2026-10-01
[01:50:03.0608] <Luke Warlow>
https://github.com/whatwg/html/issues/11136 - this is the issue fwiw

[01:53:25.0437] <annevk>
How does that relate to the formAssociatedCallback? Your example doesn't even have it.

[03:43:30.0596] <smaug>
anyone has the magic skills to trigger pr diff creation https://github.com/whatwg/html/pull/10995#issuecomment-5924635469 ?

[04:17:06.0917] <annevk>
smaug: I've made an attempt. If you edit OP (which I think you can do too now) by adding a space or some such the bot should come looking again.

[04:17:20.0931] <smaug>
thanks

[04:50:20.0937] <Luke Warlow>
> <@annevk:matrix.org> How does that relate to the formAssociatedCallback? Your example doesn't even have it.

I need to look into that a bit further. But it's the same underlying issue I believe. The conversation in the issue was relevant even if the title maybe isn't quite

[04:52:11.0931] <annevk>
Interesting. Might well be so.


2026-10-02
[05:49:45.0594] <Ms2ger>
annevk: my head hurts as usual when I think about questions like this, so I'll pass it to you: if I queue a task, do I have a "current realm" when the steps run?

[05:50:15.0480] <Ms2ger>
(Or anyone else who knows)

[05:50:20.0011] <Shannon Booth>
you do not!

[05:51:18.0481] <Shannon Booth>
(its kind of not really defined in the spec, and an execution context needs to be pushed as a workaround)

[05:51:47.0190] <Ms2ger>
Okay, I was right to be paranoid, then :)

[05:51:54.0963] <Shannon Booth>
i think every implementation must have their own solution for it

[05:51:55.0505] <Ms2ger>
(Is it paranoia if they're really out to get you?)

[06:28:09.0985] <annevk>
Maybe there should be an issue? Not entirely sure when this comes up. Usually I put the global in a variable before going in parallel and reuse that global for the task. With queue a global task you could conceivably have a current global. (Remember, don't use "queue a task".)

[08:04:37.0048] <Shannon Booth>
it is quite common since it also impacts cases a promise is reacted within a task as there is nothing on incumbent stack for the job callback, so prepare to run a callback may also be needed. I’ve sort of lumped this under https://github.com/whatwg/html/issues/4342 and https://github.com/whatwg/webidl/issues/135#issuecomment-772716243 when looking at it before, but maybe is it different

[08:05:16.0578] <annevk>
That first one looks on point for sure.


2026-10-03
[02:22:54.0361] <annevk>
Anyone around for a cleanup PR, maybe sideshowbarker? https://github.com/whatwg/html/pull/13029

[05:21:22.0322] <annevk>
emilio: so I filed https://bugzilla.mozilla.org/show_bug.cgi?id=2077692 and wanted to fix the title, but I'm getting 429 back. I don't think I made all that many requests though.

[05:23:29.0462] <annevk>
Ooh, maybe it's Private Relay. Oh well.


2026-10-04
[05:52:21.0059] <smaug>
fun, element-reflecting attributes are all over the place. Implementations not following the spec and spec being likely buggy (but https://github.com/whatwg/html/pull/8932 fixing that)

[05:53:11.0168] <smaug>
(and implementations disagree with each others)

[05:54:45.0917] <annevk>
I had forgotten about that PR. I have too many PRs outstanding and too many too review.

[05:54:54.0991] <annevk>
* I had forgotten about that PR. I have too many PRs outstanding and too many to review.

[05:56:00.0793] <smaug>
webkit based implementations could finally fix their treescope thingie, which defaults somehow to document as root when dealing with detached elements

[06:17:08.0488] <smaug>
(which I think leads to blink following that spec PR, but not with detached elements, and webkit following the current spec, but not with detached elements)


2026-10-05
[01:41:09.0411] <annevk>
https://infra.spec.whatwg.org/#identity-and-equality is now a thing, please go use it and let Noam Rosenthal and I know if you run into any issues. Adoption at WHATWG starts with https://github.com/whatwg/infra/pull/710#issuecomment-5991033332 If you know other places please file an issue or let someone know. Thanks!

[01:41:53.0199] <Noam Rosenthal>
TabAtkins: I think a few CSS places might be able to benefit from this?

[01:42:28.0487] <Noam Rosenthal>
(I'm preparing a PR for connections allowlist but might wait until xrefs are updated first)

[02:36:26.0120] <Ms2ger>
Mm, is there any reason to use "is" with lists or maps

[02:36:32.0505] <Ms2ger>
* Mm, is there any reason to use "is" with lists or maps?

