2021-05-27 [01:20:02.0872] sideshowbarker: I changed "who can read history" to "Anyone", but I'm not sure that's correct [01:20:24.0387] Or, rather, not sure it's sufficient [01:20:55.0296] Ah, but https://view.matrix.org/room/%21AGetWbsMpFPdSgUrbs:matrix.org/ seems to work now [01:24:17.0877] jgraham: yup — thanks much [01:49:38.0818] jgraham: can I get admin to fixup the logo? [01:52:55.0290] Ah yeah, I eventually worked out that one doesn't do that from "settings" [02:01:20.0175] annevk: https://gist.github.com/andreubotella/bb62bfe0489baa5fe9a1155e0b37f92d [02:01:27.0073] I'm not sure if this is any good [02:02:20.0213] Andreu Botella (he/they): cool, "This is how" seems to end abruptly [02:02:33.0692] yeah, I just noticed that 😅 [02:04:00.0606] "Firefox normalizes to CRLF as to the other browsers." as do* [02:05:21.0517] fixed [02:06:54.0236] Andreu Botella (he/they): so tl;dr: this looks amazing and it'd be great to publish this on the blog; I can get you an account [02:07:01.0397] (but I'll also continue reading it now) [02:07:46.0123] sure, that'd be great [02:09:31.0536] jgraham: my PR uses https://app.element.io/#/room/#whatwg:matrix.org as the link for this channel. Is there any reason to instead prefer https://app.element.io/#/room/!AGetWbsMpFPdSgUrbs:matrix.org ? [02:10:20.0723] I believe the latter is the internal room ID, which was meant to stay hidden to the user but isn't really in actual implementations [02:10:30.0723] OK [02:11:15.0483] Andreu Botella (he/they): I notice though that https://view.matrix.org/ uses those IDs for the log links [02:11:47.0948] yeah, I mentioned yesterday that view.matrix.org seems to be a thin wrapper over the client-server API, which I believe uses internal IDs [02:12:24.0169] I see [02:14:09.0351] "the current spec mandates that no normalization _happens_" – that should use the subjunctive [02:28:57.0654] meta nit: Do we want to keep the name of this room at _whatwg_ (lowercase)? or should we change it to _WHATWG_ (uppercase)? [02:30:57.0121] I don’t have a strong opinion about it either way, but if we’re gonna change it, it seems slightly better to change it now rather than later (for one thing, it affects how to we refer to it in https://whatwg.org/chat) [02:31:06.0940] * I don’t have a strong opinion about it either way, but if we’re gonna change it, it seems slightly better to change it now rather than later (for one thing, it affects how to we refer to it in https://whatwg.org/chat) [02:31:16.0585] I vote for WHATWG [02:36:59.0223] The address is still lowercase though. It seems you can also get uppercase addresses, but that seems a bit weird? [02:38:19.0143] It definitely looks more shouty in uppercase [02:41:00.0915] On Twitter we actually have the URL part in uppercase, but I'm not a big fan of uppercase letters in URLs personally [02:51:17.0946] So while creating the new blog account I also found out we had 24 comments on our latest post that had to be approved. Got excited for a moment, but they were all spam. [03:31:34.0898] I'm running into Wordpress bugs about escaping character references 😩 [03:49:27.0135] Oh that is annoying. I thought we had explicitly disabled any WordPress meddling for that reason [03:55:02.0442] I guess it depends on what other rooms you have in your list. In my Favorites list, it looks like that ⬆️ [03:55:24.0464] * I guess it depends on what other rooms you have in your list. In my Favorites list, it looks like that ⬆️ [03:56:16.0982] but I guess some other Mozilla channels, use lowercase — e.g., **developers** [06:44:09.0129] Is it expected that trying to force-reload whatwg.org produces a network error at the moment? [06:44:57.0813] From testing on irc: `13:43 < plh> html.spec.whatwg.org refused to connect. :(` [06:45:47.0250] sideshowbarker: ^ [06:46:12.0123] yeah, my fault [06:46:24.0709] fixing it now [07:44:54.0042] ok, I have no clue what the wordpress editor is doing with character entities [07:45:29.0747] `` outputs ` ` in the