| 00:01 | <rniwa> | Domenic: see my reply on https://github.com/w3c/web-platform-tests/pull/4164 |
| 00:01 | <rniwa> | Domenic: I can modify the test case so that I mutate it thrice |
| 00:52 | <Domenic> | I'm amazed nobody had yet specced how <li>s are assigned numbers accurately |
| 04:00 | <gsnedders> | Domenic: oh god I feel like this is gonna be a wormhole |
| 04:01 | <Domenic> | It's pretty amazing |
| 04:01 | <Domenic> | I think I've got it but I forgot that you could assign display: list-item to random things |
| 04:01 | <Domenic> | I guess that's not implemented yet... maybe will never be |
| 04:03 | <gsnedders> | Domenic: what I want to know is what resets the counter, because that… could be messy |
| 04:03 | <gsnedders> | though I wouldn't be surprised if a bunch of HTML elements just had counter-reset defined on them |
| 04:04 | <Domenic> | Oh, no, it is http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=4651 |
| 04:04 | <Domenic> | gsnedders: hah! as if you could reliably define this in terms of CSS counters |
| 04:04 | <gsnedders> | Domenic: what stops it? |
| 04:08 | <Domenic> | gsnedders: it appears most browsers search for the closest ancestor <ol> **or <ul>**, and if there is no such ancestor, they go for the parent |
| 04:09 | <Domenic> | https://github.com/whatwg/html/pull/2002 is our latest excursion |
| 04:11 | <gsnedders> | oh man, that's so much worse than I expected |
| 04:11 | <gsnedders> | I mean, I shouldn't be surprised. |
| 04:11 | <gsnedders> | I should've left my sense of logic at the door, etc. :) |
| 04:12 | <Domenic> | that's the spirit |
| 04:23 | gsnedders | wonders how much stuff actually relies on the "if there is no such ancestor" case |
| 05:54 | <annevk> | A lot of stuff relies on leaving your sense of logic at the door |
| 07:11 | <gsnedders> | anyone with opinions on documentation for web-platform-tests, plz speak up with high-level issues: https://github.com/w3c/web-platform-tests/issues/4165 |
| 07:15 | <annevk> | gsnedders: the things I look up most often: 1) how to update/reset submodules 2a) various asserts one can use 2b) how to test/async_test/single-page 4) server-side API and examples |
| 07:16 | <annevk> | gsnedders: the other thing I often look for is JavaScript test boilerplate, which I often end up copying from another test and then adjust <title> and some other things |
| 07:30 | <gsnedders> | annevk: plz comment there! |
| 07:31 | <annevk> | gsnedders: are those the kind of comments you're looking for? |
| 07:32 | <gsnedders> | annevk: yeah, or specific opinions as to how you'd arrange stuff to get the information you want from the docs homepage (i.e., "writing tests" -> "testharness.js API") |
| 07:33 | <gsnedders> | s/i.e./e.g./ |
| 07:54 | <annevk> | gsnedders: you got it |
| 08:20 | <gsnedders> | annevk: thanks, exact;y wjat I wanted! |
| 08:30 | <tobie> | gsnedders: you also need to fix the sidebar--it's all broken |
| 08:33 | <tobie> | gsnedders: glad you're picking that up. |
| 08:35 | <tobie> | gsnedders: when you refactor, you might want to redirect folks to the new pages using HTML redirects. |
| 08:35 | <tobie> | gsnedders: I've got a layout setup for that: https://github.com/w3c/testtwf-website-src/blob/master/sanfrancisco-2012.html |
| 08:36 | <tobie> | gsnedders: https://github.com/w3c/testtwf-website-src/blob/master/_layouts/redirect.html |
| 09:07 | <tobie> | gsnedders: for ref (and laughs), pre-testtwf resources: https://github.com/w3c/testtwf-website-src/blob/master/RESOURCES.md |
| 09:24 | <gsnedders> | tobie: fwiw, I think the plan is to simply kill TTWF and have all the docs redirect |
| 09:24 | <tobie> | gsnedders: yeah. That works too. :) |
| 09:25 | <gsnedders> | tobie: first off, tho, my plan is to get the content right |
| 10:00 | <tobie> | Today's magic incantation: bikeshed && git add index.html && git rebase --continue |
| 10:03 | <annevk> | tobie: those kind of things must be why folks add shortcuts here and there |
| 10:04 | <annevk> | tobie: I just repeat the instructions each time like some kind of animal |
| 10:06 | <tobie> | annevk: yeah, I'm a mildly evolved animal, as I found the arrow-up key to cycle through previous commands. |
| 10:06 | <tobie> | annevk: and && |
| 11:28 | smaug | tries to recall why stable state was changed to use microtasks |
| 11:29 | <annevk> | smaug: git log says https://www.w3.org/Bugs/Public/show_bug.cgi?id=24724 |
| 11:29 | annevk | hasn't followed the link |
| 11:31 | <smaug> | and I was against it then because of the same reason I wonder why on earth that change happened. But too late to change |
| 12:52 | <tobie> | Domenic: when you're around, could you please have a look at: https://github.com/heycam/webidl/pull/205#discussion_r86335055 |
| 12:53 | <tobie> | Domenic: thanks. :) |
| 13:37 | <beverloo> | annevk, mind if I remove `sticky` and `noscreen` from notifications? |
| 13:37 | <beverloo> | annevk, there are no implementations and I'm not aware of anyone planning to implement them either |
| 14:34 | <annevk> | beverloo: I don't mind |
