Hi. Thanks for the feedback, Ojan and Ryosuke.
>From the experiments I've been doing, I think it will be a smooth landind. The platform specific editing behavior code paths are not that common in the tests in LayoutTests/editing, and the tests that go through them were converted to use the LayoutTestController::setEditingBehavior machinery, so should not be affected. See for example bugs 47471 and 47472 (among others), where tests touching platform specific editing behaviors were converted to use LayoutTestController::setEditingBehavior. For instance, I have access to tree platform builds: Mac port on SL, and webkit-gtk and qtwebkit on linux. I am able, for example, to run all tests in editing/ in a Mac build after changing the editing behavior to "Win" instead of the default "Mac". Similar situation on gtk and qt builds. Cheers, On 10/27/10, Ryosuke Niwa <ryosuke.n...@gmail.com> wrote: > On Wed, Oct 27, 2010 at 2:17 PM, Antonio Gomes <toniki...@gmail.com> wrote: > >> Well, despite the naming to be used, in the same bug I am proposing >> that the editing behavior of each platform to be set according to the >> OS it is running on. As an example, Chromium when built on Mac would >> have the MacEditingBehavior set, while when built on Windows it would >> have the Window editing behavior, etc. The same logic would apply to >> all cross-platform ports, including QtWebKit. > > > I think this is an improvement. But can we coordinate when landing patches > so that we can maintain chromium bots green? I don't want to see 100+ > editing test failures all of sudden without a pre-caution. This shouldn't > be too hard given that chromium has a try bot. > > - Ryosuke > -- --Antonio Gomes _______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev