| 00:01 | <gsnedders> | because, like, if it's spec'd now we get have shared tests for it |
| 00:01 | <gsnedders> | YES, I'M JUST CONSPIRING TO GET TESTS FOR EVERYTHING, OKAY?! |
| 00:01 | <gsnedders> | I don't actually care about your silly specs! :P |
| 00:01 | <TabAtkins> | I mean, good. That's what I'm paying you for. ^_^ |
| 00:01 | <gsnedders> | Pretty sure that's Rich. ;P |
| 00:02 | <TabAtkins> | Not while he's on vacation! |
| 00:02 | <gsnedders> | Well, except this motnh. |
| 00:02 | <TabAtkins> | *Rick |
| 00:02 | <gsnedders> | …I can't type. |
| 00:06 | <gsnedders> | TabAtkins: is http://test.csswg.org/suites/css21_dev/nightly-unstable/html4/uri-013.htm wrong/outdated? |
| 00:06 | <gsnedders> | TabAtkins: per 2.1? per Syntax? |
| 00:06 | <TabAtkins> | ooh, malformed url() parsing. *probably* outdated a little bit, let me check. |
| 00:06 | <TabAtkins> | Since Syntax locked down and simplified the rules for url(). |
| 00:08 | <gsnedders> | do we need errata on 2.1? |
| 00:08 | <TabAtkins> | I'm neutral on how much needs to be applied backwards to 2.1, particularly when the changes are difficult to backport (like parsing) |
| 00:10 | <TabAtkins> | ah, yup, #3 is definitely outdated. still looking at the rest. |
| 00:10 | <TabAtkins> | and #4 |
| 00:13 | <gsnedders> | the problem with no backporting changes to 2.1 is we end up needing different tests if we actually want to claim we're still testing 2.1 |
| 00:14 | <TabAtkins> | That's why the correct answer is to start chopping out parts of 2.1 that have been moved to other specs. |
| 00:14 | <gsnedders> | I agree. |
| 00:14 | <TabAtkins> | But people keep fighting against this for some reason. |
| 00:14 | <gsnedders> | I'd be totally fine with errata just making large parts of 2.1 undefined behaviour provided we have a better spec to point at |
| 00:14 | <gsnedders> | so we don't actually need to do anything useful |
| 00:15 | <TabAtkins> | But anyway, Syntax can't be accurately ported to 2.1 without just copying it over. 2.1 has a totally different parsing definition that can't be modded into precision in the same way. |
| 00:15 | <gsnedders> | oh, I'm well aware |
| 00:15 | <gsnedders> | I have no idea what a sensible errata would look like |
| 00:15 | gsnedders | wonders if this is a GitHub issue or a www-style issue :) |
| 00:15 | <TabAtkins> | GH plz |
| 00:17 | <gsnedders> | should I be in some group so I can add labels etc. in csswg-drafts? |
| 00:17 | <TabAtkins> | yes |
| 00:17 | <gsnedders> | or is that limited to the cool kids? |
| 00:17 | <TabAtkins> | #12 is invalid |
| 00:17 | <TabAtkins> | you're a wg member |
| 00:18 | <TabAtkins> | #13 is valid, but not testing *precisely* what it thinks it's testing anymore. It passes by accident, in other words. (It's making the same legacy assumption that #12 is, just happens to still succeed.) |
| 00:18 | <gsnedders> | Still not sure how I managed to wangle my way into the WG. It's not like I'm even that great at CSS. :) |
| 00:20 | <TabAtkins> | #14 is invalid |
| 00:21 | <gsnedders> | so Fx and Blink are correct? |
| 00:21 | <TabAtkins> | blink is, at least. not looking at ff righ tnow |
| 00:21 | <gsnedders> | identical to Blink |
| 00:22 | <TabAtkins> | #15 and #16 are like #13 - technically valid, but not actually testing exactly what they think they are. |
| 00:22 | <TabAtkins> | Oh, and #17, sorry. |
| 00:25 | <TabAtkins> | and done |
| 00:25 | <TabAtkins> | so yeah, Blink is totally correct - all its failures are test mistakes. |
| 00:25 | <TabAtkins> | (Or rather, tests against 2.1 which have had reality changed underneath them.) |
| 00:26 | <gsnedders> | unless the latest 2.1 errata deals with this? |
