96109 2012-09-07 08:00:46 -0700 [chromium] Global Selection test broken on Chromium Linux 2013-04-09 17:07:09 -0700 1 1 1 Unclassified WebKit Layout and Rendering 528+ (Nightly build) Unspecified Unspecified RESOLVED WONTFIX P2 Normal --- 1 fsamuel webkit-unassigned allan.jensen tony oldest_to_newest 714524 0 fsamuel 2012-09-07 08:00:46 -0700 editing/pasteboard/paste-global-selection.html is broken and should be fixed. 714704 1 tony 2012-09-07 10:34:35 -0700 From https://bugs.webkit.org/show_bug.cgi?id=88238#c49 : There is no code in WebKit that updates the global selection for Chromium. The chromium code that does that the update is outside WebKit, which is why the automated test will not work in the chromium port. -- For non-selection clipboard tests, we use a mock clipboard. We probably need to make a mock selection clipboard for this test. 714706 2 tony 2012-09-07 10:35:16 -0700 Alternately, we should add a method to internals for setting the global selection clipboard that all ports can use (and doesn't depend on X). 714764 3 allan.jensen 2012-09-07 11:12:27 -0700 A(In reply to comment #2) > Alternately, we should add a method to internals for setting the global selection clipboard that all ports can use (and doesn't depend on X). Not sure about internals, while it would be nice if all ports could pass the tests the same; I think it makes more sense to me to add it to TestRunner. For Chromium that only means updating DumpRenderTree and adding something saving and recalling the selection (like the Chrome/Chromium application would). 729375 4 allan.jensen 2012-09-27 03:13:26 -0700 An alternative solution is to move the chromium update of the global selection into WebKit. Unless there is a good reason it is outside, I think that would be the best solution, and would mean we could also unify the code updating of the selection buffer. 729793 5 tony 2012-09-27 13:10:37 -0700 (In reply to comment #4) > An alternative solution is to move the chromium update of the global selection into WebKit. Unless there is a good reason it is outside, I think that would be the best solution, and would mean we could also unify the code updating of the selection buffer. We could do this for DRT, but in a real Chromium. WebKit is in a sandbox that doesn't have direct access to X. We make an IPC call to the browser process that does sets the global selection clipboard. 730408 6 allan.jensen 2012-09-28 02:40:31 -0700 (In reply to comment #5) > (In reply to comment #4) > > An alternative solution is to move the chromium update of the global selection into WebKit. Unless there is a good reason it is outside, I think that would be the best solution, and would mean we could also unify the code updating of the selection buffer. > > We could do this for DRT, but in a real Chromium. WebKit is in a sandbox that doesn't have direct access to X. We make an IPC call to the browser process that does sets the global selection clipboard. My point was to move the code that captures the changes to the selection to WebKit, the part that makes the X calls would still be hidden behind the PasteboardChromium.cpp api. 872493 7 schenney 2013-04-09 17:07:09 -0700 Marked LayoutTest bugs, bugs with Chromium IDs, and some others as WontFix. Test failure bugs still are trackable via TestExpectations or disabled unit tests.