| 01:36 | <kochi> | annevk: thanks for everything. I was not well available last week due to private matters, but will be doing more stuff for the f2f this week and the next week. |
| 01:56 | <kochi> | annevk: are you coming to Tokyo, or if not, could you join VC remotely? |
| 03:25 | <smccandlish> | Any of the autoconfirmed around? Need to set up wiki.whatwg.org account |
| 03:49 | <smccandlish> | Domenic, you around? |
| 03:51 | <smccandlish> | GPHemsley: What about you? I'm going down the autoconfirmed list. Need help with getting wiki.whatwg.org account. |
| 03:53 | <smccandlish> | Is anyone active at all right now? |
| 04:00 | <MikeSmith> | smccandlish: I can create an account for you |
| 04:00 | <smccandlish> | Yay! |
| 04:00 | <smccandlish> | I requested one via e-mail a long time ago, but never saw a response, so this seems the way to go. |
| 04:00 | <MikeSmith> | yeah |
| 04:01 | <smccandlish> | What will you need from me? |
| 04:01 | <MikeSmith> | I get those e-mail messages but ignore them at this point |
| 04:01 | <smccandlish> | Too much spam? |
| 04:01 | <MikeSmith> | no, just that I don’t want to encourage people to keep using that wiki |
| 04:01 | <smccandlish> | Is a replacement in the works? |
| 04:02 | <MikeSmith> | well I personally think we should take all the current content there and move it to a github wiki |
| 04:02 | <MikeSmith> | but others disagree |
| 04:03 | <MikeSmith> | anyway, to set up an account for you, I need an e-mail address and your preferred username |
| 04:03 | <MikeSmith> | feel free to PM them if you want |
| 04:05 | <smccandlish> | That sounds reasonable, since people could use their existing accounts. Anyway, I'm a pretty public figure, so no secrecy needed. UserID: SMcCandlish, e-mail smccandlish⊙gc |
| 04:06 | <MikeSmith> | OK, gimme a minute, doing it right now |
| 04:06 | <smccandlish> | Thanks much. |
| 04:07 | <MikeSmith> | hmm, it says “username already in use”？ |
| 04:08 | <smccandlish> | ?? Hmm. I don't have a PW on file for that site. I can try stabbing guesses. |
| 04:09 | <MikeSmith> | maybe I can have it send you a reminder |
| 04:09 | <MikeSmith> | oh, you can yourself I think |
| 04:09 | <MikeSmith> | https://wiki.whatwg.org/wiki/Special:PasswordReset |
| 04:09 | <smccandlish> | Yeah, I just clicked on that. I had no idea the account already existed, and now feel very silly. Sorry for wasting your time. |
| 04:09 | <MikeSmith> | no worries |
| 04:11 | <smccandlish> | I'll go look for the reset. Have a good one. :-) |
| 04:12 | <MikeSmith> | cheers |
| 04:30 | <kochi> | MikeSmith: do you think if we have remote participants for the coming F2F at Google Tokyo, they can connect using Google Hangouts? |
| 04:30 | <MikeSmith> | kochi: yes |
| 04:30 | <kochi> | (depends on who they are, of course, but is "webex" required thing?) |
| 04:30 | <MikeSmith> | no need to set up anything else |
| 04:30 | <MikeSmith> | no, webex is not required |
| 04:30 | <MikeSmith> | it is better to avoid it, actually |
| 04:30 | <kochi> | ok, cool, then not setup is needed, then. |
| 04:30 | <MikeSmith> | hai |
| 04:31 | <kochi> | s/not setup/no setup/ |
| 06:07 | <annevk> | kochi: I land the Sunday before and depart Friday |
| 06:19 | <kochi> | annevk: cool. In the w3c meetings page https://github.com/w3c/WebPlatformWG/blob/gh-pages/meetings/18-03-Web-components.md I didn't find your name and was wondering if you could not make it. |
| 06:21 | <kochi> | we're trying to make the logistic as lightweight as possible, but in any case you need a blue security badge for entering the building. we are figuring out if name/company are required for passing those badges to guests. |
| 06:25 | <annevk> | kochi: I guess I should update that |
| 06:26 | <kochi> | annevk: thanks. we expect there are last-minute updates on the list, so no hurries. |
| 06:31 | <annevk> | kochi: added |
| 06:31 | <kochi> | annevk: thanks!! |
