It doesn't matter how you try to downplay it. Google shipped a huge useful chunk of kernel changes to Linux, plus a mature, ready, out of the box piece of open source that anyone can fork and use as a foundation to make another free OS, like FireOS, Tizen, FirefoxOS, etc, and they successfully got a platform deployed on a huge number of devices around the world that don't completely lock down who can install what.
Even if they had completely evil motives, good has come out of it. If you're a rabid fanboy who wanted Steve Job's platform to have a monopoly on mobile, it's depressing for you, but it's liberating for everyone else.
Google wanted a platform that prevented Microsoft and Apple from beating them to control of the next big advertising platform.
Nothing more, nothing less. "Open" has been shown conclusively to take a back seat to monetization. It is as much a marketing tactic (in the sense that it's useful to promote the platform, but can and will be discarded, with spin to explain why, when its usefulness decreases) as anything in Microsoft's or Apple's ads.
And it doesn't matter how you try to downplay it: companies other than Google have, whether they intended to or not, accomplished undeniable good through things they've shipped. Apple, for example, has contributed monumentally to LLVM, which is a large and useful public good. But it did so more or less explicitly to screw the GPL.
So, faced with all the major players doing this in some form or another, how, precisely, do you plan to rank them relative to each other in order to discern which one is "good" or at least "least bad"?
For the average consumer, control of "the next big advertising platform" is a meaningless abstraction, it matters to people in the ad buying business, not consumers who generally don't like ads in the first place, and tolerate them mostly to get free stuff. And how does Android "control" mobile web advertising anyway?
But for the average consumer, telling them they can't install App X on their device that they spent $300 on because Apple censors don't like it is much more direct.
No company is perfect, but I can say as a Googler, your cynicism is basically wrong, that open source isn't just a marketing tactic, but that the rank and file at the lowest levels of the company, culturally believe in it. Google has disappointed in the past on the degree of openness, and there is much to improve, so I am not uncritical of it in this regard, but if you were to query the rank of file of Microsoft and Apple with regards to open source, and Googlers, you'd find a lot more zealots at Google.
The world isn't black and white, it's shades of gray, and some are just a much brighter shade of gray than others.
While I work for Google, and was involved in the CPTN and Rockstar stuff (and thus can't comment on this article at all, though I wish I could), i can actually point out that this statement is not right.
First, companies that wish to contribute to LLVM are asked to non-assert patents involved in their contributions. This is covered in the dev policy, and has been for a long time. Their are non-assertions/grants on file for a number of companies now.
As for whether Apple owns any patents on LLVM core, Chris Lattner himself told me Apple held none, which is why they had not filed a non-assertion.
Apple may or may not hold patents on LLVM code that is not "in tree" (so yes, they may have non-public frontends or backends they own patents on), and I never specifically asked about the clang frontend, but I trust Chris enough to know that if there was in-project code, be it clang or core or whatever, he would have followed the dev policy and Apple would have non-asserted them.
There is simply nothing that prevents apple from using patents against LLVM users. We can only hope that none of their numerous vaguely written patents cover any technology used by LLVM or any of their frontends like Clang.
Wow, you used just a single line to describe an entire, complex, (possibly superior) open source operating system?
I'm done reading comments here.