As someone who's been working with Android from the beginning, I have to say: About Time.
This is a required step if Google wants the platform to be taken seriously. The only way to keep fragmentation out is to exert a fair amount of control over the OEMs and Carriers who wish to differentiate (or Fragment) Android for their own purposes.
It's going to be a fine balancing act for Rubin. If he pushes the carriers/oems too far, they'll walk. If he doesn't push them hard enough, the platform will disintegrate Java ME style.
Google is using the only leverage they have (early access and the google apps) to make the platform one worth developing for. I, as someone who makes his living doing it right now, am all for this move.
As much as people online complain about fragmentation, there is no evidence (yet) that the general consumer cares that much. "Google experience" devices are not outselling the customized devices.
People need to tone down the rhetoric, in my opinion. Android already is being "taken seriously". Let's not make small problems out to be big problems. Android doesn't have any big problems, as it's being adopted by manufacturers and consumers at a rapid pace.
Perhaps that statement came out slightly more hyperbolic than I meant it.
The fact remains that Android's fragmentation is something I deal with on a daily basis. Any time spent fixing issues that crop up on particular phones is time I don't spend adding features to our application. Keeping customers from knowing the joys of fragmentation is one of the things I spend a lot of thought and work at.
Yeah, that's fair, and I don't want to fall on the side of "there's nothing to see here" either. Talking about a platform's deficiencies is essential. I'm just speaking to the general "fire! fire!" rhetoric that comes out a lot of the time.
They were only problems in so far as they damaged the Windows brand name and lowered consumer confidence. There's no evidence the same is happening here.
The source will be released when Google is happy with the product. For people to be up in arms that they aren't releasing the source to an unfinished product is ridiculous. Seems like they are damned if they do, damned it they don't.
>The source will be released when Google is happy with the product. For people to be up in arms that they aren't releasing the source to an unfinished product is ridiculous. Seems like they are damned if they do, damned it they don't
The Motorola Xoom is already out.. so how is Honeycomb an unfinished product if it's in consumers' hands? The basic spirit of FOSS is that users and developers should be able to modify the code. The article you linked to says this:
>The lack of Honeycomb code availability is especially bad for enthusiasts who were hoping to be able to install custom firmware on their Android tablets. Without the code, it will be difficult for the modding community to produce custom builds that fix the software problems that plague the Xoom and other upcoming Android tablets. Users who were looking forward to better Honeycomb builds for the Nook Color and other budget devices are also going to have to wait.
>For now, only a privileged few hardware vendors will have access to Honeycomb while the rest are left with uncertainty about the future of the platform. Even after the matter is resolved, the fact that Google is willing to withhold source code at its whim for competitive reasons is going to continue to cast a dark shadow over the company's increasingly hollow claim that Android is an open platform.
I don't see how the above is wrong.
>The source will be released when Google is happy with the product.
Doesn't Android use the Linux kernel which is GPL'ed (among other parts)? Can they legally withhold code for a shipping product by saying the software isn't finished?
> Doesn't Android use the Linux kernel which is GPL'ed (among other parts)? Can they legally withhold code for a shipping product by saying the software isn't finished?
Yes, the kernel is GPL, so they have released that source. It's the rest of the platform they are withholding, which is generally under an Apache license.
> The Motorola Xoom is already out.. so how is Honeycomb an unfinished product if it's in consumers' hands?
It's unfinished because it doesn't ship with major features that it claims to support such as LTE and Adobe Flash. Every review has said that the Xoom feels like a beta at best and it is simply not ready for the masses, and a large part of that is due to Android 3.
The anger is explained in the introduction to that article. Google presented Android as an open gift to the world in a fight against strictly controlled platforms, yet Google is now withholding source code and requiring final approval over modifications to the operating system.
That they're allegedly holding up Android phones which utilize a rival search engine, Bing, is especially troubling. Is Android supposed to be an open platform or a Google platform?
> "Google is using the only leverage they have (early access and the google apps)"
Walking isn't such a big issue these days. Which is likely why Google feels secure doing this. Windows CE is dead. Phone 7 is struggling. Nokia essentially gave up, removing MeeGo as a concern.
So where does, say, HTC go if Google pushes them to drop the Sense UI?
They certainly don't have the software expertise to fork and run. None of the Android OEMs do. At the rate mobile is advancing, they'd be irrelevant in 18 months. And there is literally no-one else who is delivering a product they can use to maintain their sales.
I honestly wonder if Google isn't more concerned about Amazon than the OEMs.
As for carriers... they're only a real issue in the US; where Android's growth isn't that strong anyway. I mean, is it even possible for AT&T and Verizon to support Android less these days? I haven't seen a single Droid ad since January.
HTC in particular has a long history of keeping a few different systems in their lines. They can prioritise Windows Phone more. But with Nokia there already they might want to keep it approx. 50-50 anyway.
Others like Motorola and Samsung, I'd expect them to stick to Android whatever happens, at least for the high end. This is why Google is doing this.
Is fragmentation really the issue? I think the issue is quality. Every android phone I've used was very different from the last. My friends G1 isn't a lot like my Vibrant and my Vibrant isn't a lot like my old EVO. Three different GUIs! Heck, Samsung decided to put in their own filesystem on the Vibrant and all Vibrant owners suffer from random lag.
Each phone has a different camera app, different gallery, different everything. Its annoying. They're all more or less phoned in - if you pardon the pun.
What google really needs to do is address its ugly stock GUI and put some shiny on there. Put in a decent media player. Put in lots of Apple-quality apps so that OEMs don't feel the need to completely redo everything because the stock android looks like something a defense contractor would make.
Maybe they can even make custom GUI enhancements run in userspace so that updates don't require redoing them. There should be a skinnable layer on top of system widgets. Heck, once you have that then google can start pushing out its own updates to phones and OEMs won't need to wait 8 months to port them over.
One can dream, I guess. Or I can buy an iphone again or move to Win7. My little Android experiment isn't really going anywhere. Its just as locked down as an iphone in practice and I suffer with Samsung's or HTC's low quality enhancements.
Agreed, the fragmentation might be caused by the semi-quality of apps and looks. But the mess that the hardware makers are forcing upon arent that much better at the moment.
I've been amazed each time i've put on cyanogenmod to replace the default android that comes with the phone with how much more useful it gets. Smoother, better working and less lag.
I would love to see htc(or any other hardware maker) ship their phones with a cyanogenmodversion that will track cyanogenmods updates. That might even save them some work.
This is a required step if Google wants the platform to be taken seriously. The only way to keep fragmentation out is to exert a fair amount of control over the OEMs and Carriers who wish to differentiate (or Fragment) Android for their own purposes.
It's going to be a fine balancing act for Rubin. If he pushes the carriers/oems too far, they'll walk. If he doesn't push them hard enough, the platform will disintegrate Java ME style.
Google is using the only leverage they have (early access and the google apps) to make the platform one worth developing for. I, as someone who makes his living doing it right now, am all for this move.