| 09:32 | <annevk> | MikeSmith: I didn't realize that currently if you use <h1> followed by <h3> there's no conformance error |
| 09:33 | <annevk> | MikeSmith: so this outline change will definitely affect the validator |
| 10:04 | <MikeSmith> | yeah |
| 10:05 | <MikeSmith> | which is fine as far as I’m concerned |
| 10:05 | <MikeSmith> | happy to see that be made an error |
| 10:07 | <annevk> | It used to be an error too I think in HTML4 |
| 10:07 | <annevk> | Insofar that required more than a quoted desk |
| 10:07 | <annevk> | (I'm probably messing up that joke) |
| 10:09 | <MikeSmith> | it was never an error in HTML5 as far as I can recall |
| 10:10 | <MikeSmith> | the (second) outline view in the checker exposes it as problem |
| 10:10 | <MikeSmith> | by design |
| 10:10 | <MikeSmith> | by showing an empty placeholder for any heading levels that are missing in the hierarchy |
| 10:11 | <MikeSmith> | oh I see that’s the first outline view now |
| 10:11 | <MikeSmith> | I guess I reversed the order a while back |
| 10:12 | <MikeSmith> | I guess at this point I really should just remove the (now-second) one, the “Structural outline” (generated from the outline algorithm) |
| 11:40 | <nox> | HTML4? What year is this |
| 12:32 | <annevk> | kochi: if you think there's other things worth categorizing with the existing issues, let me know |
| 12:32 | <annevk> | kochi: I think I've mostly done my bit, except for working on fixes |
| 13:14 | <Domenic> | annevk: re https://github.com/whatwg/fetch/issues/505#issuecomment-366671504, in what way did HTTP start dealing with fragments? I hadn't heard. |
| 13:17 | <annevk> | Domenic: they made it more explicit about how they work with redirects |
| 13:17 | <annevk> | Domenic: so that kinda means the fragment has to go into the HTTP subsystem |
| 13:20 | <Domenic> | Huh interesting |
| 13:20 | <Domenic> | Probably a good thing |
| 13:26 | <annevk> | I kinda like it actually, it's just a little risky at this point due to me not fully considering it early on |
| 13:26 | <annevk> | And nobody reviewing noticing I suppos |
| 13:26 | <annevk> | e |
| 13:44 | <annevk> | smaug____: are you around? I want to ask about the strategy for https://github.com/whatwg/html/issues/1127 |
| 13:45 | <annevk> | smaug____: do we basically run the insertion steps twice for DocumentFragment with some kind of flag that indicates when it's safe to execute scripts or do we just special case script elements or something else? |
| 13:46 | <annevk> | smaug____: I'm inclined in the specification to special case script elements, unless there's other things we need to run the steps twice for |
| 13:46 | <annevk> | smaug____: nm, that doen't work |
| 14:45 | <nox> | annevk: Did the mixin stuff etc land in our WebIDL parser? |
| 14:45 | <nox> | Wrong channel. |
| 15:13 | annevk | wonders if esprehn is still planning to get around to making mutation events "async" |
| 17:08 | <wanderview> | is there an up-to-date draft svg spec? |
| 17:10 | <annevk> | wanderview: https://svgwg.org/svg2-draft/ maybe? |
| 17:12 | <annevk> | wanderview: https://github.com/w3c/svgwg/commits/master does not look like it's very active though |
| 17:12 | <wanderview> | annevk: thanks, mainly looking for any kind of processing model for fetching stuff |
| 17:12 | <annevk> | wanderview: kinda doubt that exists 😕 |
| 17:13 | <annevk> | wanderview: SVG rendered as image is somewhat defined at https://svgwg.org/specs/integration/ I think, but not really |
| 17:14 | <annevk> | wanderview: at some point I drafted together some notes for what SVG needs, but nobody is really available to do the work |
| 17:14 | <wanderview> | annevk: just trying to figure out the fragment stuff |
| 17:14 | <wanderview> | annevk: or if anything else should keep the fragment |
| 17:47 | <railgun> | annevk: fixed #3473 |
| 17:59 | <annevk> | railgun: cool, will look tomorrow |
| 18:00 | <annevk> | wanderview: per HTTP it'd be best if everything preserved it |
