| 08:04 | <annevk> | mathiasbynens: did you write a PR for https://github.com/whatwg/encoding/issues/73#issuecomment-255601209? |
| 08:43 | <annevk> | Do we need some kind of "bug" label for stuff like https://github.com/whatwg/html/issues/1950? |
| 08:44 | <annevk> | For those "normative change" doesn't really seem appropriate |
| 08:44 | <annevk> | Although technically it is one |
| 08:54 | <annevk> | heycam: ping for https://github.com/whatwg/dom/pull/347 |
| 08:54 | <botie> | will do |
| 08:54 | heycam | looks |
| 09:19 | <annevk> | zcorpan: https://github.com/whatwg/html/pull/1717 is in need of some action |
| 09:19 | <annevk> | zcorpan: it's quite old and not marked as "do not merge" |
| 09:20 | <annevk> | zcorpan: I guess at this point I'd prefer that we have some tests before landing |
| 09:20 | <zcorpan> | annevk: yeah, i wanted to write tests for it |
| 09:20 | <zcorpan> | sorry for the delay |
| 09:21 | <annevk> | No worries, I thought it might have dropped from your radar |
| 09:21 | <annevk> | I also merged a trivial 12-day old PR just now that was seemingly forgotten |
| 09:29 | <annevk> | foolip: would be good to get https://github.com/w3c/web-platform-tests/pull/4043 merged as the currently checked in version of the test doesn't run properly in Firefox |
| 09:29 | <foolip> | annevk: just needs rubberstamping? |
| 09:32 | <annevk> | foolip: yeah, pretty much |
| 09:34 | <foolip> | annevk: "pick the first with activation behavior <input type=checkbox>" seems to fail everywhere, is it for a good reason? |
| 09:38 | <annevk> | foolip: is that the one about bubbling? |
| 09:38 | <foolip> | annevk: oh, wait, it's noe one of the new tests in this PR, nevermind |
| 09:40 | <annevk> | foolip: there might be one that fails everywhere where I argued browsers should change |
| 09:40 | <foolip> | annevk: but why is the disabled test using nested input elements, is there some significance to that? |
| 09:40 | <annevk> | foolip: yeah, to make sure the disabled input is not skipped when finding an element with activation |
| 09:41 | <foolip> | annevk: oh, so it could have been any parent with activation behavior |
| 09:41 | <annevk> | Yeah |
| 09:45 | <foolip> | annevk: should I merge? you have the power, right? |
| 09:52 | <annevk> | foolip: I can do it post-break |
| 11:54 | <mathiasbynens> | annevk: ah, no I didn’t, sorry |
| 11:56 | <annevk> | mathiasbynens: I ended up writing one, it wasn't too hard |
| 12:36 | <smaug> | annevk: I can't seem to find in the spec where 'enter' causes a click to be dispatched |
| 12:37 | <annevk> | smaug: "An element that has its tabindex focus flag set but does not otherwise have an activation behavior defined has an activation behavior that does nothing." |
| 12:38 | <smaug> | annevk: that does say anything about 'enter' |
| 12:38 | <smaug> | doesn't |
| 12:38 | <annevk> | smaug: there's another section that says |
| 12:38 | <smaug> | I mean literally, what somewhere says keyevents for 'enter' should trigger click |
| 12:38 | <annevk> | "The user agent should allow the user to manually trigger elements that have an activation behavior, for instance using keyboard or voice input, or through mouse clicks. When the user triggers an element with a defined activation behavior in a manner other than clicking it, the default action of the interaction event must be to fire a click event at the |
| 12:38 | <annevk> | element." |
| 12:39 | <smaug> | ok, not defined then in this spec |
| 12:39 | <annevk> | smaug: both of those quotes are from HTML |
| 12:40 | <smaug> | annevk: right. Nothing talks about 'enter'. |
| 12:43 | <annevk> | smaug: oh right, that's somewhat up to the user agent, but I don't think Firefox makes it work with any key |
| 12:43 | <annevk> | smaug: and "Enter" (or maybe "Space") is a rather obvious choice since it works elsewhere already, like for <a> and <button type=submit> |
| 12:44 | <smaug> | for links at least it is 'enter' |