[03:25:21.0467] <annevk>
Well, I think you want to have identity comparison available for them, but I suspect it won't be used much.

[03:39:13.0520] <Noam Rosenthal>
yea I think erring on the side of being explicit is better even if we have some branches that are less useful

[15:55:33.0242] <Johnny Stenback>
Hey folks. A skeleton of a schedule for WHATUP at TPAC is now up at the top of https://github.com/whatwg/meta/issues/398. It's mostly empty at this point. There's a few requests in that issue for more slots, which I will incorporate as well. Please comment in that issue with further topics!

[15:56:19.0540] <Johnny Stenback>
You can also reach out to me either here, or at jstenback@google.com if you have coordination requests for cross WG sessions or what not.


2026-10-06
[03:15:15.0237] <emilio>
annevk: I'm curious about https://github.com/whatwg/html/pull/13040#discussion_r4185780341, what kind of styles would turn you into a listbox? IIRC it's just purely DOM, right?

[04:55:23.0054] <zcorpan>
Psychpsyo: The ipr check is red for w3c/editing

[05:30:25.0043] <Psychpsyo>
Thanks for the notice. Since my current PRs seem to be merged now, I'll look into this when I next file one.

[07:54:42.0608] <annevk>
emilio: we have a couple cases where base vs native can do that in Safari 

[07:55:07.0161] <emilio>
annevk: huh, why? That seems... unexpected / unintended?

[07:58:53.0715] <annevk>
emilio: there's no listbox on iOS for instance. It's been that way since 2007.

[07:59:06.0168] <emilio>
Huh

[07:59:22.0790] <emilio>
In Gecko on android we render a listbox, but when you click it shows up a popup

[08:01:03.0845] <annevk>
I think Chrome on Android does the same, but maybe jarhar changed it. I was actually kinda curious about that to see if there is an opportunity for us to do some alignment. Although I doubt we could align it across apps and the web.

[08:21:33.0872] <Luke Warlow>
Chrome on Android historically always rendered as drop downs, but a change rolled out to where they actually have listboxes and they work in page and don't show a picker anymore.

[08:21:52.0349] <Luke Warlow>
* Chrome on Android historically always rendered as drop downs, but a change rolled out to where they actually have listboxes and they work in page and don't show a picker anymore. This is the same across appearances.

[08:57:40.0100] <jarhar>
yep luke is right

[08:58:04.0688] <jarhar>
i just gave android the exact same listboxes that desktop has

[08:59:20.0898] <jarhar>
not sure why they werent on android in the first place

[09:30:15.0415] <eemeli>
Is there any particular reason why `lang` counts as a [translatable attribute](https://html.spec.whatwg.org/multipage/dom.html#translatable-attributes), but `dir` does not?

[09:36:52.0598] <mgaudet>
👋 Now that [Thenable Curtailment](https://github.com/tc39/proposal-thenable-curtailment) is at Stage 3, I am hoping to push forward on the WebIDL changes. I've put up a [PR on the WebIDL repo](https://github.com/whatwg/webidl/pull/1645) but I need to get some read on Chromium / WebKit folk, and I'm not sure who to ping about this (or the expected procedure here really... ) 


2026-10-07
[17:52:24.0850] <dbaron>
Sounds reasonable to me, with the caveat that translation sometimes requires adding new bidirectionality embeddings that weren't present for the original language.  For example, if you're translating to an RTL language you likely want to wrap a phone number in `<span dir=ltr>`, I think.

[17:52:41.0517] <dbaron>
I haven't previously been involved with that part of the spec, though.

[23:09:36.0245] <annevk>
jarhar: since when has that shipped? It didn't break any websites?

[23:11:16.0676] <annevk>
Yeah seems like a bug. And I guess it might indeed warrant warning of sorts that just changing direction can be insufficient. File an issue/PR?

[23:12:34.0738] <annevk>
Could try Sam Weinig for WebKit (`@weinig`). Maybe it's me. Noam Rosenthal might have advice for Chromium.

[00:27:00.0689] <Noam Rosenthal>
> <@annevk:matrix.org> Could try Sam Weinig for WebKit (`@weinig`). Maybe it's me. Noam Rosenthal might have advice for Chromium.

Happy to ping the right folks. Its not clear from the PR what the web observable implication is if ant? Tests  will help

[03:15:53.0527] <annevk>
zcorpan: Noam Rosenthal addressed the last round of feedback for that HTML parser PR and TT PR. I'll leave it to you to do the final check.

[03:42:17.0128] <zcorpan>
annevk: r+ but I don't have permission to merge for trusted-types

[03:46:02.0092] <Noam Rosenthal>
I can merge. Thanks!

