Bugzilla – Bug 21444
FutureResolver's resolve algorithm uses wrong resolver in new handlers
Last modified: 2013-03-31 12:50:42 UTC
The resolve algorithm currently states:
If value is a future, run these substeps and then terminate these steps:
Let accept be a new future handler whose type is resolve and resolver is value's associated resolver.
Let reject be a new future handler whose type is reject and resolver is value's associated resolver.
Append accept and reject to value.
When implementing this in WebKit, I noticed that it did not work. I may be mistaken but I believe the new handlers' resolver should not be "value"'s associated resolver. Instead, it should be the *current* resolver.