| 12:45 | <smaug> | annevk: and if 'enter' is defaultPrevented = true, click won't be dispatched |
| 12:45 | <smaug> | stuff to spec somewhere I guess |
| 12:45 | <annevk> | That's what the quote already says |
| 12:45 | <annevk> | Firing "click" would be the "default action" (so when canceled not fire) |
| 12:46 | <annevk> | Of course, "default action" needs to be reworded and indeed, we need an interaction event spec |
| 12:46 | <annevk> | I guess the main question here is if we can make this change at all |
| 12:47 | <smaug> | annevk: which change? |
| 12:47 | <smaug> | you mean browsers following the spec? |
| 12:47 | <annevk> | smaug: allow <div tabindex onclick=...> to be triggered through a keyboard |
| 12:47 | <smaug> | right |
| 12:48 | <smaug> | would need to write some quite complicated telemetry before such change could be done |
| 12:49 | <smaug> | and even then the telemetry wouldn't tell anything too accurate |
| 12:49 | <nox> | annevk: Shouldn't we put in the spec that 20 nested document.write is the limit? |
| 12:50 | <nox> | (AFAIK WebKit and Firefox both do that.) |
| 12:59 | <annevk> | nox: I guess we could |
| 13:02 | <nox> | annevk: Ok. Filed. https://github.com/whatwg/html/issues/1954 |
| 13:24 | <smaug> | annevk: is XHR exposed to service workers? |
| 13:25 | <annevk> | smaug: no |
| 13:25 | <smaug> | though, doesn't matter. Fetch without streaming could use blobs |
| 13:25 | <smaug> | in efficient way |
| 13:53 | <zcorpan> | annevk: re https://github.com/w3c/web-platform-tests/pull/4043#issuecomment-255387868 arrow functions have an implied return i think |
| 13:53 | <zcorpan> | oh Domenic commented on that already |
| 13:54 | <annevk> | yeah |
| 13:57 | <annevk> | Hmm, an update of macOS requires an update to command-line tools? |
| 13:57 | <annevk> | Got a linker error while playing with Pascal |
| 13:59 | <Ms2ger> | You might need to agree to the eula again or something |
| 14:00 | <annevk> | The amount of boilerplate Hixie's DOM API requires compared to JavaScript seems rather excessive |
| 14:10 | <gsnedders> | annevk, nox: IIRC document.write limit is needed for web compat (!) |
| 14:13 | <nox> | gsnedders: All the reason more to specify it, right? |
| 14:31 | <annevk> | I might have figured out a fix for https://github.com/whatwg/html/issues/46 |
| 14:31 | <annevk> | Not making me a big fan of Pascal though |
| 14:32 | <jgraham> | I'm sure both pascal users will be heartbroken |
| 14:32 | Ms2ger | is |
| 14:33 | <annevk> | jgraham: I'm somewhat surprised I actually got it to work |
| 14:35 | <annevk> | zcorpan: what's the status on https://github.com/whatwg/wattsi/pull/21? wontfix? |
| 14:44 | <Domenic> | annevk: woohoo on https://github.com/whatwg/wattsi/pull/30 |
| 14:45 | <zcorpan> | annevk: yeah i guess |
| 14:51 | <MikeSmit1> | just now opened https://github.com/whatwg/html/pull/1958 for annevk’s https://github.com/whatwg/html/issues/1943 “Meta: update TEAM PR advice” |
| 14:52 | <MikeSmit1> | I guess all editors should review that to make sure that in their environment the steps actually work as described |
| 14:53 | MikeSmith | steps away for the next ~2 hours |
| 15:21 | <annevk> | MikeSmith: wow, ended up a lot more detailed than I thought |
| 15:22 | <Ms2ger> | Domenic, is there any reason to use the incumbent script settings object for importScripts()? It's always going to be the same as the relevant or current one anyway |
| 15:26 | <zcorpan> | MikeSmith: FYI https://github.com/whatwg/html/pull/1945#discussion_r84686760 |
| 15:31 | <Domenic> | Ms2ger: no, we should fix that, nice |
| 15:32 | <Ms2ger> | Yay :) |
| 15:33 | <Ms2ger> | Domenic, fwiw, I'm trying to audit all the places where Servo sets the "current Realm", so more bugs may be coming :) |
| 15:34 | <Domenic> | Ms2ger: woohoo, should be fun! |
| 15:34 | <Ms2ger> | Well, I sure am glad I'm paid for it :) |
| 15:51 | <Domenic> | I've asked before but I'll try again: anyone know a way to bypass the wizard in Firefox's Bugzilla? |
| 15:52 | <jgraham> | Which wizard? |