HTML, but `8x0D;` does the right thing [07:45:54.0552] anyway, this appears to work [07:46:18.0477] annevk: Should I publish it now? [07:49:01.0670] Andreu Botella (he/they): yeah, sounds good, if you got it to work [07:49:12.0772] I didn't know there was a new editor interface, looks kinda awful [07:49:41.0039] Any preference for the tweet? [07:51:17.0065] Let me come up with a shorter excerpt [07:54:38.0149] In form submissions, form values containing newlines get normalized to CRLF. But that newline normalization runs deeper than might be thought. Here is a deep dive on the recent spec changes on that area. [07:54:45.0383] Or something like that, even that's probably too long [07:59:14.0911] > During form submission, newlines get normalized to CRLF. That newline normalization is more complex than you might think. Here is a deep dive on the recent @htmlstandard changes in that area: [07:59:32.0291] Either way, it should fit, Twitter tells me [08:03:56.0095] I'm now off the post, at least until I find some misspelling or something, so feel free to aprove it [08:04:03.0759] * I'm now off the post, at least until I find some misspelling or something, so feel free to aprove it [08:06:18.0055] Andreu Botella (he/they): it's up at https://blog.whatwg.org/newline-normalizations-in-form-submission now, if it looks good I'll tweet (sorry, in a meeting) [08:06:26.0488] ok [08:08:05.0058] \o/ very nice work [08:09:03.0146] intended to replace freenode channel? [08:09:48.0013] aja: yeah, see https://github.com/whatwg/meta/issues/210 for context [08:09:59.0084] cool [08:18:11.0003] Now that I'm done with all of this, I think I have like over 30 tabs to close 😁 [08:19:17.0777] You can _close_ tabs‽ This changes _everything_ ;) [12:02:15.0340] /me waves [13:20:16.0078] hober: Hi! [13:30:10.0098] Howdy hober [13:39:34.0125] 👋 [13:40:31.0526] Greetings aja [13:44:46.0075] 👊 [13:46:23.0697] Can't believe it took me an hour to move channels. Didn't count on libera verification email ending up in Promotions folder. 2021-05-28 [18:49:08.0857] annevk: https://stackoverflow.com/questions/67730604/should-a-304-not-modified-response-to-a-cors-request-contain-cors-headers [18:50:51.0064] I see the non-normative note at https://fetch.spec.whatwg.org/#ref-for-concept-cors-check%E2%91%A0, which says: > As the CORS check is not to be applied to responses whose status is 304 or 407, or responses from a service worker for that matter, it is applied here. [18:52:53.0240] …but I can’t find which normative part of the spec has a requirement that requires the CORS check to not be performed for `304`s [20:04:38.0417] * …but I can’t find which normative part of the spec has a requirement that requires the CORS check to not be performed for `304`s [22:01:48.0270] Mornin' [22:04:47.0211] Hey hey Jake Archibald you've found the place where all the excitement is now [22:05:08.0230] We tried to hide from you, but you just too clever [22:05:38.0468] Haha [22:06:44.0477] I hadn't heard of [matrix] until just now. I feel so uncool [22:09:36.0043] [matrix] actually wasn’t cool until TC39 started using it and set up #tc39-general:matrix.org. That’s what single-handedly made it finally become cool [22:15:07.0790] Those trend-setters [22:29:26.0217] sideshowbarker: thanks, I replied [22:37:20.0312] annevk: ␆ thanks [06:42:35.0241] Here I am. This feels surprisingly polished, I already don't miss irccloud :) [06:43:16.0155] jgraham: are you using Matrix for W3C IRC as well somehow? teach me how? [06:44:35.0897] 👋 [06:56:05.0175] I see there's #w3c:matrix.org [07:00:30.0536] The W3C IRC is bridged with rooms of the name `#_w3c_#:matrix.org` [07:01:14.0483] heh, bad parsing. maybe try backticks around the channel addresses? [07:01:32.0384] `#_w3c_#testing:matrix.org` for example [07:01:52.0141] * The W3C IRC is bridged with rooms of the name `#_w3c_#:matrix.org` [07:02:01.0662] * `#_w3c_#testing:matrix.org` for example [07:02:09.0465] ah, is this because they're on another