You need to
before you can comment on or make changes to this bug.
The URL (http://jsbin.com/uzesun/3/) is a test page using an extention to Modernizr to add support. Only Opera and IE9/10 should return 'true', but WebKit does as well. Safari 4 does not return true, so it seems that this is a problem in later versions of WebKit. Chrome 8 returns true, but I did not test in anything earlier than that.
This is a spec violation:
Simon, can you inform the layout reviewers to check for this kind of issue when reviewing patches for CSS support, in general? This kind of bug creates a pretty serious forwards/backwards compatibility problem since it prevents the use of the forwards-compat parsing rules to give fallback.
This is work in progress via bug 27570. We can't guarantee that WebKit nighties won't be in this state, but a shipping browser should never be.
Is anyone dealing with this bug? Chrome 18.x and Canary (20.x) are still spitting out a false positive when these values are feature detected. This is a major roadblock to using this feature because it prevents Chrome from properly displaying a fallback or polyfill since it thinks it works.
The repeat round and space were fixed here: https://bugs.webkit.org/show_bug.cgi?id=119080 and here: https://bugs.webkit.org/show_bug.cgi?id=119324.