WHATWG
2026-09-25
plaintext logs
prev next
source on github
05:15
<annevk>
What's the plan for the tests?
05:54
<Tim van der Lippe>
I ran these tests in Firefox and Chrome and took the values that they reported.
06:06
<annevk>

GitHub has an interesting new feature when blocking users, but unfortunately it doesn't work for security issues:

[ ] Close all open issues, pull requests, and discussions opened by this user

And that's really the only time I get to this checkbox as otherwise I don't need to go all the way to the settings panel of the org.

12:09
<annevk>
Tim van der Lippe: you should ask for commit access in #wpt:matrix.org
12:38
<myf>
Ah, true, pretty unbelievable. I am getting ActiveText in Chromium in Dark mode only, and it has contrast that fails tripple-A WCAG (F00 : 121212).
13:12
<Luke Warlow>
Yeah so you're on Windows which is broken in its own right, then the dark mode ActiveText is also broken (I fixed it for LinkText and VisitedText and must have just never got round to ActiveText). I have a CL that fixes it so will push it for review once I've fixed up test expectations. Do you have a link to that test page?
14:10
<Tim van der Lippe>
Ah wasn't sure whether external contributors were allowed to be added. It seems like I was added already, thanks!
20:05
<myf>

Sure, test samplers for dark | light: https://codepen.io/myf/full/KKbVpmN | https://codepen.io/myf/full/jOXWPGV .
Interesting to watch over time; from recent progress, very glad to see that Firefox abandoned semi-transparent highlight. But the fact that some colours are same in both schemes (mark, highlight, accent) feels wrong. I wonder if there could be some good, accessible and aesthetic palette developers could actually build derived palettes fitting user's preferences, and not these odd (at best) or even broken values we cannot push very far without hitting some ugly/inaccessible shades.

(Thanks for fixing stuff, much appreciated! And it is true that primarily, activeText is not displayed long enough to pose accessibility risk, I guess, so NBD with non-AAA contrast.)