Mozilla, where is Firefox OS?

There is a strong chance that Windows 8, when it arrives in 2012, will be the last conventional, installed-application, file system-driven consumer operating system. With the popularity of mobile computing, there has been an overarching shift towards simpler, application-oriented interfaces -- and on the desktop, the last few years have seen a dramatic shift in how we interact with our computers. Gone are the days when you would run multiple applications; gone are the days when your operating system would have to carefully manage memory and multitasking -- today, the vast majority of users simply use a web browser.

In recent years, the web browser has become more and more like an operating system. HTML5, with local storage, audio controls, and other low-level APIs make an ideal framework for browser (web) apps, and JavaScript performance has come along in leaps and bounds. WebGL provides a high-performance pipe to underlying graphics hardware, and SVG and CSS allow for apps that fit all form factors and screen sizes. All of the underpinnings are there -- and if you need confirmation, just take a look at Google, which has strapped Chrome to a Linux subsystem and called it an operating system.

Rather than going down the Chrome OS route, though, Microsoft has opted for a hybridized approach with Windows 8. From what we’ve seen, it looks like the Windows 7 kernel will still underpin the operating system, but on top will be an HTML5-powered tiled interface. Apps for Windows 8 will be developed in HTML5 and CSS, using new tools that Microsoft hasn’t yet released, and executed using Internet Explorer 10, or a variant of it. The key to this approach is the inclusion of the Windows 7 subsystem: Microsoft can’t just give up on the vast library of legacy Windows apps, but at the same time, Microsoft needs to usher developers towards cross-platform HTML-based apps if it wants to survive the mobile computing land rush. Basically, Windows 8 will be Windows 7, but with a transparent web browser that loads at boot time and executes HTML-based apps.

This concept is so simple, yet powerful and graceful at the same time. If a developer writes an HTML5 app, it will run across every Windows 8 form factor, from desktops to laptops, to ARM netbooks and tablets -- and to top it off, Windows Phone 8 is also expected to support these very same HTML5 apps. In other words, developers can write once and run anywhere -- and using the Windows 8 app store, monetization will be easy, too.

But therein lies the rub: if you remove the operating system -- or at least make it transparent enough that the browser becomes the platform -- then suddenly every piece of software works across every piece of hardware. Web apps are, fundamentally, installed websites -- and websites, as we know, work on any device that can run a web browser. Developers will writhe with delight at the thought of platform agnosticism, but the suits at Redmond... The problem -- and this is the stumbling block that Microsoft will run into -- is that web browsers can’t be monetized. Browsers have always been free because you pay for the OS -- but if the browser is free, and the browser is the OS... then where does that leave Microsoft? Google makes its money with advertising, which is how it can justify Chrome OS -- but if a developer can write one app that works perfectly on Windows 8 and Chrome OS... why would you pay for Windows 8?

Which leads us neatly onto Mozilla, the creators of Firefox. Unlike Google, Microsoft, Apple, and Opera, Mozilla is a charity. Mozilla is in a unique position: it doesn’t need to make a profit, it doesn’t need to please its shareholders, and it doesn’t have to dance around tricky, conflicting areas like privacy and digital rights management. Mozilla exists for one purpose only: to promote innovation, openness, and opportunity on the web. Google could turn around tomorrow and put an immovable ad in the bottom corner of every page. Microsoft could block Flash video and only allow Silverlight. Opera could bombard us with pop-ups for adopt-a-seal programs. But Firefox -- Firefox will always be there, free, unencumbered, and open source.

Firefox might have started off as an open alternative to Internet Explorer, but today it represents hundreds of millions of internet surfers that believe in the web’s openness, and its continued innovation. In a world that is rapidly moving towards browser-based computing, and with the web and our surfing habits growing steadily more commercialized, Firefox is really the only bastion of openness. It doesn’t even matter that its market share seems to be shrinking: just the sheer fact that Firefox exists is enough to keep other browser makers, and the web’s commercial interests, in check.

Mozilla, in fact, is in the perfect position to create an open, free alternative to Chrome OS and Windows 8. It could be called Firefox OS. Mozilla could even build it on Chromium OS, the open source brother of Chrome OS. You could strip out the Chrome browser and replace it with Firefox. You could retain most of the features that make Chrome OS awesome, and replace proprietary bits with open source equivalents. There would be a few teething issues, but most of the work has already been done by Google and the Chromium team. If that seems like an unconscionable bastardization, Mozilla could just build a custom version of Ubuntu or Gentoo that runs Firefox on boot.

There are still some hurdles to overcome, though. With the exception of IE9 and its Chakra JavaScript engine, web browsers are still fairly rudimentary when it comes to scheduling and JIT compilation; in Chrome and Firefox a slow script can slow down the entire browser -- and in a browser-based OS, that’s a problem. Furthermore, because every JavaScript compiler operates in 32-bit space, no browser process can use more than 2GB of memory -- a problem, if you want to turn Photoshop or Final Cut Pro into a web app. There is still the matter of raw performance, too: JavaScript in 2011 is a lot faster than just a few years ago, but it’s still many times slower than managed languages like Java and C#, and in some cases hundreds of times slower than a native language like C++.

There are solutions like Google’s Native Client which allow for a mix of C++ and JavaScript, but Mozilla has made it clear that it has no interest in such hacks. If JavaScript is to become a primary programming language, Mozilla will need to build core, C++ libraries into Firefox that JavaScript can leverage. Microsoft, incidentally, if it is serious about using JavaScript for “native” Windows 8 apps, will have to do the same thing -- and if we’re lucky, the same feature might percolate down into Internet Explorer 10.

On the HTML front, HTML5 makes deep inroads into historically problematic areas like bookmarking and navigation, but we shouldn’t forget that browsers still have a long way to go until their nascent implementations of HTML5 are stable, standardized, and feature complete. We’re at least a year or two away from having web apps that render equally across the five main browsers. Even then, with the big three -- Microsoft, Google, and Mozilla -- using different app packaging techniques, we’ll still be some way away from that write-once-run-anywhere utopia.

There are a few hills to climb before a free and open browser-based OS can take off, then, but nothing that even comes close to being insurmountable. The problem as it stands -- if you can call it that -- is that Mozilla is focused on making its browser as all-round awesome as possible. Instead of focusing on native-like JavaScript performance and hammering out HTML5’s quirks, Mozilla is working on tangential features like better privacy controls and a built-in contacts manager. Firefox 5, 6, and 7 are all shaping up to be excellent browsers, but they’ll be useless when we shift away from the big-OS-and-installed-software paradigm.