Temporary fix for the Cocoa-in-renderer problem.

Temporary fix for the Cocoa-in-renderer problem.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=10326

[John Grabowski, 2009-02-24 23:31:08]

This is not the final answer, but it'll end (for now) the throwing around of
patches.  The longer term goal is to remove use of Cocoa from WebKit in our
renderer.  I experimented a bit but was not able to find any reasonable way to
make Cocoa happy.  This CL is the least objectionable thing to do for now, even
with it's problems (e.g. it eliminates dock bounce when Chrome.app launches).

LGTM from any reviewer on the To: list is fine.

[Amanda Walker, 2009-02-25 00:18:32]

LGTM for now, but please open a bug to track coming back to it.

[pink (ping after 24hrs), 2009-02-25 13:59:34]

http://codereview.chromium.org/27108/diff/1/3
File chrome/browser/browser_main_mac.mm (right):

http://codereview.chromium.org/27108/diff/1/3#newcode36
Line 36: // the dock.  These 3 lines make it happen.
would be helpful to have the bug# here for reference so we can mark it fixed or
comment there appropriately.

http://codereview.chromium.org/27108/diff/1/3#newcode39
Line 39: TransformProcessType(&psn,  kProcessTransformToForegroundApplication);
extra space after comma?

[Mark Mentovai, 2009-02-25 14:19:39]

http://codereview.chromium.org/27108/diff/1/5
File chrome/renderer/renderer_main_platform_delegate_mac.mm (right):

http://codereview.chromium.org/27108/diff/1/5#newcode58
Line 58: if (![NSThread isMultiThreaded]) {
FYI, base::InitThreading() from platform_thread_mac.mm does exactly the same
thing.  If this is revived we should share.

(but it's not declared in any header, so we'd need to fix that.)