Bug 251226

Summary: Regression(STP162): 268 editing/other tests regressed and failing now
Product: WebKit Reporter: Ahmad Saleem <ahmad.saleem792>
Component: HTML EditingAssignee: Nobody <webkit-unassigned>
Status: RESOLVED CONFIGURATION CHANGED    
Severity: Normal CC: karlcow, rniwa, webkit-bug-importer, wenson_hsieh
Priority: P2 Keywords: InRadar, WPTImpact
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Regression Screenshots - WPT Editing (STP162 vs STP161) none

Description Ahmad Saleem 2023-01-26 14:12:44 PST
Created attachment 464675 [details]
Regression Screenshots - WPT Editing (STP162 vs STP161)

Hi Team,

Another regression in STP162, in HTML Editing:

https://wpt.fyi/results/editing/other?diff&filter=ADC&run_id=5183180772999168&run_id=5177001724346368

Just wanted to raise it and will try to see which commit might have caused this.

Thanks!
Comment 1 Radar WebKit Bug Importer 2023-01-26 16:05:04 PST
<rdar://problem/104718120>
Comment 3 Ryosuke Niwa 2023-02-01 16:02:43 PST
These tests pass when I run locally.
Comment 4 Ahmad Saleem 2023-02-08 16:44:36 PST
http://wpt.live/editing/other/keeping-attributes-at-joining-elements.tentative.html?method=backspace - WebKit ToT is failing 32 tests

http://wpt.live/editing/other/keeping-attributes-at-joining-elements.tentative.html?method=forwarddelete - WebKit ToT is failing 32 tests

typing-around-link-element-at-non-collapsed-selection.tentative.html?target=ContentEditable - WebKit ToT is failing 31 tests

__

^ I am going to use these three tests as sample and revert some contenteditable changes in STP161 to STP162 ranges, I have slight guess, which might have caused these failures.

@rniwa - any suggestions or input?
Comment 5 Ahmad Saleem 2023-03-09 04:10:05 PST
WebKit ToT (261417@main)

Reverted - https://github.com/WebKit/WebKit/commit/f61cf23abe063f69ce6408c47a6454825ce011e8#diff-8448f5db214cf11ac521a0d8d7f3fc5dde40e9afc0d6f9a063ba5cff93d9be91

^ Still fails - http://wpt.live/editing/other/keeping-attributes-at-joining-elements.tentative.html?method=backspace

^^ Funny enough - this now fails on Safari 16.3 as well. :-(

Reverted - https://github.com/WebKit/WebKit/commit/97d658482be252a38bb1128072ec6a3372928605

^ Still fails - https://github.com/WebKit/WebKit/commit/97d658482be252a38bb1128072ec6a3372928605

________

These were two potential candidate for this regression but now since even Safari 16.3 is failing at least when running directly. It seems that it might be something on WPT end that we started failing.

____

In Safari 16.3, we get following error:

promise_test: Unhandled rejection with value: object "Error: unimplemented"
promiseReactionJob@[native code]

and WebKit ToT:

promise_test: Unhandled rejection with value: object "Error: unimplemented"
get_stack@http://wpt.live/resources/testharness.js:4534:30
AssertionError@http://wpt.live/resources/testharness.js:4527:31
assert@http://wpt.live/resources/testharness.js:4511:37
@http://wpt.live/resources/testharness.js:759:35
@http://wpt.live/resources/testharness.js:2595:30
@http://wpt.live/resources/testharness.js:2642:40
Asserts run
No asserts ran

_______

@Karl & @rniwa - Any help or guidance on what I can do to get this regression fixed on my side?
Comment 6 Ahmad Saleem 2023-04-20 16:55:07 PDT
It seems to have progressed / fixed with STP168.

Marking this as "RESOLVED CONFIGURATION CHANGED".
Comment 7 Ryosuke Niwa 2023-04-20 17:01:06 PDT
(In reply to Ahmad Saleem from comment #6)
> It seems to have progressed / fixed with STP168.
> 
> Marking this as "RESOLVED CONFIGURATION CHANGED".

Nice! Although it's a bit unsettling that we didn't figure out what caused it. Maybe web driver changes??
Comment 8 Ahmad Saleem 2023-04-20 17:15:42 PDT
(In reply to Ryosuke Niwa from comment #7)
> (In reply to Ahmad Saleem from comment #6)
> > It seems to have progressed / fixed with STP168.
> > 
> > Marking this as "RESOLVED CONFIGURATION CHANGED".
> 
> Nice! Although it's a bit unsettling that we didn't figure out what caused
> it. Maybe web driver changes??

Yeah! Because the tests didn't change from 2021 and they were failing till STP167, so it might be some web driver changes or some other underlying HTML Editing changes in STP168 but I think, it is more to do with Web Driver and something on WPT end because these tests were failing in Safari 16.3 as well (passing but then suddenly start failing), which didn't had any of HTML editing fixes.
Comment 9 Ahmad Saleem 2023-05-04 00:29:48 PDT
(In reply to Ahmad Saleem from comment #6)
> It seems to have progressed / fixed with STP168.
> 
> Marking this as "RESOLVED CONFIGURATION CHANGED".

It seems to now regressed with STP169 release. :-(

https://wpt.fyi/results/editing/other?diff=&filter=ADC&run_id=5082264357044224&run_id=5198643877904384
Comment 10 Ahmad Saleem 2023-07-26 02:44:12 PDT
I am closing this as 'RESOLVED CONFIGURATION CHANGED' again because it seems it is flaky on WPT side and some runs have it as passed and some as failed.

https://wpt.fyi/results/editing/other?diff=&filter=ADC&run_id=5183180772999168&run_id=5177001724346368&run_id=5082951824441344

^ Not good link but shows STP174 passing these tests.