| 01:23 | <Domenic> | realm craziness https://github.com/w3c/battery/issues/2 |
| 02:41 | <MikeSmith> | wow |
| 05:03 | <annevk> | Domenic: prolly just poorly reviewed and not implemented that way |
| 05:03 | <terinjokes> | Domenic: 👏 |
| 05:03 | <annevk> | Domenic: unclear how it is implemented though… |
| 05:04 | <annevk> | krijn: any ETA on log migration? |
| 05:49 | <krijn> | Will try this week, but very busy |
| 07:36 | <annevk> | krijn: ta, understood |
| 08:58 | <zcorpan> | Error: missing <dfn> for topic "math:mi" explicitly from <code> element containing "mi"; previous heading contents are "12.2.3.2 The stack of open elements" |
| 08:58 | <zcorpan> | annevk: do you get that also? |
| 08:59 | <annevk> | zcorpan: you might need to do git pull, I think Domenic fixed that in a follow up commit |
| 08:59 | <annevk> | zcorpan: your menuitem change was incorrectly rebased I suppose |
| 09:00 | <zcorpan> | i see |
| 09:52 | <Ms2ger> | annevk, is there a non-historical reason that "origin" is defined in HTML rather than URL? |
| 09:54 | <annevk> | Ms2ger: it seems a bit bigger than URLs, since many things can have them, dunno |
| 11:05 | <zcorpan> | MikeSmith: thoughts on checking lang against language analysis in the html checker? https://www.w3.org/Bugs/Public/show_bug.cgi?id=26942#c13 |
| 11:23 | MikeSmith | looks at zcorpan link |
| 11:26 | <MikeSmith> | zcorpan: https://github.com/shuyo/language-detection |
| 11:27 | <MikeSmith> | 「99% accuracy for 53 languages」 |
| 11:27 | <MikeSmith> | (if that is what you had in mind) |
| 11:28 | <MikeSmith> | would be a fun thing to implement |
| 11:29 | <MikeSmith> | not sure how to do it without just collecting all the text content of the entire document |
| 11:31 | <MikeSmith> | or I guess I wouldn’t need to save/store the text content but instead just check each text node as the parser exposes it |
| 12:02 | <smaug____> | what is sicking's github handle |
| 12:06 | <Ms2ger> | https://github.com/sicking |
| 12:33 | <zcorpan> | MikeSmith: yeah |
| 13:18 | <SimonSapin> | annevk: what do you think? https://github.com/whatwg/url/issues/104 |
| 13:20 | <annevk> | SimonSapin: wouldn't that allow a URL without a host to get a special scheme? |
| 13:36 | <SimonSapin> | annevk: yes, that’s a case that I forgot and should be prevented |
| 13:36 | <SimonSapin> | still, nothing wrong going between http://example.com/ and view-source+http://example.com/ |
| 13:37 | <annevk> | I think that might be blocked for Location actually |
| 13:37 | <SimonSapin> | but for URL objects? |
| 13:37 | <annevk> | But that might require some further fixes |
| 13:38 | <annevk> | Can we make that distinction easily? |
| 13:38 | <SimonSapin> | annevk: now that URLUtils is gone and everything defines their own getters and setters yes |
| 13:39 | <annevk> | Well, they all still use the same parser |
| 13:39 | <annevk> | But sure, everything is possible |
| 13:49 | <annevk> | SimonSapin: anyway, prolly reasonable to fix in some way, just need to find a bit of time |
| 13:49 | <annevk> | SimonSapin: if you are blocking on me I'll try to get to it by Friday |
| 13:51 | <SimonSapin> | annevk: I’m writting some tests for setters and need to write them one way or another, but I can change them later |
| 13:51 | <SimonSapin> | I’ll submit them to wpt |
| 13:59 | <Domenic> | yessss setter tests yessss |
| 13:59 | <Domenic> | need those badly |
| 13:59 | <annevk> | Sorry for the delay in fixes, between Shadow DOM, whatwg/html, and this presentation I need to finish I'm suddenly strap for time |
| 14:00 | <annevk> | SimonSapin: if you ever want to start editing again 😊 |
| 14:02 | <SimonSapin> | Domenic: want to write a JS harness for https://github.com/servo/rust-url/blob/1.0/tests/setters_tests.json ? |
| 14:02 | <Domenic> | SimonSapin: not if you're going to do it first! :) |
| 14:03 | <SimonSapin> | Domenic: testing my rust implementation for now |
| 14:03 | <zcorpan> | hsivonen: fyi I get a 502 again for *.validator.nu |
| 14:08 | <annevk> | zcorpan: https://checker.html5.org/ |
| 14:08 | <zcorpan> | annevk: thx, though i actually wanted to use https://parsetree.validator.nu/ :-) |
| 14:08 | <annevk> | aah, maybe MikeSmith should add that to checker.html5.org somewhere |
| 14:10 | <SimonSapin> | annevk: no worries re delays. I don’t mind sending PRs for edits once they’re decided, but there’s plenty of corner cases where I don’t know if there a reason if something is the way it is or if it’s a copy/paste mistake or something |
| 14:11 | Ms2ger | unsubscribed from whatwg/{html,dom} |
| 14:12 | <Domenic> | aww |
| 14:15 | <Ms2ger> | Feel free to cc me if you think I'd be interested, but y'all spam too much :) |
