King Ludd #4 : Android is Dead

Tweet

With Mobile World Congress in Barcelona kicking off this week, the assertion that Android is dead in the face of an inevitable onslaught of new devices might be seen as slightly out there, even compared to the Rainbow Islands article last week. One time Android fanboy that I am, sadly, it’s the truth. With iOS 7 languishing in a sea of user dissatisfaction and Windows Phone failing to make any dents on the market, Google’s main motivations to improve have fallen away.

Android was, briefly, an independent company formed by Andy Rubin. Prior to Android he was known for having founded Danger (of Sidekick fame, or more accurately of Paris Hilton’s Sidekick getting hacked fame) and having worked at an obscure but strangely influential company called General Magic, alongside people such as the Nest founders. After a brush with near backruptcy, and long before a usable product had appeared, Android was acquired by Google in August 2005 with Rubin continuing to run the unit at Google as a fairly independent enterprise.

As such, he’s responsible for having led the development of what is now the world’s most widely deployed mobile operating system, but he failed to secure a position to prevent a wildly costly distraction in the shape of the Sun/Oracle lawsuit over usage of Java (not helped by various ex-Sun employees showing up on the Android team at Google).  Evidence from that trial shows what amounts to bullying of the then independent Motorola not to adopt the usage of a non-Google source for location services (in that case Skyhook) or to lose the right to install Gmail and other Google products on Motorola phones. The later acquisition of Motorola, largely to acquire intellectual property to protect Android device manufacturers from being sued by Apple, Microsoft and others, was expensive and not as dramatically successful at preventing the ongoing patent litigation-fest as hoped.

Whatever your position on Rubin, he’s not in charge of Android anymore, but runs Google’s nascent robotics unit. Instead Android is now subservient to Chrome under Sundar Pichai. How exactly that went down is anyone’s guess, but Sundar is strongly of the view that developers only want to write applications once for a common platform instead of for Android, iOS, and other platforms separately. And which platform can that possibly be? Well, Chrome of course, because by this logic if you can run an “App” in Chrome then it will work on anything Chrome works on, including Android, iOS, and your PC/Mac.

This clearly isn’t the reality today, so the Chrome team are busy trying to make it one. There are many technical obstacles in the way, largely based in two problematic areas: html layout engines are by necessity large and complex (much more so than native user interface toolkits) and JavaScript is not exactly known for performance, even when implemented by people as talented as those behind V8, which is the part of Chrome responsible for that initially revolutionary burst of JavaScript speed. JavaScript has problems being efficiently converted into the underlying machine instructions, since in certain situations it is essentially ambiguous until right before it is executed, meaning the computer has to do more work. This has led Google to create the “Dart” project, which is an all new language for writing programs that run in Chrome but that can run faster, and because it wasn’t designed in a week (like JavaScript famously was by now Mozilla CTO Brendan Eich while he was employed at Netscape) the long term view was considered, and so it is actually easier to write large programs in too. The downside is pretty much no one outside of Google has any interest in Dart whatsoever, and thanks to their talent for losing interest in things just before they look finished you’d have to be a special kind of nuts to bet the farm on it.

Then, of course, there is the strange beast of Chrome OS. This system is essentially Linux + Chrome, with a lot of the usual Linuxy crap removed (I’m especially pleased to say there’s nothing even resembling X Windows in there) and locked down to prevent breakage by soccer moms. Chromebooks, machines with Chrome OS preinstalled, are currently the best selling laptops around, and have easily the best trackpad implementations this side of a MacBook. Chrome OS is strangely compelling in spite of limitations, especially for developers in unlocked mode, and is likely to be a subject of future rantings.

In a beautiful twist of the embrace, extend, extinguish knife that Microsoft used to wield so effectively, Chrome on Windows 8 can run in a full screen mode that is to the untrained eye indistinguishable from Chrome OS, thus turning any Windows 8 machine into a full on Chrome machine, should that user want to, easing any eventual switch away from Windows.

To pile on the confusion Mozilla believe they’ve already got there with Firefox OS, which runs only web apps on seriously underpowered phone hardware, in typically patronising fashion, for use primarily in developing countries since those of us that can afford to avoid it would do so. While the world would be better for a strong Mozilla this technically inept direction is merely going to prove an expensive waste of time when they’d be better off aiming at their far closer in spirit competitor Chrome OS in the netbook market.

The problem is this leaves Android in a void where it’s useful today but Google clearly don’t see it as the future (rumour has it we’ve seen the last of the Nexus devices), and are embarking on what is quite likely to be a wild goose chase to persuade the world to use Chrome as an app platform instead, with the only likely way to provoke further major investment in Android being if Windows Phone suddenly becomes popular. The Chinese continue to develop their Android ghetto, Nokia may even launch an Android phone, but the Google-centred Android-ecosystem is unlikely to be the source of anything interesting from now on.

Nigel Birkenshaw runs atomirex.