“As the Android kernel code is now gone from the Linux kernel, as of the 2.6.33 kernel release, I’m starting to get a lot of questions about what happened, and what to do next with regards to Android. So here’s my opinion on the whole matter.”
“As the Android kernel code is now gone from the Linux kernel, as of the 2.6.33 kernel release, I’m starting to get a lot of questions about what happened, and what to do next with regards to Android. So here’s my opinion on the whole matter.”
That’s weak.
It’s sad that in this matter Google has not worked at all with the community. They dropped a huge amount of code without consulting the kernel community on any matter, and they haven’t done anything to improve the situation….I hope this situation change…
Android is very different from a typical GNU/Linux system. Here’s a comparison between Android and Palm’s WebOS, which under the hood, is much more similar to a GNU/Linux system in spite of having somewhat more closed-source components: http://mjg59.livejournal.com/111453.html
“WebOS is recognizably Linux in a way that Android isn’t”, according to the article.
Sorry for trumpeting WebOS again, but the comparison at least should be educational, and illustrate why Android code is difficult to integrate with the main Linux code branch.
But gratuitously so, it would seem. Things like writing their own framebuffer infrastructure instead of using the kernel standard – it’s hard to see a good reason for doing something like that, and practically guarantees isolation from upstream.
Yeah, that makes making Android apps that don’t use the Android API more difficult. Early graphical native apps for WebOS just used the Linux framebuffer code and were a straightforward port. (Now that 1.3.5 added native accelerated app support, they use the standard SDL library and/or OpenGLES instead, making Linux games a straightforward port, or iPhone apps get ported easier with the latter API.)
Who says stardard way is best way? Why should dictators of Linux kernel say what is right and wrong, it goes against whole point of open source. What Google did was decide that standard stuff was bad, rewrite own version. They didn’t broke any license stuff since they do offer source code. Whole point of evolution in open source world is to take old, rewrite and then offer that. Just because it doesn’t follow plans of dictators, doesn’t mean it’s bad thing.
Sadly, collaboration instead of forking is the whole point of evolution of open source.
I agree. However Google made a contribution in either form. It is the responsibility of the community to take advantage of this. They provided drivers, something all OpenSource projects need. and the locking mechanism they provide could be possibly more suitable for the desktop than corporate servers. If it cannot be merged upstream there is an issue and I do not believe it is a google responsibility only. Forking is bad, but it can be inevitable and beneficial for the evolution. This is the main problem with the Con Kolivas scheduler. People say that Linux can scale from embedded to desktop and servers. I believe that it is not possible without compromises for the desktop user. Small distros for old hardware can benefit from these two projects. Android contributions should be kept somewhere and not deleted until suitable people take up. Maybe it does not build with upstream. But it is not useless.
Yeah. Whoever said that Android will only benefit Linux. Right.
One thing I find odd is that Microsoft claims that the Linux kernel supposedly infringes some of it’s patents.
Yet Microsoft hasn’t tried to use this as leverage against Google, since both Android and ChromeOS use the Linux kernel.
Well, unless the patent infringements are baseless or Microsoft just doesn’t consider it worth the effort.
Yeah.
I am sure that it was actually Microsoft that caused this conflict too…
Linux is meant to be a general purpose OS capable of powering everything from supercomputers to DVRs. Android is designed to run on embedded systems. these two have very different design goals and constraints. while it would be NICE to have the same source code, i dont see why they have to merge the code.