| 14:15 | <Ms2ger> | Especially with github's one-email-per-review-comment crap |
| 14:16 | <annevk> | Enabling email for GitHub is beginner's mode |
| 14:16 | <annevk> | But you'll be paged as appropriate |
| 14:17 | <SimonSapin> | subscribing to repos is a terrible idea http://i.imgur.com/uVDIXm4.png |
| 14:17 | <Domenic> | somehow it works for me. but i guess it is my job. |
| 14:18 | <Domenic> | i unsubscribe from individual threads aggressively, but i haven't done that on html |
| 14:18 | <SimonSapin> | ("Mentions" these days is header :is ["X-GitHub-Reason"] ["manual", "author", "comment", "mention", "assign"]) |
| 14:19 | <annevk> | This is why email seems so terrible for this, if there was a long discussion, you'll have 25 emails, just one notification though |
| 14:23 | <MikeSmith> | zcorpan: https://checker.html5.org/parsetree/ |
| 14:29 | <zcorpan> | MikeSmith: ty! |
| 14:43 | <Ms2ger> | Oh look: https://www.w3.org/TR/2016/WD-CSS22-20160412/ |
| 15:12 | <Ms2ger> | annevk, do you know of any spec that defines onfoo idl attributes well? |
| 15:23 | <annevk> | Ms2ger: not for nodes or global |
| 15:24 | <Ms2ger> | This is for https://github.com/w3c/page-visibility/issues/22 |
| 15:52 | <hsivonen> | Validator.nu is back. The server had got so stuck that the watchdog didn't work and I couldn't log in, either. |
| 16:24 | <zcorpan> | what is the right design for https://lists.w3.org/Archives/Public/www-style/2016Apr/0171.html ? Domenic? |
| 16:25 | <Domenic> | zcorpan: uhhhhh definitely not an event listener that fires when you add it |
| 16:26 | <zcorpan> | right |
| 16:26 | <Domenic> | zcorpan: promises are appropriate when you care about the *next* change, not all future ones. not sure if that's tehe case here |
| 16:26 | <Domenic> | I haven't backed up to read the whole thread yet but I'd guess property + event? |
| 16:27 | <zcorpan> | the problem with property is that it could either be out of date or it would flush layout |
| 16:31 | <Ms2ger> | Fire along with DOMContentLoaded? |
| 16:32 | <Domenic> | maybe make it a method so it's more obvious it flushes layout? |
| 16:55 | <smaug____> | does anyone recall why we decided (I assume we decided it at some point) why shadow scripts run scripts in global scope? |
| 19:15 | <annevk> | smaug____: I think other engines do not have much options here, we did discuss the various levels of encapsulation |
| 19:16 | <annevk> | smaug____: isolation is some kind of v2 concept some want, though processing model is unclear |
| 19:16 | <smaug____> | well, I wasn't thinking about encapsulation here |
| 19:16 | <smaug____> | nor isolation |
| 19:16 | <smaug____> | but about the scope |
| 19:17 | <annevk> | Hmm well without a global there is not much API |
| 19:17 | <annevk> | But maybe I do not understand what you are thinking off |
| 19:18 | <smaug____> | there are different options |
| 19:18 | <smaug____> | like, 'this' could point to the shadow <script> |
| 19:19 | <smaug____> | similar to event handler I guess |
| 19:20 | <benjamingr__> | I have a question: `var embed = document.createElement("embed"); embed()` - this is correct behavior according to the spec, but I'm not sure why - or how I can utilize it or why it's there - if it's some legacy reason is there any reference about why it's there somewhere? |
| 19:22 | <jyasskin> | Ms2ger: Is https://webbluetoothcg.github.io/web-bluetooth/#idl-event-handlers a reasonable way to define the onfoo idl attributes? |
| 20:21 | <Domenic> | benjamingr__: because websites depend on it? |
| 21:12 | <TabAtkins> | jyasskin: I think so, yes. |
| 21:12 | <jyasskin> | TabAtkins: Thx. I sent it to https://github.com/w3c/page-visibility/issues/22. |
| 21:19 | <benjamingr__> | Domenic: yes, but I'm wondering why? What plugin utilized this so that websites would depend on it? |
| 21:20 | <Domenic> | benjamingr__: I'd guess Java or Flash. |
| 21:21 | <benjamingr__> | I tried both and couldn't find info for either about how to control what `()` does on the embed element. Trying it for some random flash files and java applets didn't produce anything meaningful in Chrome or IE9, maybe I'll check earlier |
| 22:05 | <jyasskin> | TabAtkins: Can you ingest https://encoding.spec.whatwg.org/ into Shepherd? |
| 23:55 | <jyasskin> | TabAtkins: Also ping on adding https://w3c.github.io/permissions/ |
| 23:57 | <TabAtkins> | Permissions is easy. Encoding won't have any of their definitions exported, so you'll have to provide a `spec` argument for all the links (or set them up in link-defaults) - is that ok? |
| 23:58 | <jyasskin> | TabAtkins: I've sent https://github.com/whatwg/encoding/pull/43 to export the definitions. |
| 23:58 | <TabAtkins> | Ah, cool. |
| 23:59 | <jyasskin> | Thanks! |
