So you consider bug 89572 required for enabling sub pixel layout on GTK as well? Do you see the same issue that I observed initially that there's too much rebaselining required on GTK without that patch? Maybe together we can make another attempt and working with the Chromium gardeners to get it landed. Last time I tried I couldn't quite get them to rebase these 6000+ cases - but if we're two (ports), we might have more luck.

(In reply to comment #1) > So you consider bug 89572 required for enabling sub pixel layout on GTK as well? Do you see the same issue that I observed initially that there's too much rebaselining required on GTK without that patch? > Maybe together we can make another attempt and working with the Chromium gardeners to get it landed. Last time I tried I couldn't quite get them to rebase these 6000+ cases - but if we're two (ports), we might have more luck. I haven't really checked out how this patch affects the number of tests that would need rebaselining, but it wouldn't come around as smart to enable the subpixel layout, rebaseline gazillion of tests, get that patch landed and rebaseline gazillion of tests again. I'd rather help out the Chromium gardeners do their rebaselining when this patch lands.

As you've probably seen on some other bugs, we've enabled it on EFL, and bug 89572 wasn't actually the core of the issue. Updating FreeType was. And no rebaselines required on the Chromium side. However, after my discussions with Werner Lemberg, there will be another change in FreeType that might affect baselines - so, if you plan to enable sub-pixel layout on GTK, I would wait for FreeType 2.4.12 - or in other words any release that includes http://git.savannah.gnu.org/cgit/freetype/freetype2.git/commit/?id=e0469372be3870a5ad60b2c4586e9c281357bd28

Created attachment 198186 [details] Patch

Comment on attachment 198186 [details] Patch Clearing flags on attachment: 198186 Committed r148793: <http://trac.webkit.org/changeset/148793>

All reviewed patches have been landed. Closing bug.