| 00:27 | <gsnedders> | I don't think it does? |
| 00:28 | <gsnedders> | no, that's all about at-rules |
| 02:36 | <MikeSmith> | annevk: added you to https://github.com/w3c/resource-timing with admin access. heads up igrigorik; see http://logs.glob.uno/?c=freenode%23whatwg&s=16+Aug+2016&e=16+Aug+2016#c1002845 for context |
| 03:00 | <MikeSmith> | annevk: http://stackoverflow.com/questions/38957103/fetch-api-rejects-with-networkerror-when-hosted-on-localhost-firefox-only |
| 05:24 | <MikeSmith> | annevk: https://github.com/jsdom/whatwg-url/pull/58#issuecomment-240170242 |
| 05:35 | <MikeSmith> | hayato: http://stackoverflow.com/questions/38976819/composing-v1-nested-web-components |
| 06:37 | <annevk> | MikeSmith: yeah, that jsdom issue is why I have https://github.com/whatwg/url/issues/140 |
| 06:37 | <annevk> | MikeSmith: folks assume obvious meaning rather than follow the link |
| 06:39 | MikeSmith | looks |
| 06:39 | <MikeSmith> | ah yeah OK |
| 06:54 | <mkwst> | annevk: I'm adding more bikeshed metadata to Fetch; do you care whether it's multiple commits in a single PR, or multiple PRs? Trying to give it to you in reviewable chunks... |
| 06:56 | <annevk> | mkwst: multiple commits in one PR seems fine |
| 06:56 | <annevk> | mkwst: then we can add it as a single commit |
| 06:56 | <annevk> | (in the end) |
| 06:56 | <mkwst> | Ok. Then I'll extend https://github.com/whatwg/fetch/pull/370 a bit. |
| 06:58 | <annevk> | cool |
| 07:48 | <mkwst> | annevk: https://github.com/whatwg/fetch/pull/370 has enough for you to chew through, I think. I'll spend another hour on it sometime later. |
| 07:48 | <mkwst> | Can I assume you'd accept similar patches to HTML? |
| 07:58 | <annevk> | mkwst: yeah, we landed one yesterday that I created |
| 07:59 | <mkwst> | annevk: Ok, great. I hope you won't be too sad if I'm opportunistic there and only tag the definitions I need. :) |
| 07:59 | <annevk> | mkwst: nope |
| 08:34 | <annevk> | mkwst: sorry, I'm commenting on the generated file again :-/ |
| 08:34 | <mkwst> | annevk: No worries. The line numbers aren't terribly useful anyway; I'm just searching for the text. |
| 08:34 | <mkwst> | I'm fiddling with HTML right now, will get to your Fetch feedback in a few minutes. |
| 08:34 | <annevk> | mkwst: yeah, but the 100 col comments are likely wrong due to that |
| 08:35 | <annevk> | mkwst: I removed the comments |
| 08:35 | <mkwst> | Oh, well. *shrug* Whatever. I'll figure it out. :) |
| 08:35 | <mkwst> | This is going to make my life a bunch simpler, so I can deal with a little annoyance up front. :) |
| 08:35 | <mkwst> | Well, this tiny part of my life, but still. :) |
| 08:36 | <annevk> | mkwst: let me know when you consider it ready and I'll give it another go |
| 08:36 | <mkwst> | Sure, thanks! |
| 08:38 | <mkwst> | (For clarity, HTML breaks on the last whitespace before 100col, regardless of content? |
| 08:38 | <mkwst> | ) |
| 08:41 | <annevk> | mkwst: yeah |
| 08:41 | <annevk> | HTML requires \s+ to find terms, Fetch doesn't |
| 08:47 | <annevk> | Anyone interested in reviewing Attr's comeback? https://github.com/whatwg/dom/pull/299 |
| 09:39 | <annevk> | Anyone interested in reviewing improved global object references in DOM? https://github.com/whatwg/dom/pull/296 |
| 09:51 | <mkwst> | annevk: Took another pass at Fetch, found one more newline issue. I think the rest is taken care of. Mind taking another look when you have a minute? |
| 09:57 | <nox> | annevk: Attr as Node, what is the value of prevSibling and nextSibling etc? |
| 09:57 | <Ms2ger> | Null |
| 09:58 | <nox> | Ok. |
| 10:04 | <annevk> | mkwst: later today |
| 10:04 | <annevk> | nox: parent is null too |