IRC network? [07:02:42.0588] Right [07:02:43.0063] https://github.com/matrix-org/matrix-appservice-irc/wiki/Bridged-IRC-networks [07:03:03.0072] gotcha. btw, the libera bridge is _almost_ finished 😀 [07:03:13.0388] See the matrix-appservice-irc usage guide to configure your IRC username and the like [07:03:17.0811] https://matrix-org.github.io/matrix-appservice-irc/latest/usage.html [07:03:25.0317] * See the matrix-appservice-irc usage guide to configure your IRC username and the like [07:28:57.0863] Hmm, so I've joined #_w3c_#testing:matrix.org but it's called "FDNPK0PMG5N0 (IRC) and 245 others" :) [07:29:08.0901] I guess I'll figure it out eventually [07:29:22.0612] Yeah, I'm not sure why that is [07:29:42.0628] Andreu Botella (he/they): are you seeing similarly unhelpful room names? [07:29:43.0988] Apparently that room doesn't have a name and matrix clients synthetize one from the room's users [07:30:03.0948] But I'm not sure why it wouldn't have a name [07:30:16.0447] /me opens the issue [07:30:34.0260] that happens sometimes, restarting the app/refreshing the window should fix that eventually. [07:30:38.0079] it's an element bug [07:31:07.0066] Usually leaving/rejoining or clearing cache in Help fixes that. [07:31:29.0548] right, exactly. nice to see you here, blue_penquin 😀 [07:32:28.0594] Hmm, restarting didn't help. I'll just leave it alone and figure it out another year. [08:49:02.0037] foolip: Yeah, I hadn't figured out #testing either. I joined #webdriver from here, but I think I'll still need to be on IRC to control the bots [08:49:20.0267] jgraham: I guess I'll stay on both too :) [08:50:25.0432] (if I understand correctly, one could issue raw irc commands to invite the bots, and after that normal usage probably works, but I'd at least have to learn what the raw command for `/invite` looks like ;) ) [08:52:18.0105] > <@foolip:matrix.org> Hmm, restarting didn't help. I'll just leave it alone and figure it out another year. settings -> help & about -> clear cache and reload is the surefire way to 'fix' the problem :) [08:52:22.0370] (until the actual bug gets resolved) 2021-05-29 [17:33:40.0415] Found an interoperability issue with flexbox vs img elements, I think. However, I can't tell whether the (accidentally useful?) behaviour in Firefox is correct, or whether what Safari/Chrome do is correct. Reduced test case at https://codepen.io/Krinkle/full/wvJrxQO [17:33:56.0019] In Firefox, the images shrink as needed to fit within the row. [17:34:51.0720] in Safari/Chrome, the image's box model becomes wider than natural and pushes out the third and fourth image [17:35:18.0264] I can workaround it by changing `width: 100%` to `width: calc(100% / 4)`, but seems like that shouldn't be needed. [05:39:42.0979] Timo Tijhof: What specs define the required behavior for that? Just the Flexbox spec? Or are there also requirements in HTML spec that are relevant? [05:39:54.0976] * Timo Tijhof: What specs define the required behavior for that? Just the Flexbox spec? Or are there also requirements in HTML spec that are relevant? [05:40:25.0837] hmm I guess just the Flexbox spec [05:40:42.0029] maybe this is underspecified [11:38:25.0629] For the record, the #whatwg channel at libera.chat is now bridged at #whatwg:libera.chat [11:38:36.0238] although the bridge is still buggy 2021-05-30 2021-05-31 [22:17:08.0624] As I mentioned elsewhere, I hope we can end up exclusively using Matrix and not need IRC alternatives, but maybe the community is big enough to be splintered? [22:20:18.0724] IRC could be bridged directly to this room, but as mentioned, appservice bridges tend to be flaky... [22:24:17.0008] So far we have had only one strong objection for moving to Matrix, and that objection argued that there’s nothing wrong with the current Freenode situation and that therefore we have no need to move *anywhere*. [22:25:02.0856] so given that, the Libera channel anyway essentially does nothing to help resolve that objection [22:25:31.0838] > there’s