| 00:02 | <TabAtkins> | +1 to entry |
| 00:05 | <tobie> | Domenic: https://github.com/whatwg/infra/pull/53 |
| 01:43 | <gsnedders> | is there no easy way to get a version of ChromeDriver that works with Canary? |
| 05:20 | <MikeSmith> | botie, inform zcorpan the cause of that https://checker.html5.org/parsetree/?parser=html5&content=<%21doctype+html><template><plaintext>a<%2Ftemplate>b&submit=Print+Tree difference is a regression I introduced in the validator htmlparser branch when I fixed a <template>-related bug there https://github.com/validator/htmlparser/commit/140b79631dffaea0f4dfe77144eb86e0b863f3f5 |
| 05:20 | <botie> | will do |
| 06:19 | <annevk> | gsnedders: interesting, so does Python 3 use a more specific strip in that place or was it something else? |
| 06:21 | <annevk> | gsnedders: https://github.com/python/cpython/commit/2a05a314547cded33cdf9169dd4a4ced937a64bd |
| 06:21 | <annevk> | gsnedders: fixed only very recently it seems |
| 07:23 | <gsnedders> | annevk: that's not the bug |
| 07:23 | <gsnedders> | annevk: that's it splitting headers on all unicode line breaks |
| 07:25 | <gsnedders> | annevk: https://github.com/python/cpython/commit/15a312f0238df6add3ca64e9934ab258e06a5006 is where it was fixed, accidentally |
| 07:31 | <annevk> | gsnedders: would all line breaks not include 0C at least? |
| 07:33 | <gsnedders> | annevk: oh, that'll be a fix for a Py3-only bug that broke it in a different way |
| 07:33 | <gsnedders> | hah |
| 07:35 | <annevk> | gsnedders: having looked at that code I'm not sold on it being correct |
| 08:02 | <annevk> | Anyone else getting the GitHub Unicorn? |
| 08:03 | <annevk> | downforeveryone says it's me |
| 08:03 | <tobie> | Yup, me. :( |
| 08:03 | <annevk> | but I guess it would say that even if it got the Unicorn |
| 08:03 | <annevk> | since that's still up in a way |
| 08:03 | <tobie> | yup |
| 08:04 | <tobie> | annevk: I was in the process of getting back to you. |
| 08:04 | <annevk> | About Infra? |
| 08:04 | <tobie> | yes |
| 08:05 | <tobie> | annevk: It seemed you were only x-referencing terms once per paragraph. |
| 08:05 | <annevk> | tobie: oh okay, ignore that then |
| 08:05 | <annevk> | tobie: I'm not sure that was intentional, but it very well might be the case |
| 08:05 | <tobie> | alright |
| 08:05 | <tobie> | annevk: it looked quite systematic |
| 08:24 | <tobie> | annevk: GH is back up for me. |
| 08:26 | <annevk> | same |
| 08:27 | <tobie> | annevk: the 100 character line limit partially breaks diff highlighting in GH. |
| 08:27 | <tobie> | annevk: have you considered what I think TabAtkins refers to as semantic line breaks? |
| 08:29 | <annevk> | tobie: nobody really likes them |
| 08:29 | <tobie> | annevk: well, outside of Tab and I, I guess. :) |
| 08:29 | <annevk> | With nobody I meant the folks working on HTML mostly I suppose |
| 08:30 | <annevk> | I don't really care, but I don't want to reformat everything either |
| 08:30 | <tobie> | annevk: that said, I prefer 100 character limit and not having to work on CSS. |
| 08:30 | <annevk> | And having some consistency within WHATWG is nice for new contributors |
| 08:31 | <tobie> | sure |
| 08:35 | <tobie> | annevk: alright, well, LMK, if you need anything else from me on this PR. |
| 08:49 | <annevk> | tobie: thanks, you weren't in a rush right? going to ping pong back to Domenic |
| 08:50 | <tobie> | annevk: yeah, np. Would me nice if I could get them in shepherd tomorrow, though. |
| 08:50 | <tobie> | annevk: not sure what the timing for *that* is like |
| 08:51 | <annevk> | tobie: if Domenic lands it we should be good for that |
| 08:51 | <tobie> | annevk: awesome |
| 11:40 | <annevk> | If anyone with Bikeshed chops would know the answer to https://github.com/whatwg/xhr/pull/97#issuecomment-272142369 that'd be great |
| 11:40 | <annevk> | I guess converting XMLHttpRequest is blocked another day |
| 11:53 | <annevk> | mkwst: ping for https://github.com/whatwg/url/pull/175 |