[03:51:14.0642] <Noam Rosenthal>
Merged TT. Can you merge HTML plz? (I don't have merge rights for that...)

[04:36:45.0730] <Luke Warlow>
Chrome 142 (according to chrome status) so shipped to stable end of October last year.

[07:57:08.0619] <annevk>
TabAtkins: I have another case of Spec Generator and WHATWG's build pipeline disagreeing: https://github.com/whatwg/fs/pull/189 

[07:57:59.0166] <annevk>
Some things will happen a microtask later.

[09:58:06.0944] <Noam Rosenthal>
OK I pinged around

[12:14:19.0579] <jarhar>
yes, luke is right. there was one webview app that filed a bug but after i explained to them that they can get the dropdown back by adding size=1 to their select multiple they just changed their app and everything was fine


2026-10-08
[07:29:12.0766] <annevk>
Tim van der Lippe Shannon Booth Luke Warlow: https://github.com/whatwg/dom/pull/1507 impacts Servo and Ladybird afaik, and WebKit’s moveBefore(). Would appreciate your review.

[08:27:46.0001] <annevk>
Thanks!

[10:02:08.0996] <mgaudet>
Yes, sorry -- basically, if you _could_ run user code by resolving a promise, the promise is latched, but then the actual resolution (where the user code is run) is deferred to another microtask. 

The end goal here is to kill the litany of security bugs caused by creating `then` on `Object.prototype` 

[10:02:39.0214] <mgaudet>
One thing that's not clear to me is how to deal with tests in this transitional state; I can prepare a WPT PR I guess that would just be a draft? 

[10:14:17.0430] <Kurt Catti-Schmidt>
I would like reconsider whether to use <link type=module> for https://github.com/whatwg/html/pull/12860 at next week's sync. Last time it was discussed, this was rejected because type is already used on <link> for MIME type enforcement. 

I updated the PR to instead use a new attribute as the trigger (<link rel=stylesheet import=foo>), which would also allow for fallback on unsupported browsers via href, which I think is pretty nice. It also semantically indicates that it's a declarative version of a Javascript import.

However, there have been some comments in support of <link type=module> since I made the move away from it: https://github.com/whatwg/html/pull/12860#issuecomment-6064803258

We might want to use a new attribute instead of href anyways (since it behaves differently). If we do, type=module becomes redundant.

Curious what everyone's thoughts are - this was discussed briefly a few weeks ago, but members wanted some more time to think about introducing a new attribute.


[10:16:58.0051] <Kurt Catti-Schmidt>
* I would like reconsider whether to use \<link type=module> for https://github.com/whatwg/html/pull/12860 at next week's sync. Last time it was discussed, annevk was against this because type is already used on \<link> for MIME type enforcement. 

I updated the PR to instead use a new attribute as the trigger (\<link rel=stylesheet import=foo>), which would also allow for fallback on unsupported browsers via href, which I think is pretty nice. It also semantically indicates that it's a declarative version of a Javascript import.

However, there have been some comments in support of \<link type=module> since I made the move away from it: https://github.com/whatwg/html/pull/12860#issuecomment-6064803258

The argument is essentially that type=module brings in more in sync with <script type=module>, and <script> has a carve-out for this, but <script> only allows "text/javascript" MIME types, while <link> supports many MIME types, so it's a higher bar for a carve-out.

We might want to use a new attribute instead of href anyways (since it behaves differently). If we do, type=module becomes redundant.

Curious what everyone's thoughts are - this was discussed briefly a few weeks ago, but members wanted some more time to think about introducing a new attribute.


[10:45:19.0474] <annevk>
I think going along with `<script>` is probably okay. However, is this equivalent to script in that it ends up fetching a whole graph or does it still only fetch the top-level resource as CSS module scripts do? Because if we go for parity we should have parity I think.

[12:52:09.0820] <Noam Rosenthal>
It might fetch a whole graph if we end up supporting fetching from a module graph in `@import`. Currently module scripts don't support `@import`. But it does *populate* the graph like <script>

[12:52:23.0229] <Noam Rosenthal>
* It might fetch a whole graph if we end up supporting fetching from a module graph in `@import`. Currently module scripts don't support `@import`. But it does _populate_ the graph (with itself) like \<script>

[13:05:16.0024] <Noam Rosenthal>
But I don't see why this detail matters? It's a property of @import in css and not of the element that import the top level style 


2026-10-09
[22:53:18.0790] <annevk>
I'm pretty sure that when we did CSS imports we decided we would solve this before adding declarative markup.

[22:54:10.0603] <annevk>
Also from a CSP and security perspective it would be very weird if it suddenly started fetching more things.