nothing wrong with the current Freenode situation the situation hasn't developed quite favorably for that argument... does the person still hold this view since freenode started maliciously taking over channels? [22:25:42.0757] yes [22:28:18.0715] yeah see https://github.com/whatwg/meta/issues/210#issuecomment-845242195 [22:29:41.0945] …although now I’m seeing that one of the objecting-to-matrix comments I saw there before has since been either removed or edited — so maybe there is some more of a chance for resolution than I had thought [22:31:00.0567] > we should stick with the current Freenode solution until/unless something bad happens but many bad things have happened, even since the comment was made. > I personally don't plan to move before then. that's perfectly okay, of course. Many people connect to both networks for now anyway... [22:31:48.0046] * > we should stick with the current Freenode solution until/unless something bad happens but many bad things have happened, even since the comment was made. > I personally don't plan to move before then. that's perfectly okay, of course. Many people connect to both networks for now anyway... [22:32:02.0167] ah no, I realize now the comment I had been thinking of was not in that issue but instead on the IRC channel: https://freenode.logbot.info/whatwg/20210526#c8144303 [22:32:47.0497] oh, I see. This is interesting. [22:32:59.0706] Well, I guess I just... respectfully disagree then. [22:33:35.0860] This is not _just_ political. It's about a provably malicious party taking control of the network. [22:34:08.0386] See for example the discussion around the change of policy regarding hate speech... [22:36:40.0646] I had not seen the change in policy about hate speech yet — but then, I quit paying attention a few days ago, and disconnected my IRC client from the #whatwg Freenode channel and all other Freenode channels [22:37:19.0901] I’m now re-connected to the #whatwg IRC channel from the Matrix bridge, but I’d honestly rather not need to be [22:37:54.0176] I decided to stay connected there for cases where somebody happens to show up and ask about something I can answer [22:38:16.0546] I personally share your views. In the TC39 Inclusion Group, we looked deep into the whole matter for a long time before concluding that IRC was pretty provably hostile for newcomers. [22:39:11.0063] matrix/element is not perfect but it's a major improvement IMO [22:39:27.0592] > <@usharma:igalia.com> I personally share your views. In the TC39 Inclusion Group, we looked deep into the whole matter for a long time before concluding that IRC was pretty provably hostile for newcomers. yeah I think that fact about IRC is widely agreed on by a lot of people at this point [22:40:14.0206] right. I don't hold anything against people using IRC for personal reasons, I do the same, but using it for community projects hurts inclusion. [22:41:43.0018] yeah the problems inherent with IRC are also the reasons why we’d be very unlikely to get agreement about bridging any IRC channel into this Matrix room [22:42:10.0207] as mentioned at https://github.com/whatwg/meta/issues/210#issuecomment-851012643 [22:43:06.0939] oh I see that actual precise term of art is “plumbing” an IRC channel into this room [22:43:46.0715] oh, because the IRC people wouldn't agree on a server either? [22:44:27.0181] I mean because I guess it would mean then we’d have traffic from the IRC coming directly to this room [22:44:58.0820] …and I am not sure we would have direct means to moderate that traffic [22:45:21.0336] yeah, the approach we've agreed upon at TC39 is to avoid bridging and just nudge folks who stumble across the IRC channel to Matrix [22:45:48.0959] I uploaded this yesterday btw 🙈 [22:46:29.0510] > <@sideshowbarker:mozilla.org> …and I am not sure we would have direct means to moderate that traffic exactly, especially if the hate speech policy thing is as serious as I think it is. [22:48:18.0690] > <@usharma:igalia.com> yeah, the