Changeset 149548 was committed without updating the SunSpider 1.0 website. A patch will follow.
Created attachment 205405 [details] Patch
Comment on attachment 205405 [details] Patch This makes a lot more sense to me.
I think we owe it to the community to do proper versioning here. Please prepare a SunSpider 1.0.1, and make sure that SunSpider 1.0 still has the bug. In fact, I'm shocked that you guys landed the fix in the SunSpider 1.0 directory in the first place - I would revert this. The whole point of having versions is that running a particular version always gives the same results.
Changing the bug title to reflect the upcoming work. A patch will follow.
Created attachment 206011 [details] Patch So, Filip, we reverted the SunSpider 1.0 change and prepared a new version 1.0.1, as you suggested. I will appreciate your input. Is a public announcement, or other things, needed? We are affected by this issue, so we need to move on and get it fixed.
(In reply to comment #3) > I think we owe it to the community to do proper versioning here. Please prepare a SunSpider 1.0.1, and make sure that SunSpider 1.0 still has the bug. In fact, I'm shocked that you guys landed the fix in the SunSpider 1.0 directory in the first place - I would revert this. The whole point of having versions is that running a particular version always gives the same results. I'll leave the new r+ to Filip since it's his request, but really SunSpider 1.0 is broken. This isn't a new version, it's just a fix of the old one.
Filip, Can you review new patch? Its just a kind reminder.
Comment on attachment 206011 [details] Patch This seems very reasonable to me.
Comment on attachment 206011 [details] Patch Thank you for the review. Committing. By the way, I noticed this change fixes the 3d-morph execution on Windows Phone also.
Comment on attachment 206011 [details] Patch Clearing flags on attachment: 206011 Committed r153824: <http://trac.webkit.org/changeset/153824>
All reviewed patches have been landed. Closing bug.