| 15:08 | <Domenic> | botie: ask annevk why his dfn.js's are id=head |
| 15:08 | <botie> | will do |
| 15:13 | <botie> | annevk, at 2016-11-03 15:08 UTC, Domenic said: why his dfn.js's are id=head |
| 15:26 | <zcorpan> | maybe dfn.js should do its work using requestIdleCallback |
| 15:29 | <Domenic> | oooh interesting |
| 15:30 | <Domenic> | tobie: still NewTarget problems |
| 15:31 | <tobie> | Domenic: ty. Wasn't clear to me what Allen meant by "top-level algorithm." Fixing. |
| 15:44 | <annevk> | Domenic: I think dfn.js requires it |
| 15:45 | <Domenic> | Hmm can't see how |
| 15:46 | <annevk> | Maybe that was removed? |
| 15:46 | <Domenic> | Yeah maybe |
| 15:46 | <Domenic> | We could consider using Bikeshed's version but it has some styling issues last I checked where it will squish the contents of the box |
| 15:47 | <Domenic> | And in general (like with caniuse) it's a bit frustrating to move these things outside our control |
| 15:48 | <annevk> | Domenic: https://github.com/whatwg/resources.whatwg.org/commit/b39d40c5e77f8ce0119eadc864364f93aa4f7979 removed it |
| 16:01 | <Domenic> | Ah I see |
| 16:16 | <zcorpan> | Domenic: https://github.com/whatwg/resources.whatwg.org/pull/22 |
| 16:17 | <Domenic> | zcorpan: did you test locally? |
| 16:18 | <zcorpan> | Domenic: yes. seems to work in chrome and firefox. it invoked the callback about 40 times for me |
| 16:18 | <tobie> | Domenic, so: Let |R| be the result of performing the actions listed in the description of |constructor| with |values| as the argument values. Rethrow whaetver is raized. (or can I use exception to signify the thing that is thrown?) |
| 16:18 | <Domenic> | tobie: "Rethrow any exceptions." |
| 16:18 | zcorpan | gotta go |
| 16:18 | <Domenic> | And yes, exception = any value, as long as you don't link it to things |
| 16:18 | <tobie> | Domenic: thanks. |
| 16:19 | <Domenic> | zcorpan: seems good then; let's see what happens! |
| 20:47 | <tobie> | Domenic: I saw you were setting-up bikeshed build for the dom standard. Think you could do the same for WebIDL? |
| 20:48 | <tobie> | Domenic: I feel like I'm spending my days fixing merge conflits on index.html. |
| 20:48 | <Domenic> | tobie: sure, although won't that break your scripts that rely on rawgit? |
| 20:48 | <tobie> | That's a good point. |
| 20:49 | <tobie> | Domenic: I'll render the PR branches and stick them on AWS |
| 20:49 | <Domenic> | ok |
| 20:52 | <TabAtkins> | Wait, why are you using rawgit? |
| 21:00 | <tobie> | TabAtkins: to display the rendered PRs |
| 21:00 | <tobie> | TabAtkins: see for example here: https://github.com/heycam/webidl/pull/220#issue-187083299 |
| 21:01 | <TabAtkins> | Oh, for PR rendering. Carry on then. |
| 21:01 | <tobie> | :) |
| 22:10 | <tobie> | Domenic: do we have a list of where to file bugs against UA for WebIDL stuff? |
| 22:13 | <Domenic> | tobie: the usual places; bugzilla.mozilla.org, bugs.webkit.org, crbug.com, and https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/ |
| 22:13 | <Domenic> | tobie: I use http://browser-issue-tracker-search.appspot.com/ to make sure there are no dupes or find related bugs I can reference |
| 22:57 | <laughinghan> | so I have some very specific bidi questions |
| 22:57 | <laughinghan> | first of all, this algorithm for dir=auto https://html.spec.whatwg.org/multipage/dom.html#attr-dir-auto |
| 22:57 | <laughinghan> | is that supposed to be exactly equivalent to these style rules [1] interpreted according to [2]? |
| 22:58 | <laughinghan> | [1]: https://html.spec.whatwg.org/multipage/rendering.html#bidi-rendering |
| 22:58 | <laughinghan> | shoot no |
| 22:58 | <laughinghan> | wait no that was right |
| 22:58 | <laughinghan> | [2]: https://drafts.csswg.org/css-writing-modes/#valdef-unicode-bidi-plaintext |
| 22:58 | <laughinghan> | I may be misunderstanding, but those seem either conflicting or redundant (hopefully, redundant) |
| 22:59 | <laughinghan> | or maybe they're complementary in some way I don't understand |
| 23:00 | <laughinghan> | the reason I'm asking is I'm working on a span that pretends to be an <input type=text> |
| 23:01 | <laughinghan> | and I'm wondering if it's necessary or sufficient to have a CSS rule like span.my-special-kind-of-span[dir=auto i] { unicode-bidi: plaintext; } |
| 23:02 | <laughinghan> | if that dir=auto behavior is complementary or overrides something about that CSS unicode-bidi behavior, then that won't be sufficient, which would be unfortunate |
| 23:06 | <laughinghan> | another thing I don't understand is that, from my understanding (and according to this plain-English, non-spec explanation [1]), all unicode-bidi:plaintext does differently from unicode-bidi:isolate has to do with if there are multiple paragraphs, so why do the rules above set it on <input type=text>s? |
| 23:06 | <laughinghan> | [1]: https://www.w3.org/International/questions/qa-html-dir#autotextarea |
| 23:18 | <tobie> | Domenic: I had forgotten the existence of browser-issue-tracker-search.appspot.com. So amazing. |