approach we've agreed upon at TC39 is to avoid bridging and just nudge folks who stumble across the IRC channel to Matrix yeah but I guess you know the problem there is that if you change the IRC channel topic to say something like, *Avoid using this channel; we have a Matrix room at #freenode_#whatwg:matrix.org you can use instead*, then you’d run afoul of the “No advertising other networks” policy that Andrew retroactively created and applied (by seizing control of the hundreds of channels that he did). [22:49:07.0289] exactly, which is why we cannot just leave it in the topic and would have to do it manually... might end up making a bot for it 😅 [22:49:40.0292] but yeah, note that policy and it's clear how hostile the new management is 😕 😓 [22:49:48.0346] * I’m now re-connected to the #whatwg IRC channel from the Matrix bridge, but I’d honestly rather not need to be [22:49:49.0981] * but yeah, note that policy and tell me again that Andrew is not malicious 😕 😓 [22:50:10.0489] * but yeah, note that policy and it's clear how hostile the new management is 😕 😓 [22:55:38.0277] yeah, thanks for the heads-up — I hadn’t actually seen https://github.com/freenode/web-7.0/pull/513/files yet [22:56:25.0970] that’s awful — and seems to pretty clearly be at odds with the spirit of the WHATWG’s own documented code of conduct [22:56:35.0796] exactly [22:59:00.0559] yeah — anyway, I’ve been a daily user of IRC for 20 years, and on the #whatwg IRC channel for 15 or 16 years, so what’s happened with Freenode is a pretty big disappointment for me personally [23:01:08.0887] same 😕 [23:02:23.0654] * same 😕 [00:24:44.0786] So I’m considering how to resolve https://github.com/mdn/content/issues/5487 [00:25:14.0446] …which leads me to read https://developer.mozilla.org/en-US/docs/Web/API/Text [00:25:28.0225] > The Text interface represents the textual content of Element or Attr. [00:25:51.0338] > If an element has no markup within its content, it has a single child implementing Text that contains the element's text. [00:26:42.0337] …which leads me to wonder, *Who wrote this and how did they get it so wrong?* [00:26:57.0395] then I take a look at https://www.w3.org/TR/DOM-Level-3-Core/core.html#ID-1312295772 [00:27:17.0885] > The Text interface inherits from CharacterData and represents the textual content (termed character data in XML) of an Element or Attr. If there is no markup inside an element's content, the text is contained in a single object implementing the Text interface that is the only child of the element. If there is markup, it is parsed into the information items (elements, comments, etc.) and Text nodes that form the list of children of the element. [00:27:55.0855] so that MDN text is verbatim from the DOM3 spec… [00:29:11.0339] That explains how they got it so wrong [00:29:24.0982] heh [00:29:26.0417] Let's blame plh :) [00:29:59.0999] I hope plh didn’t actually write that part [00:30:31.0017] I mean, when did it ever make sense to talk about “markup” in the DOM? [00:31:52.0318] and was there ever some point where attributes actually had Text nodes in the DOM? (rather than just a string value) [00:32:21.0761] Oh, it's straight out of DOM 1 Core [00:32:42.0271] Attrs definitely had a Text child, I'm not even sure if I managed to remove that [00:32:52.0243] ah yeah, I see https://www.w3.org/TR/REC-DOM-Level-1/level-one-core.html#ID-1312295772 [00:33:32.0491] > <@ms2ger:igalia.com> Attrs definitely had a Text child, I'm not even sure if I managed to remove that oh, well TIL about that part at least [00:33:45.0772] Seems like they don't anymore in Firefox [00:34:35.0633] well that’s progress at least [00:35:18.0141] I thought that issue would be “10 minute task” level-of-effort kind of issue [00:36:38.0156] … but I often find that when I go to try to resolve an issue like that, and start reading the surrounding text… it’s something that’s gonna take a lot longer than just 10 minutes to unwind [00:37:07.0220] I am now