozilla and it’s not about whether Mozilla wins or loses. It’s about the threat to what Mozilla stands for and it’s about whether all of us win or lose. That’s more important than a few unempathetic, blunt individuals.

  5. Note that the preview talks of Windows 8 make it look a lot like one of those in the non-win scenario you pointed out. And ChromeOS looks like a potential one of those as well, as does Android and of course iOS. The difference to 2001 is that it’s multiple monolithic, mostly tightly controlled platforms.
    As strange as it sounds, as Mozilla always said we don’t want to get into any OS “business”, an effort like B2G or Webian will probably be needed to break that somewhat, though I hope our “web application runtime layer” will be able to run on a number of different base OSes behind it, and I also hope we’ll have a really open stack to base upon, which Android is not, even if it’s the best available to experiment right now.

  6. Gerv, we’re already starting to end there again.

    WebKit, an engine whose Standards implementation is much less than perfect (less than 90% of the CSS 2.1 Test Suite are passed for example)[1] is gaining more and more market share.

    Google and Apple, two of the currently most successfull companies can’t (or maybe don’t want to) improve their standards support to match the level of quality IE and Gecko can provide.

    And it’s not only CSS, a lot of my colleagues are complaining every once in a while about some bug in WebKit even in Script work.

    It’s very sad that Chrome hurts the web so much even though it is only as successful as it is now because of Mozilla and its Mission.

    [1] http://test.csswg.org/suites/css2.1/20110323/report/results.html
    Note: Opera scores less, but they don’t cover the whole Test Suite. Older results show about 94% passing of the Test Suite.

  7. the_dees: Let’s say all your facts are correct. How do we stop ourselves losing in the way I outline? That’s my question for discussion. :-)

  8. @Gerv: If I only knew what the solution has to look like…

    Of course, more Standards are always important. Maybe we need closer relations between vendors. At the moment each vendor has it’s own research department (for example Mozilla created multi-column layout, Apple created Transforms and Gradients, Microsoft created positioned floats). I think we need to work together much earlier in the creative phase. maybe then wen can grasp the need of developers and create a truly creative idea.

  9. We Fight For the User

    My last blog post looked at what winning (and losing) might look like. So now we ask: how do we win? Answer: like this (audio version). Key points and aspirations from Brendan’s talk: Open web apps – cross-platform and cross-renderer Multiple competing…

  10. @the_dees Actually, multi-column layout was created at W3C many years ago (see drafts circa 1999); Mozilla was just one of the first to implement it.

    @Gerv the_dees has a good point wrt standards support, and I think we can make a lot of progress on that front if we coordinate with W3C to make our tests part of the standard test suites and help them make test results more visible and interesting. Right now we’re writing thousands of tests which are only useful to us and we are therefore not using to press for standards-compliance in other engines.