| Summary: | ImageDocument should support dark mode | ||||||
|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Pascal Abresch <nep-webkit> | ||||
| Component: | Images | Assignee: | Nobody <webkit-unassigned> | ||||
| Status: | NEW --- | ||||||
| Severity: | Normal | CC: | bugs-noreply, cdumez, changseok, darin, esprehn+autocc, ews-watchlist, gyuyoung.kim, nekohayo, sabouhallawa, webkit-bug-importer | ||||
| Priority: | P2 | Keywords: | InRadar | ||||
| Version: | Other | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| Attachments: |
|
||||||
|
Description
Pascal Abresch
2022-07-26 01:32:17 PDT
Created attachment 461339 [details]
Patch
Comment on attachment 461339 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=461339&action=review Thanks, I think this is a good idea. > LayoutTests/fast/loader/image-document-dark-mode.html:5 > +<iframe src="resources/image@test.png"> Can we instead construct a test that uses a dark mode document with an image element in it in the expected.html? The reason I suggest this is that an expected.html test is even more useful than a mismatch.html for catching future mistakes. And also, a test structured in that way could probably be turned on everywhere; it would automatically do the right thing on platforms with no dark mode support. The only tricky thing about such a test might be matching the layout that you’d get with an iframe that directly references an image, but I think that’s something likely to stay stable for a long time. I wish we had done that for bug 242950 too. I wouldn’t block landing this for that test change, but I think we should improve these tests. I can work on improving the test next week. |