reminded of https://twitter.com/bhathos/status/1390109363180904453 [00:37:26.0192] https://twitter.com/bhathos/status/1390109778945486851 [00:39:14.0776] to which I would respond to say, *I hear you*, but also to say, *You can edit the text of MDN articles directly in the GitHub UI* (though yeah it does now also require an additional step of raising a PR and getting it approved) [00:40:36.0347] but the fact a PR is now required hopefully helps mitigate wrong info getting added into MDN to begin with (compared to the old days when anybody could go in there and add/change anything they wanted, whether it was right or wrong, without review) [00:41:53.0587] I lost count of how many times in the old days I had to revert a change that somebody made to “correct” the JavaScript media type to make it `application/javascript`, as the “standard” says it should be [00:42:22.0499] anyway, end of monologue [00:50:34.0851] Duh, it should be `text/css` [00:50:44.0807] heh [01:07:18.0411] If only the whole XML project had succeeded we could have `application/javascript+xml` where programmers would simply write out the AST in XML syntax and use XSLT to implement macros. [01:12:29.0064] jgraham: we can still hope [01:12:57.0016] /me tries to figure out which MDN article to sneak that quote from jgraham into [01:18:33.0985] Surely they already have a page "Bad ideas (jgraham)" [01:36:30.0060] Ms2ger: I successfully removed Text children of Attr in Chromium, somewhere in https://crbug.com/305105 saga. [01:57:31.0361] hmm yeah I now vaguely recall hearing before about the _“make Attr no longer a subtype of Node”_ thing — but I didn’t care, because it didn’t affect me personally and instead I just felt slightly sorry for the people who had to spend time on fixing it 😂 [01:58:07.0474] * hmm yeah I now vaguely recall hearing before about the _“make Attr no longer a subtype of Node”_ thing — but I didn’t care, because it didn’t affect me personally and instead I just felt slightly sorry for the people who had to spend time on fixing it 😂 [01:59:21.0622] sideshowbarker: That was good procrastination material for me, but ultimately the simplification didn't get all the way to the original goal. But I think where things ended up isn't terrible. [01:59:46.0032] "less terrible", I'd say :) [01:59:55.0249] I'll take it. [02:00:31.0437] Oh, and contrats on the house, btw :) [02:00:50.0931] Thanks! [02:01:59.0700] My hands are now messed up from strong chemicals (cleaning products) last week, I don't recommend that part of moving. [02:02:11.0329] * My hands are now messed up from strong chemicals (cleaning products) last week, I don't recommend that part of moving. [02:02:52.0579] /me > <@foolip:matrix.org> sideshowbarker: That was good procrastination material for me, but ultimately the simplification didn't get all the way to the original goal. But I think where things ended up isn't terrible. hands foolip a shiny coin [02:03:23.0372] Do you have two? :) [02:03:36.0714] Moving is Sweden is, I think, a slightly different experience from moving elsewhere. Or at least in the UK there's no requirement to have magazine-quality photos of the house done when you're listing it, and lower (but not zero) expectations of cleanliness when you move in/out. [02:03:45.0677] * Moving is Sweden is, I think, a slightly different experience from moving elsewhere. Or at least in the UK there's no requirement to have magazine-quality photos of the house done when you're listing it, and lower (but not zero) expectations of cleanliness when you move in/out. [02:03:54.0490] Ms2ger: two hands yes, two houses no, but an apartment [02:04:28.0732] jgraham: magazine-quality photos pretty much sums up the expectation, yeah. [02:04:32.0114] I meant two coins, but glad you have two hands :) [02:05:29.0507] Oh :) [02:07:04.0006] Ms2ger: I got a coin made especially for you 👆 [02:07:28.0122] 🙏 [02:08:57.0115] Back to work for me [02:13:56.0545] foolip: I think I know a way to give a name to