| 11:53 | <botie> | will do |
| 11:54 | <annevk> | mkwst: also ping for https://github.com/whatwg/fetch/issues/364 |
| 11:54 | <annevk> | (there's probably more, but I noticed these recently) |
| 15:27 | <annevk> | miketaylr: I'm going inline the Compatibility Standard image |
| 15:27 | <annevk> | miketaylr: pinging your blog each time someone reads the standard seems a little bad |
| 15:29 | <jgraham> | Well not notiably worse than pinging whatwg.org (except from a performance pov I guess) |
| 15:30 | <annevk> | And privacy |
| 15:30 | <jgraham> | Not really |
| 15:30 | <jgraham> | I don't particularly trust you more than miketaylr |
| 15:31 | <jgraham> | Or any of the other people who have access to whatwg server logs |
| 15:31 | <annevk> | jgraham: but you made a decision to visit WHATWG, not miketaylr |
| 15:31 | <annevk> | jgraham: so WHATWG letting miketaylr know about your visits is bad |
| 15:31 | <jgraham> | Right, but if I was really concerned about the difference I would block all cross-origin loads |
| 15:32 | <annevk> | Why should we put the difficulty on the visitor? |
| 15:32 | <annevk> | I really don't understand how this can be controversial, seems like an obvious bad thing |
| 15:32 | <jgraham> | So I don't think it's a bad thing to inline it. I just think it doesn't make a realistic increase in privacy |
| 15:33 | <jgraham> | Because in reality I have no idea who has access to WHATWG server logs, and if I actually care then the internet is so bad in general I would already be reading sites using curl or something |
| 15:35 | <annevk> | I don't understand why you are particularly relevant |
| 15:35 | <annevk> | If you already know most folks involved it doesn't really matter |
| 15:35 | <annevk> | Anyway, whatever |
| 15:35 | <jgraham> | I don't think that's true |
| 15:39 | <jgraham> | It just seems unrealistic that someone would be fine with their ISP, dreamhost, and whatever random group of people have access to whatwg log data knowing about their visits, but be super concerend about an externally hosted image |
| 15:39 | <jgraham> | I mean I agree that in theory it's a smaller set of people who can track you |
| 15:39 | <jgraham> | I'm just not sure the difference is meaningful |
| 15:56 | annevk | wonders where zcorpan is |
| 15:57 | <MikeSmith> | annevk: thought he is at CSS WG meeting this week |
| 15:57 | <MikeSmith> | so, wehrever that meeting is at I guess |
| 16:01 | <Ms2ger> | Seattle |
| 16:01 | <MikeSmith> | ok |
| 16:02 | <MikeSmith> | so 8am there now |
| 16:19 | <annevk> | How long does it generally take for w3c-test.org to be updated? |
| 16:23 | <tobie> | Oh, what's up with > 50 characters showing up on GH commits? |
| 16:23 | <MikeSmith> | annevk: for a PR? |
| 16:23 | <MikeSmith> | to mirror something to submissions ? |
| 16:23 | <MikeSmith> | that should only take a few minutes |
| 16:24 | <jgraham> | tobie: ? |
| 16:25 | <annevk> | MikeSmith: merged PR |
| 16:25 | <tobie> | jgraham: yeah, I thought the first line was truncated above 50. |
| 16:26 | <tobie> | jgraham: I assume I'm not the only one given the WHATWG recs |
| 16:26 | <jgraham> | Oh, idk |
| 16:27 | jgraham | considers 50 character limits to be pointless busywork |
| 16:28 | <tobie> | tobie: considers arguing with repo owners about their stylistic preferences to be pointless busywork |
| 16:28 | tobie | is talking to himself. |
| 16:29 | <jgraham> | Well I'm not arguing with them |
| 16:30 | <jgraham> | I agree they're free to impose whatever unecessary work they like on contributors |
| 16:30 | <jgraham> | It just means I don't plan to worry about your problem :) |
| 16:32 | <tobie> | jgraham: I wasn't suggesting you should be arguing with them, just that I thought my time was better sent complying with stylistic requests than arguing against them. |
| 16:32 | <jgraham> | Agreed :) |
