[Page]Up/Down should not commit a CSS property value altogether but instead apply it to the style and continue editing. Patch to follow
Created attachment 76660 [details] [PATCH] Suggested fix
Comment on attachment 76660 [details] [PATCH] Suggested fix View in context: https://bugs.webkit.org/attachment.cgi?id=76660&action=review > WebCore/inspector/front-end/ElementsPanel.js:-89 > - this.sidebarPanes.styles.addEventListener("style property toggled", this._stylesPaneEdited, this); _stylesPaneEdited is now only called from _styleSheetChanged, please get rid of _stylesPaneEdited
Created attachment 76665 [details] [PATCH] Fix reimplemented correctly pfeldman@ has explained the idea behind the event handling. This patch reverts the erroneous fix from r74038 and augments the stylesheet onRevert handling
Comment on attachment 76665 [details] [PATCH] Fix reimplemented correctly How do we test this?
Created attachment 76751 [details] [PATCH] Test added
(In reply to comment #4) > (From update of attachment 76665 [details]) > How do we test this? Eric, could you please not blindly r- inspector changes that are missing tests without meaningful review comments? They go away from my to review report and this slows down the development. If you are willing to help us with the testing, as you know, inspector tests became flaky at some point and something in the harness needs to be fixed. I already had similar discussion with Adam the other day.
Committing to http://svn.webkit.org/repository/webkit/trunk ... M LayoutTests/ChangeLog M LayoutTests/inspector/styles-add-blank-property-expected.txt M LayoutTests/inspector/styles-add-blank-property.html M WebCore/ChangeLog M WebCore/inspector/front-end/ElementsPanel.js Committed r74185
http://trac.webkit.org/changeset/74185 might have broken Leopard Intel Debug (Tests)