the #testing room, if you're a mod on the IRC side. [02:15:02.0264] I think sideshowbarker can get ops on that room, maybe? [02:21:40.0501] Andreu Botella (he/they): looks like I'm not [02:22:42.0134] For reference, the way to do it is by linking your matrix account with your IRC username: https://matrix-org.github.io/matrix-appservice-irc/latest/usage.html#customising-your-experience [02:23:06.0971] I believe that should now give you Matrix permissions in the room, which would let you go into Settings and set a name [02:24:03.0244] Either that or follow the more complicated way of changing things in https://matrix-org.github.io/matrix-appservice-irc/latest/room_configuration.html [03:08:57.0991] > <@jgraham_:matrix.org> I think sideshowbarker can get ops on that room, maybe? OK, I’ve got ops on #testing now — but jgraham you now do too [06:59:46.0126] zcorpan_: about https://github.com/w3c/aria-practices/issues/960, I don’t have a Windows environment in which to try to reproduce that. As far as what Nick Schonning said, I don’t think the `asciiquotes` option is actually going to have any effect. And if it’s not due to the Windows shell environment not having its `LANG` equivalent set to UTF-8 encoding, then I don’t have much idea what else the problem could be. [07:02:07.0179] oh, maybe invoking `java` with `-Dfile.encoding=UTF-8` is an idea — but other than that and checking the `LANG` encoding, I don’t know [07:02:55.0491] I definitely have no idea what "page files" (that Nick refers to) means in the Windows context [07:03:18.0169] sideshowbarker: ok, thx [07:03:24.0924] * oh, maybe invoking `java` with `-Dfile.encoding=UTF-8` is an idea — but other than that and checking the `LANG` encoding, I don’t know [07:27:17.0142] Hi! I have a spec question but I'm not sure if it belongs here. The WHATWG HTML spec defines the **unfocusing steps** https://html.spec.whatwg.org/multipage/interaction.html#unfocusing-steps, but it's not readily apparent what triggers these besides calling _Element.blur()_. Specifically, I'm interested in whether the element no longer **being rendered** https://html.spec.whatwg.org/multipage/rendering.html#being-rendered should trigger these steps. Intuitively I would say yes, but I seem to get conflicting behaviour from browsers: Firefox and Safari no, Chrome yes. [07:37:33.0654] lpd: per spec I think Chrome is wrong and I don't think there's an issue on changing this [07:39:47.0788] > <@annevk:mozilla.org> lpd: per spec I think Chrome is wrong and I don't think there's an issue on changing this Wrong because no spec anywhere says these steps may be taken? Do you think this is a more appropriate scenario to file a spec issue or a chrome bug? [07:40:45.0275] lpd: I guess it depends on what you want, but a Chrome bug seems like a good start to me [07:41:04.0531] lpd: and yeah, unless a spec states something ought to happen, it shouldn't happen [07:41:13.0728] Cool, thanks! [07:42:47.0051] zcorpan_: there was already a zcorpan without underscore? :-) [08:07:22.0776] (matrix is not very forgiving with lost passwords. the account is forever lost when you don't haven an email address set) [08:54:08.0626] freddy: Yeah matrix has a lot of quirks. There are also some issues with how it handles SSO — or at least with how some matrix clients do. In my experience some matrix clients don’t work at all work at all with chat.mozilla.org’s SSO at least — though Element seems to, reliably. [09:13:44.0445] zcorpan_: I posted https://github.com/w3c/aria-practices/issues/960#issuecomment-851580808. I guess the Windows way to set the encoding of the shell environment to UTF-8 is to run the command `chcp 65001`. [09:16:53.0211] sideshowbarker: OFC, perfectly obvious! [09:45:44.0964] > <@fbraun:mozilla.org> (matrix is not very forgiving with lost passwords. the account is forever lost when you don't haven an email address set) Well, you can ask your homeserver admin to manually intervene, but that's easier with small servers.