| 06:42 | <annevk> | keithamus: if you lock a conversation and give the rationale "spam", does that mean the entire conversation is flagged as spam? That's kinda what the UI implies to me. Whereas I mainly want to lock certain PRs because they attract spam. |
| 06:45 | <keithamus> | keithamus: if you lock a conversation and give the rationale "spam", does that mean the entire conversation is flagged as spam? That's kinda what the UI implies to me. Whereas I mainly want to lock certain PRs because they attract spam. |
| 07:16 | <annevk> | Domenic: Initially I wanted to keep the "A number" wording as-is, but I'll change it to "Each a number". I thought one should be able to insert and remove dt elements without impacting the dd, and vice versa, but the "Authors" example in the specification clearly contradicts that view. As in, we pluralize the term when there's multiple entries for it. |
| 09:54 | <Jake Archibald> | smaug: huh, I don't recognise that part of the spec. Domenic was this introduced as part of close watcher? |
| 09:55 | <smaug> | aha, it could have been around that time, not when navigable stuff was added |
| 09:55 | <smaug> | I should have looked at the blame |
| 09:56 | <smaug> | I wish there was a variant of the spec where it showed when some change was added (maybe when hovering over some marker). Looking at https://searchfox.org/whatwg-html/source/ works too, but since it is not what I read normally, I just don't check it too often. |
| 10:41 | <Domenic> | Yes, it follows "consume user activation". There is a note explaining the asymmetry between consumption and notification below both algorithms. |
| 10:41 | <Domenic> | "consume user activation" has been that way since it was first introduced many years ago. |
| 11:12 | <zcorpan> | Psychpsyo: are you still interested in https://issues.chromium.org/u/1/issues/367992694 ? :) |
| 11:16 | <Psychpsyo> | Interested yes, interested in actually working on it in Chrome, not really, no. Developing for Chrome is very slow on my Machine and I've gotten involved with Ladybird a lot more. I hadn't said anything on the Gerrit since that site has somehow started using my other google account and does not let me in anymore. TL;DR: No, Chromium-dev is too tedious |
| 11:17 | <zcorpan> | Psychpsyo: ok thx, I unset the assignee |
| 11:17 | <Psychpsyo> | Thank you |
| 11:18 | <zcorpan> | jarhar: fyi ^ |
| 11:19 | <Psychpsyo> | Here's the WIP commit that I had written up: https://chromium-review.googlesource.com/c/chromium/src/+/5866772 |
| 11:27 | <zcorpan> | Aha... Gecko has a mechanism to use count unsupported properties (but lacks httparchive integration) https://searchfox.org/mozilla-central/source/dom/base/UseCounters.conf#184-186 |
| 11:28 | <Psychpsyo> | Oh neat! I remember that a tricky thing in Chrome was adding a use-counter for something that didn't exist. (making no-op getters/setters would still have been observable from JS) |
| 14:10 | <acr> | Sorry for this off-topic question, but is there any Matrix channel were one could "informally" ask about w3c specs? (to put an example, about Battery API, or WebNN) |
| 14:10 | <acr> | I didn't find any, but suspect there must be some, and I just don't know the right keyword. |
| 14:16 | <annevk> | acr W3C is mainly on Slack: https://www.w3.org/wiki/Slack (and IRC) |
| 14:19 | <acr> | Thank you, I may reinstall my IRC client. |
| 15:38 | <jarhar> | toggle events are not composed. in the case that the source is from a different tree, i used the retargeting algorithm. what impact does that have on this? |
| 16:06 | <AtkinsSJ> | annevk: I've been chasing some WPT regressions with the scoped custom element registries, and trying to figure out if they're our bugs or spec ones. Just wanted to check: Is the "active custom element constructor map" supposed to ever get null registries in it? Because I'm finding it does sometimes and that was causing crashes. |
| 16:15 | <AtkinsSJ> | Never mind, I'll post an issue. |
| 16:20 | <annevk> | AtkinsSJ: interesting, I don't know for sure. WebKit doesn't have regressions... |
| 16:22 | <annevk> | But this is why it's great to have a second implementation! |
| 16:24 | <AtkinsSJ> | I submitted an issue in the html repo. I've been staring at all this trying to make sense of things for a couple of days now so my brain is mush. 😅 I don't want to start filing spec bugs until I'm vaguely sure I'm not just making silly mistakes |
| 16:31 | <annevk> | AtkinsSJ: that's very nice of you, but a little sooner is definitely okay. |
| 16:31 | <annevk> | I think my personal threshold is a couple of hours at best, unless I'm fairly positive it's my fault. |
| 16:54 | <AtkinsSJ> | Thanks! |
| 17:04 | <smaug> | the whole internal resource link stuff, which is only used by rel=expect in underspecified , as far as I see. Or perhaps its use in rel=expect. |
| 17:07 | <smaug> | pushState seems to break it |
| 18:34 | <annevk> | Noam Rosenthal: ^^ |
| 19:58 | <smaug> | Hmm, what might https://html.spec.whatwg.org/#dom-navigation-reload step 2 mean. 'Exists' is an operation related to ordered map, but options is a dictionary |
| 19:58 | <smaug> | Domenic: ^ |
| 20:32 | <jmdyck> | step 3 you mean? |
| 20:53 | <smaug> | er, step 3 yes |
| 22:34 | <akaster> | if a fetch request's client is null, should its window become "no-window" during fetching? (re: https://github.com/whatwg/html/pull/11250) |
| 22:36 | <akaster> | i.e. should 11250 update create_navigation_params_by_fetching to set request's window to no-window (not doing so crashes fwiw) |
| 22:42 | <akaster> | bleh. https://w3c.github.io/webappsec-mixed-content/#upgrade-algorithm step 1.3 doesn't consider whether request's client is null |
| 22:43 | <akaster> | presumably a browser-ui initiated navigation doesn't care about mixed security contexts and should just keep going with whatever the user set |