Bug 12713
| Summary: | Import Anne van Kestern's offset test suite | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | David Kilzer (:ddkilzer) <ddkilzer> |
| Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | Normal | CC: | abob, ahmad.saleem792, annevk, ap, bfulgham, koivisto, rniwa, sam |
| Priority: | P2 | ||
| Version: | 420+ | ||
| Hardware: | Mac | ||
| OS: | OS X 10.4 | ||
| URL: | http://dump.testsuite.org/2006/dom/style/offset/ | ||
David Kilzer (:ddkilzer)
See Bug 12687 Comment #9.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
David Kilzer (:ddkilzer)
Implied with this activity is to file bugs on all tests where WebKit ToT is failing. :)
Sam Weinig
As long as we are importing the offset tests from Anne's suite, we might as well also grab some of the other related directories. For instance http://dump.testsuite.org/2006/dom/style/offset-more/ , http://dump.testsuite.org/2006/dom/style/client/ , etc. We should also check to with Anne about this.
Anne van Kesteren
I'm perfectly happy with you guys importing these tests (and others from testsuite.org). Go ahead!
Antti Koivisto
While these tests are useful, we probably should not try to actually pass them at this point. It is more important to have consistent behavior with other browsers (mostly firefox). Offset properites are not specified anywhere authoritatively as far as I know so passing and failing is pretty subjective.
Ahmad Saleem
I tested following test suite in Safari 15.5 on macOS 12.4 and focused on only failures within Safari (not where other browsers might have different behavior on test cases):
From the failures, these are following where Safari behavior was different from other browsers (Chrome Canary 105 or Firefox Nightly 104):
NOTE - I didn't have Ahem font installed. (So this might be pointless testing but still wanted to share the results)
1) Below test case -> offsetTop is incorrect in Safari and Chrome but correct in Firefox
https://dump.testsuite.org/2006/dom/style/offset/048.htm
2) Below test case -> offsetTop is incorrect in Safari and Chrome but correct in Firefox (all browser fail on offsetBottom)
https://dump.testsuite.org/2006/dom/style/offset/049.htm
3) Below test cases -> after dialog boxes, all browsers rendered the output different
https://dump.testsuite.org/2006/dom/style/offset/056.htm
Anne van Kesteren
Let's close this and rely on WPT instead. I'm pretty sure the tests that are still relevant have been migrated there long ago.