04:49
<MikeSmith>
annevk: about the wattsi issue, today I have only been able to make enough time so far to look only briefly. I will have more time tonight
04:55
<annevk>
MikeSmith: thanks for taking a look
07:43
<annevk>
zcorpan_: you around? Looking at your CSP + COOP tests and it might help to go through them together
09:31
<MikeSmith>
annevk: about the wattsi issue, can you tell me explicitly what you want the behavior to be?
09:32
<MikeSmith>
currently it halts and emits an error if there’s an *empty* span in a heading
09:32
<MikeSmith>
that is, <h5 id="rel-alternate"><span id="link-type-alternate"></span>
09:33
<MikeSmith>
but it does not emit an error if there’s a non-empty span
09:33
<MikeSmith>
<h5 id="rel-alternate"><span id="link-type-alternate">foo</span>
09:33
<MikeSmith>
so you want it to be the other way around?
09:35
<annevk>
MikeSmith: that's surprising since we already have empty spans in headings I believe
09:36
<annevk>
MikeSmith: e.g., <h5 id="image-map-processing-model"><span id="processing-model"></span>Processing model</h5>
09:38
<MikeSmith>
hmm OK
09:38
<MikeSmith>
so that is in the source but not causing an error?
09:39
<annevk>
MikeSmith: that's in there right now, yes
09:39
<MikeSmith>
OK
09:39
<annevk>
MikeSmith: which is why it was surprising that adding it elsewhere didn't work
09:39
<MikeSmith>
yeah
09:40
<MikeSmith>
so I will step through the code execution and see what’s happening differently for those two cases
10:05
<annevk>
<3
11:07
<annevk>
Python 3 really makes you appreciate "don't break the web"
12:17
<zcorpan_>
annevk: I'm here now
12:18
<zcorpan_>
happy to go through them
12:45
<zcorpan_>
annevk: reading your comment https://github.com/web-platform-tests/wpt/pull/21111#issuecomment-681730038 again, I think it finally clicked for me. I misremembered how Window/WindowProxy works on navigation. `popup` is the old window which still has the CSP sandbox applied
12:50
<annevk>
zcorpan_: yup
12:50
<annevk>
zcorpan_: and the subtlety with the whole thing is that the CSP is applied to the document/window and not the browsing context
12:51
<zcorpan_>
annevk: yeah. I'll fix the test and try to explain this with comments
15:44
<Domenic>
annevk: was "do not merge yet" on https://github.com/whatwg/html/pull/5868 just about tests, or is there more to resolve?
15:45
<annevk>
Domenic: I was hoping to have at least a response from the CSS WG to the concerns raised by Richard
15:45
<Domenic>
Makes sense
15:45
<annevk>
well, not the whole CSS WG, but someone who can answer
15:46
<Domenic>
As I said, I'd like to not hold this up indefinitely (like what happened with list marker styles, which are still out of sync between HTML and CSS), but we can give it a few days.
15:46
<annevk>
Yeah, that makes sense, but it's only a couple hours old at this point
16:04
<astearns>
annevk I haven't looked very close at Richard's comment, but if Emilio's response is enough that's good enough for me