| 18:00 | <annevk> | wanderview: I recall there also being some proposals for fragments to work on media resources (video, audio) so that seems sensible in retrospect |
| 18:00 | <wanderview> | annevk: the question I have is... when the Response.url has a fragment that the original Request.url did not have... is that possible with service worker? |
| 18:01 | <annevk> | wanderview: it's possible with redirects and service workers, I think |
| 18:01 | <wanderview> | annevk: should we just take the fragment off of the Request.url and attach it to the Response.url in HandleFetch? |
| 18:01 | <wanderview> | hmm, ok |
| 18:01 | <annevk> | wanderview: we wanna forward, but explicit response fragment wins |
| 18:02 | <annevk> | wanderview: see some of the tests I wrote |
| 18:02 | <annevk> | wanderview: they don't cover service workers yet though |
| 18:02 | <wanderview> | annevk: so service worker HandleFetch should forward the fragment? |
| 18:02 | <wanderview> | ok |
| 18:02 | <wanderview> | annevk: I think that would solve this particular problem that xidorn has |
| 18:02 | <annevk> | wanderview: I need to figure out a local HTTPS setup for WPT |
| 18:02 | <annevk> | wanderview: yeah it should |
| 18:03 | <wanderview> | annevk: it would also solve it without having to figure out the Response.url fragment situation |
| 18:03 | <annevk> | wanderview: if we just kept it internal you mean? |
| 18:03 | <annevk> | wanderview: sure, that'd work, it just doesn't give the same power to apps |
| 18:04 | <wanderview> | annevk: well, we have the fragment on FetchEvent.request.url... we could just apply it to whatever the Response.url is right now |
| 18:04 | <astearns> | annevk: When I last talked to esprehn he had a new, non-Blink (and non-mutation-event) project he was excited about. |
| 18:04 | <annevk> | wanderview: what about redirects? |
| 18:04 | <wanderview> | annevk: I'm not saying we stop there... but it solves the immediate compat problem xidorn raised while we figure out the full picture |
| 18:05 | <wanderview> | annevk: and its easy for me to do this without touching Response.url yet |
| 18:05 | <annevk> | astearns: hmm good for him I guess |
| 18:06 | <annevk> | wanderview: sure, intermediate fixes seem fine, esp since Chrome/Safari already work, right? |
| 18:07 | <wanderview> | annevk: I don't think they are doing the Response.url propagation stuff yet in this case, so not sure they have run into the problem yet |
| 18:09 | <wanderview> | annevk: maybe I should file a separate spec issue on applying the Request.url fragment if present when ServiceWorker does an interception |
| 18:12 | <annevk> | wanderview: I think we should tackle it all at once |
| 18:13 | <annevk> | wanderview: btw, for your intermediate solution, if request.url contains a fragment, response.url wins? |
| 18:13 | <annevk> | wanderview: but response.url only wins if it also has a fragment? |
| 18:13 | <wanderview> | not sure I understand |
| 18:14 | <annevk> | wanderview: if you request /hi#bye and you get /hi#yo |
| 18:14 | <annevk> | wanderview: it needs to be /hi#yo, right? |
| 18:14 | <wanderview> | annevk: in the interim you cannot get a fragment on Response.url (and we don't save it internally) |
| 18:14 | <wanderview> | annevk: you will always get /hi right now |
| 18:14 | <annevk> | I see, okay |
| 18:15 | <wanderview> | annevk: saving it internally means also storing it on disk in Cache API, etc |
| 18:15 | <wanderview> | so we haven't done that |
| 21:23 | <esprehn> | annevk: I still think it's worth doing but I'm not working on Blink/Chrome at this time. I'm not sure who works on the DOM these days, maybe ask TabAtkins or ikilpatrick (who's unfortunately not in this channel) |
| 21:29 | <astearns> | maybe iank___ ? |
| 21:54 | <smaug____> | annevk: (traveled to Toronto today) |
| 21:55 | <smaug____> | I'll look at the question later today or on Wed |