| 10:18 | <smaug____> | anyone here familiar with houdini and worklets? |
| 10:43 | <annevk> | smaug____: a bit |
| 10:44 | <smaug____> | annevk: just wondering what kinds of objects are expected in worklets |
| 10:44 | <smaug____> | looks like Audio spec will use Events there |
| 10:45 | <smaug____> | so Gecko implementation will certainly need cycle collector in threads where worklets are run |
| 10:45 | <annevk> | smaug____: I see, I don't really know much about that, except that I believe it's nothing by default |
| 10:45 | <annevk> | smaug____: and I haven't seen an issue reported against DOM to expose the necessary APIs in worklets either |
| 10:46 | <smaug____> | (how events are used in Audio API seems to be all unclear https://github.com/WebAudio/web-audio-api/issues/937 ) |
| 10:57 | <annevk> | I see |
| 10:57 | <annevk> | They should probably figure that out then |
| 10:58 | <annevk> | Please make sure they patch DOM somehow if new things need to be exposed in worklets |
| 12:11 | <annevk> | smaug____: any thoughts on what we should do with https://www.w3.org/Bugs/Public/show_bug.cgi?id=25002? Edge 14 still has it |
| 12:17 | <smaug____> | annevk: if we add that, we probably need to add the rest of the IEism too |
| 12:17 | <smaug____> | window.event and src<something> |
| 12:17 | <smaug____> | and whatever there is |
| 12:20 | <annevk> | Hmm, you think those are tightly coupled? |
| 12:20 | <annevk> | There's a different open bug for those |
| 12:20 | <annevk> | With lots of compat concerns from bz |
| 12:20 | <annevk> | I guess I'm not going to touch that |
| 12:23 | <smaug____> | annevk: well, if one is using one IEism, I'd expect also other IEisms being used |
| 12:24 | <smaug____> | but this is just a guess of course |
| 12:50 | <annevk> | mkwst: could you comment on https://www.w3.org/Bugs/Public/show_bug.cgi?id=25529#c11 perhaps? |
| 12:51 | <annevk> | mkwst: cloning of elements allows you to bypass CSP style attribute checks |
| 12:51 | <annevk> | mkwst: also, I hope those CSP style attribute checks don't actually prevent the attribute from being set... |
| 13:12 | <mkwst> | annevk: Will take a look. The style attribute checks are a mess, honestly. CSSOM even more so. |
| 13:13 | <annevk> | Not really surprised |
| 13:17 | <JakeA> | Did anyone get a confirmation email when they booked one of the recommended TPAC hotels via http://www.delegatehotels.com/e/w3c2016? |
| 13:17 | <JakeA> | I'm pretty sure I booked, but struggling to find myself any proof of that |
| 13:18 | <Domenic> | annevk: no comment about how Attr being a Node is a historical artifact that we are all very sad about? |
| 13:18 | <Domenic> | (in the spec) |
| 13:19 | <MikeSmith> | JakeA: when I booked at Hotel Vila Gale Opera I got a confirmation mail from accommodation⊙bc |
| 13:20 | <MikeSmith> | JakeA: if you need confirmation, Alexandra Lacourba <alex⊙wo> might be able to help |
| 13:21 | <annevk> | Domenic: should probably be fixed as part of https://github.com/whatwg/dom/issues/205 |
| 13:21 | <JakeA> | MikeSmith: thanks. Pretty sure I booked, but can't find a confirmation. Maybe I dreamt booking it. |
| 13:21 | annevk | booked via Mozilla's agent |
| 13:21 | <Ms2ger> | Domenic, you're nyc-based, are you not? |
| 13:22 | <annevk> | Domenic: also, the really sad thing would have been it supporting child nodes, so I think the damage is fairly minor |
| 13:25 | <Domenic> | Ms2ger: yep |
| 13:27 | <Ms2ger> | Do you have an opinion on a 1h25 layover (international -> domestic) in EWR? |
| 13:29 | <Domenic> | Not really, sorry; I haven't been to EWR in a while... |
| 13:31 | <Ms2ger> | Np, thanks |
