[Webkit-unassigned] [Bug 258282] REGRESSION (264594 at main): Broke custom element built-in polyfill by making form properties non deletable

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=258282

--- Comment #17 from Chris Dumez <cdumez at apple.com> ---
(In reply to Chris Dumez from comment #16)
> (In reply to Chris Dumez from comment #15)
> > I wrote this very reduced test case which tries to delete property 0 of a
> > form:
> > https://codepen.io/cdumez/pen/JjeKgLa
> > 
> > Safari before 264594 at main: Able to delete the property
> > Safari after 264594 at main: Not able to delete the property
> > Chrome 114: Not able to delete the property
> > Firefox 114: Not able to delete the property
> > 
> > So it seems like 264594 at main fixed a bug and aligned our behavior with
> > Chrome & Firefox. Unfortunately, it also seems that the custom-elements
> > polyfill was relying on said bug.
> > 
> > At this point, I think the best we could do is detect the policy and maybe
> > quirk our behavior. I think this would also mean detecting the polyfill
> > version so that we can have our new/standard behavior when using the fixed
> > version of the polyfill.
> 
> Also, for the record, I believe that the part of 264594 at main that broke the
> polyfill was dropping this logic in the generated bindings:
> https://github.com/WebKit/WebKit/commit/
> 12544d4187162da11dad0451bceece73a83c4837#diff-
> 635979f17a2f56f0ef3529de78480ad862a207ce15c096072b4e025a1e6b7428L1484
> 
> We used to mark indexed properties as configurable, but it didn't match
> Blink or Gecko.

Doing a quirk to restore pre-264594 at main when the custom-element polyfill version < 1.2.1  is used sounds tricky. We usually do quirks based on site domains. Here the quirk would change the behavior of our bindings based on whether of not a script element with a particular URL is on the page. I'll see what I can do.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20230620/4445c11e/attachment-0001.htm>