Not that I'd want to ask strategic questions -- the strategy has IMO been transparent for some time -- but I wonder a couple technical points, which you may or may not wish to elaborate on depending on the positioning of the secrecy wall, and/or whether source will be forthcoming.
Mainly two things. First, how much machinery does this have in common with the android tree, as I'm really just picturing it now as an android configuration with the browser wired to be fullscreen, and lacking maybe dalvik and the phone-UI stack. Like, does it present a different non-X, non-android framebuffer target for chromium? Sounds pointless. Second -- refinement on first -- is there actually much in the way of a different security system under the hood? Beyond what was done in android with the linux parts lying around?
Totally understand if neither of these can be answered in any way other than silence. Exciting direction though. Congratulations once more. Maybe if you delete enough of userspace, it won't suck anymore :)
This is part of a larger topic that, perhaps, we can discuss when I'm in town next month. One less related to the particulars of browsers, and more to the social phenomenon of the independent personal computer atrophying.
I sure hope they don't go non-X. There are advantages and disadvantages to them in dropping X but mostly disadvantages to the wider linux community.
The biggest benefit these days of X is relatively good drivers. They're not great, but all of the pc unix vendors can share most of the driver code. For Google's OS thingy to work on a variety of graphics adapters they'll either need to use X drivers (directly or through other trickery) or work with all the graphics vendors. If they worked with graphics vendors and continued to use the X driver model then everyone wins.
The advantage for Google for dumping X is that they could develop a graphics architecture that would interact nicely with their sandbox model.
I came running here before I'd finished the press release, thinking "whoa, I guess Evan's point of view suddenly won out!" But yeah, this is very different from your vision. But it's exactly what I was expecting you guys to do since Chrome was released.
Exciting vision. I hope to see a product out there soon.
Can you tell me off the top of your head if Chromium WebKit on X11 uses Render acceleration for antialiasing?
I apologize for the off-topic question. In my defense, if you can give me a definitive yes or no, I'm hoping I'll be able to get a bug report to the right people -- WebKit seems immune to a X server bug.
Note that our WebKit port is distinct from the GTK or QT WebKits. We do all our own font rendering, so no, we don't use Render for fonts in particular. (We do use Render for some format conversion.)
If you have any Chrome-related bugs, it's safer to put them in our bug tracker and then let someone on our team promote them to the WebKit tracker if necessary. The WebKit devs sometimes get annoyed about getting our bugs.