Created attachment 276635 [details] Screenshot None of the other DataGrids have this issue (the grey border should butt up against the first row). This is likely a regression from the switch to half-pixel borders on tables.
<rdar://problem/25778260>
Created attachment 276866 [details] [Image] Bug There is another issue with DataGrids' half-pixel borders. They aren't aligned with the sidebar border. As per discussed in person, we should just use 1px horizontal borders for now.
Created attachment 276880 [details] WIP
Comment on attachment 276880 [details] WIP View in context: https://bugs.webkit.org/attachment.cgi?id=276880&action=review > Source/WebInspectorUI/UserInterface/Views/DataGrid.css:82 > + box-sizing: content-box; This should not be necessary. However, without this, the table header is off by half a pixel—it's 22.5 pixels tall. This is bizarre and could be a recent WebKit bug. I'm working on a reduction for Zalan.
Created attachment 277766 [details] [Image] Half pixel off (In reply to comment #4) > Comment on attachment 276880 [details] > WIP > > View in context: > https://bugs.webkit.org/attachment.cgi?id=276880&action=review > > > Source/WebInspectorUI/UserInterface/Views/DataGrid.css:82 > > + box-sizing: content-box; > > This should not be necessary. However, without this, the table header is off > by half a pixel—it's 22.5 pixels tall. This is bizarre and could be a recent > WebKit bug. > I'm working on a reduction for Zalan. This bug was introduced by bug 154437. However, reverting that change does not fix the issue! See the attached screenshot.
Comment on attachment 276880 [details] WIP Is this patch ready for review? It's named 'WIP'. Clearing r? for now.
(In reply to comment #6) > Comment on attachment 276880 [details] > WIP > > Is this patch ready for review? It's named 'WIP'. Clearing r? for now. No, it isn't. It does look like a bug in WebKit and I'm still working on a reduction.
(In reply to comment #5) > Created attachment 277766 [details] > [Image] Half pixel off I cannot reproduce the issue on this screenshot any longer. I assume it was a WebKit rendering bug and now it's fixed.
Created attachment 280190 [details] Patch
Created attachment 280191 [details] [Image] With patch applied
Comment on attachment 280190 [details] Patch Attachment 280190 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.webkit.org/results/1414901 New failing tests: fast/scrolling/scroll-container-horizontally.html
Created attachment 280198 [details] Archive of layout-test-results from ews107 for mac-yosemite-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: ews107 Port: mac-yosemite-wk2 Platform: Mac OS X 10.10.5
Created attachment 280200 [details] Patch Seems like an unrelated flaky test.
Comment on attachment 280200 [details] Patch Clearing flags on attachment: 280200 Committed r201540: <http://trac.webkit.org/changeset/201540>
All reviewed patches have been landed. Closing bug.