Google's Locked-Down Honeycomb Sets a Challenge for Open Source
Google's restrictions on Honeycomb force the question of whether anyone can create a great open-source mobile UI.
There have always been two Androids. Lazy journalists - including myself - have called Google's smartphone OS free and open source, but that's never been the whole story. Google's apparent decision last week to puts a quiet division front and center, and it throws down a gauntlet that I would love to see open-source advocates finally meet.
Let's get this "open Android" thing out of the way first. There are really two Androids. The first - let's call it Android-O - is an open-source project that Google contributes a lot to. The second - Android-G - is a proprietary Google project that happens to frequently ingest and excrete open-source code.
At any given moment, the latest, hottest Android phones and devices are running the closed Android-G, not the open Android-O. That's always been the case. Every new version of Android is introduced with Android-G devices, and eventually, once Google's mind has moved on to other things, that code gets dumped out into the Android-O repository.
That's one of the reasons why so many phones appear to get their Android upgrades "late." Google picks a few choice friends to get the Android-G code for a given version. Everyone else has to wait until the stew gets a little cold and Google dumps it into the Android-O pot, at which point they add their own seasonings and run to the microwave.
Google's become unusually strict with Honeycomb, though, and that's because the tablet market is very different from the phone market. The phone market has a shifting cast of minority players with different strengths. The tablet market, on the other hand, is dominated by one big gorilla: Apple.
The Open Mobile Challenge
The world is littered with the corpses of open-source mobile projects. Nokia's Maemo, Intel and Nokia's , LiMo, , and TuxPhone have all failed in the market so far. Back in 2009, I said that "" because wireless carriers don't like the unexpected, dynamic nature of open-source projects.
But this time, I think the problem is different. Going up against Apple, the tablet leader, Google realized it needs an industry-leading UI and a consistent brand experience for Android on tablets.
And open-source projects, as is well known, have serious problems creating industry leading UIs. For one thing, open-source projects tend to attract hard-core programmers who love adding features, not visual visionaries. But possibly more importantly, a great end-user experience is often about editing - about making things fit to a consistent vision, which is much easier when there's one consistent vision driving the project.
The only open source projects I can think of with a world-class UI are run by Mozilla, which is an extremely rare thing: a strong, guiding nonprofit hand. Android is managed by Google, a for-profit company whose primary interest is making more money for its shareholders. So Google made a calculation - it's more important to have a guiding hand behind UI quality control than to encourage an open-source community.
Linux fans are now about to erupt with "Ubuntu!" But Ubuntu hasn't had the broad, mainstream breakthrough that many of us hoped it would have a few years ago. I'm going to go out on a limb and say that's not because Ubuntu doesn't have an attractive basic UI - I think it does, in large part because of good governance - but because as soon as you start attaching third-party components, that ease of use begins to fray.
What's happened with non-Googleized Android has only proved my point about open UIs. Go look at our tablet reviews; they're a shambling, zombie-like army of bad products, with only the Samsung Galaxy Tab (designed with some Google input), Motorola Xoom (Google's flagship) and the Apple iPad standing out.
Of course, there are super-successful open-source products. They're just not super-successful in areas where an elegant UI is the primary consideration. Android-O is having success in embedded devices, for instance.
Will this be the end of Android-O? Of course not. Google still sees value in soliciting community input, and the Android-O project lets Google proliferate Android onto many devices that the company wouldn't otherwise be able to find the time to incorporate code for.
But to face down a tablet market where UI expert Apple is dominant, Google clearly decided it needed to raise its UI game - and open-source isn't the right path to do so.
I'd love to see this become a challenge for open-source advocates to finally face down one of the biggest challenges for the philosophy. It's long been proven that an open source app can be powerful, efficient and elegant. Can it be beautiful and easy to use, too?