| 15:52 | <Domenic> | The one that makes me say whether it's Firefox or Thunderbird or whatever, then asks for steps to reproduce etc. |
| 15:53 | <Domenic> | I just want something like WebKit has with a raw Bugzilla editing interface |
| 15:53 | <Domenic> | Right now I am at https://bugzilla.mozilla.org/enter_bug.cgi?format=guided#h=bugForm%7CCore%7CDOM%3A+Core+%26+HTML |
| 15:53 | <jgraham> | It's the URL without format=guided |
| 15:53 | <jgraham> | But you might get redirected there unless you have editbugs or something |
| 15:54 | <Domenic> | Yeah redirected |
| 15:54 | <jgraham> | I think you need editbugs then |
| 15:55 | <jgraham> | https://bugzilla.mozilla.org/page.cgi?id=get_permissions.html |
| 15:55 | <jgraham> | (I am sort of guessing though; I could just ask someone) |
| 15:55 | <Domenic> | Excellent, thank you |
| 15:56 | <jgraham> | (I was right) |
| 15:59 | <Ms2ger> | Domenic, you want it for d⊙.? |
| 16:00 | <Domenic> | Ms2ger: yep! |
| 16:05 | <gsnedders> | Ms2ger: if you're giving out editbugs… |
| 16:06 | <Ms2ger> | I sent email, which is being held for moderation |
| 16:06 | <Ms2ger> | I'll see who answers the call :) |
| 16:07 | <Ms2ger> | Domenic, allegedly done |
| 16:07 | <Domenic> | Ms2ger: \o/ thank you! |
| 16:07 | <Domenic> | yesss no more wizard |
| 16:07 | <Ms2ger> | gsnedders, geoffers+mozilla@? |
| 16:08 | <gsnedders> | Ms2ger: yeah |
| 16:11 | <jgraham> | Domenic: Seems like we will also now get a pref to disable it |
| 16:12 | <Ms2ger> | gsnedders, same |
| 16:20 | <Domenic> | annevk: DOM did not deploy again :( |
| 16:22 | <annevk> | Domenic: fixed manually, I'm not sure how to reproduce, since GitHub claims things are okay and script still functions |
| 16:22 | <Domenic> | annevk: yeah. Well I'll hopefully have time to rewrite it all anyway soon to do auto-building and such. |
| 16:23 | <annevk> | TabAtkins: Event's constructor and that of all its subclasses need to point to a single dfn. Is that possible? |
| 16:46 | <gsnedders> | is it right to say taht an unhandled JS error will either result in a error or unhandledrejection event on the global object? |
| 16:54 | <Domenic> | gsnedders: for some definition of unhandled, sounds correct. Although cross-origin we censor error and just don't fire unhandledrejection. |
| 16:55 | <gsnedders> | Domenic: right. |
| 19:14 | <zcorpan> | Domenic: https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/9496749/ was cut off right before "<0.0001%" - maybe add a comment with s/</less than/ |
| 20:00 | <mathiasbynens> | zcorpan: smells like XSS |
| 20:00 | <mathiasbynens> | …or an overly crappy XSS filter |
| 20:00 | <zcorpan> | right |
| 20:01 | <zcorpan> | makes one wonder if they can be actually XSSed though :-P |
| 20:01 | <Domenic> | It's just Markdown |
| 20:01 | <Domenic> | I kind of forgot what I typed so I figure I'll just let them look at the source code |
| 20:02 | <mathiasbynens> | fair enough |
| 20:03 | <zcorpan> | i'm not very confident anyone will look at the bug's source code though |
| 22:12 | <KiChjang> | how does one submit a change on whatwg/html again? |
| 22:12 | <KiChjang> | is it just by editing the source file? |
| 22:12 | <KiChjang> | nvm, i got it mixed up with fetch/xhr |
| 22:12 | <KiChjang> | there really is only 1 file to edit |
| 22:13 | <KiChjang> | when there's a src.html and a .html file in the repo, which one should be edited? |
| 22:42 | <mathiasbynens> | KiChjang: the src one |
| 22:42 | <KiChjang> | should i regenerate the .html one if i can? |
| 22:42 | <KiChjang> | using bikeshed right? |
| 22:44 | <mathiasbynens> | KiChjang: no, Anolis |
| 22:44 | <mathiasbynens> | KiChjang: but don’t worry about the generated .html; see https://github.com/whatwg/fetch#pull-requests |
| 23:31 | <jyasskin> | Tobie, TabAtkins, Domenic, annevk : I'm out for parental leave. See you in January. |
| 23:31 | <TabAtkins> | Congrats and see you later! |
| 23:31 | <jyasskin> | :) |
| 23:51 | <gsnedders> | jyasskin: congrats! |