| 16:33 | <tobie> | Anyway, I wasn't mentioning this here to start a discussion on the topic. I was just merely surprised by this and was curious if anyone had context on that behavior. |
| 16:35 | <annevk> | tobie: I think it's just for readability and making it easier to scan commit titles |
| 16:35 | <jgraham> | I think the context is that it wraps at 70 characters |
| 16:35 | <annevk> | tobie: provide the context in the body |
| 16:36 | <tobie> | jgraham: I thought wrapping was a fifty for the first line, hence the limit. |
| 16:36 | <jgraham> | Presumably because people not following the 50 character "rule" would find it super-annoying if it wrapped at 50 |
| 16:36 | <jgraham> | Yeah, but consider the number of projects that don't follow that |
| 16:36 | <tobie> | jgraham: so there's a soft 50 char rules so people don;t get annoyed by a hard 70 char rule? |
| 16:36 | <tobie> | jgraham: software development is hard. |
| 16:37 | <jgraham> | Well the 70 character rule isn't hard either |
| 16:37 | <jgraham> | PR titles can be as long as you want |
| 16:37 | <jgraham> | But GitHub will "helpfully" break it after 70 characters |
| 16:38 | <tobie> | s/;/'/ |
| 16:39 | <jgraham> | (ftr I agree that short summaries are good, but disagree that technical limits of old terminals mean that a 51, or even 70, character summary is unacceptable) |
| 16:39 | <tobie> | oh, so we have different levels of softness. |
| 16:41 | <tobie> | annevk: for context, Domenic changed my commit title from: "Add DFNs for list item & map entry, key, & value" (49) |
| 16:41 | <tobie> | annevk: to "Meta: add <dfn>s for list item and map entry, key, and value" (61) |
| 16:42 | <Domenic> | I just want to fit in the GitHub UI, so 70 is my limit |
| 16:42 | <Domenic> | I guess our documentation says 50 :_/ |
| 16:42 | <annevk> | Maybe we should stop referencing that document and write our own |
| 16:42 | <tobie> | well, what you link to from the doc. |
| 16:42 | <Domenic> | Yeah writing our own would be good I think |
| 16:43 | <annevk> | Meta: prefix is important though |
| 16:43 | <annevk> | Domenic: something for whatwg/meta to host |
| 16:43 | <Domenic> | Part of the reason I think for 50 is so that you can use "Revert of "..."" later and still not run over 70 |
| 16:43 | <tobie> | I knew there was a GH rule + knowing that article led me to believe 50 was a hard rule |
| 16:43 | <tobie> | Domenic: oh, that makes sense (sort of) |
| 16:43 | <Domenic> | annevk: I was thinking about that sort of thing (also for a revamped FAQ), but not sure a github.com URL is the best. Ideally something deployed to whatwg.org |
| 16:44 | <Domenic> | Yeah, to me allowing a potential "revert of" is not quite worth losing those 20 characters... |
| 16:45 | <tobie> | yeah. it's not. |
| 17:11 | <botie> | zcorpan, at 2017-01-12 05:20 UTC, MikeSmith said: the cause of that https://checker.html5.org/parsetree/?parser=html5&content=<%21doctype+html><template><plaintext>a<%2Ftemplate>b&submit=Print+Tree difference is a regression I introduced in the validator htmlparser branch when I fixed a <template>-related |
| 17:11 | <botie> | bug there https://github.com/validator/htmlparser/commit/140b79631dffaea0f4dfe77144eb86e0b863f3f5 |
| 17:11 | <MikeSmith> | zcorpan: fixed that already |
| 17:11 | <zcorpan> | MikeSmith: sweet |
| 17:33 | <annevk> | Domenic: maybe time for a whatwg.org repo or some such |
| 17:33 | <annevk> | Domenic: put the website there |
| 17:44 | <Domenic> | yeah IMO we really should |
| 23:17 | <nox> | zcorpan_: Is it specified in the spec what happens if a script uses document.write() to include a pending parsing blocking script in the document, when the script is executed as part of popping the elements when the end of the parser is reached? |
| 23:42 | <zcorpan_> | nox: can you ask the question in the form of a mock test case? |
| 23:42 | <nox> | zcorpan_: https://twitter.com/nokusu/status/819685935042412544 |
| 23:43 | <zcorpan_> | ah. yeah |