| 13:35 | <zcorpan> | annevk: would it help to just say integer item instead of "required line break count"? |
| 13:37 | <annevk> | zcorpan: the list can't only contain strings? |
| 13:38 | <zcorpan> | annevk: no. strings and integers |
| 13:38 | <annevk> | zcorpan: oh I see, it's mentioned in the introduction |
| 13:39 | <annevk> | zcorpan: I think it would help to say "or a positive integer (a required line break count)" |
| 13:39 | <zcorpan> | ok |
| 13:40 | <annevk> | zcorpan: and later on maybe just 2 and 1 and again have that parenthesis for clarity |
| 13:46 | <zcorpan> | annevk: ok, PTAL |
| 14:26 | <mkwst> | annevk: So, we need to adjust the HTML style rules to "break on the last whitespace before 100col, regardless of content, unless the content is hard-coded in the preprocessor scripts as being on a single line". :) |
| 14:28 | <annevk> | mkwst: yeah 😞 |
| 14:31 | <gsnedders> | Ms2ger: you're never gonna make that connection, far as I know |
| 14:31 | <gsnedders> | Ms2ger: given you can easily spend the best part of an hour in the immigration queue before you even get to picking up your luggage and checking it in for your next flight, yet alone getting through security and to the gate |
| 14:32 | <Ms2ger> | No way I'm checking luggage |
| 14:33 | <Ms2ger> | You don't happen to have experience in Montreal? |
| 14:33 | <gsnedders> | nah |
| 14:33 | <gsnedders> | I presume it'd be international to international then? |
| 14:33 | <gsnedders> | I don't think Canada make everyone off an international flight go through immigration, which probably means an equally tight connction is better, but do check. |
| 14:35 | <Ms2ger> | Would be intl -> canada, but it's only 1h05 |
| 19:37 | <JakeA> | MikeSmith: did you pay up front for the hotel (as in, will I see it in my credit card statement if I booked it) |
| 19:58 | <JakeA> | Opening that question out to others - when you booked TPAC hotel (via the W3), do you pay up-front? |
| 20:09 | <wanderview> | JakeA: last year it was cheaper to book separately from the TPAC conference block of rooms |
| 20:10 | <JakeA> | ohhh |
| 20:10 | <JakeA> | Cheers |
| 20:10 | <wanderview> | JakeA: in theory the conf organizers might be annoyed by that, though |
| 20:10 | wanderview | ascii shrugs |
| 20:11 | wanderview | is too lazy to look up the ascii shrugy |
| 20:55 | <TabAtkins> | wanderview: Just google for "shruggie" ¯\_(ツ)_/¯ |
| 20:55 | <wanderview> | TabAtkins: you underestimate my lazyness |
| 21:03 | <zcorpan> | you can set words that expand to other words in os x prefs. but i was now disappointed it doesn't work in Limechat :shrug: |
| 21:03 | <zcorpan> | Domenic: |
| 21:03 | <botie> | Domenic: is there some private discussion going on? |
| 21:03 | <zcorpan> | From https://rocallahan.github.io/innerText-spec/ |
| 21:03 | <zcorpan> | with the following normative changes: |
| 21:03 | <zcorpan> | * Defined behavior for non-CSS UAs. |
| 21:03 | <zcorpan> | * The setter is better defined. |
| 21:03 | <zcorpan> | * Added [CEReactions, TreatNullAs=EmptyString] |
| 21:03 | <zcorpan> | to the IDL. |
| 21:03 | <zcorpan> | Fixes https://github.com/whatwg/html/issues/465. |
| 21:03 | <zcorpan> | Remaining issues: #1679 |
| 21:03 | <Domenic> | zcorpan: looks good! |
| 21:04 | <zcorpan> | thanks! |
| 23:41 | <MikeSmith> | JakeA: for the TPAC hotel I did not pay up front as far as I recall |
| 23:42 | <JakeA> | Cheers! |
| 23:47 | <gsnedders> | JakeA: I def didn't |
| 23:50 | <JakeA> | Ok so I either booked it like I vaguely remember, but didn't get confirmation, or my brain is just making stuff up. |
| 23:52 | <gsnedders> | JakeA: I got an email from accomodation⊙bc FWIW |
| 23:59 | <JakeA> | yeah, I don't have one of those. Hmm |