| 00:00 | <Domenic> | Yeah, I'm done replying to you. |
| 00:00 | <Domenic> | (here and on thread) |
| 00:00 | <deyomax> | Domenic: that's not very professional. |
| 00:01 | <deyomax> | Domenic: I personally don't have a problem, I've documented this in my codebase, it's not for me that I filed this issue. |
| 00:05 | <deyomax> | I'm starting to understand Daniel Stenberg's annoyance, Domenic https://github.com/whatwg/url/issues/118#issuecomment-217866196 |
| 00:10 | <jyasskin> | deyomax: Professionalism doesn't mean someone owes you unlimited amounts of time. |
| 00:11 | <deyomax> | jyasskin: I didn't need any of his time. |
| 00:11 | <deyomax> | But currently the conclusion on this thread on GitHub is misleading. |
| 00:14 | <deyomax> | ¯\_(ツ)_/¯ |
| 01:26 | <Domenic> | I wonder why inside <picture>, <source> is only allowed to have srcset, not src. |
| 03:08 | <jyasskin> | MikeSmith: That sounds like definitely a bug in Chrome. |
| 03:16 | <domfarolino> | Hm am I missing something? The MutationObserver.observe function has an optional MutationObserverInit object, yet when you leave it out the console yells at you and can't execute the function |
| 03:19 | <MikeSmith> | jyasskin: yeah I’m doing some crude hacking on sources now to see if I can isolate where it actually errors |
| 03:19 | <MikeSmith> | and lord why does it take Chromium so long to link |
| 03:20 | <MikeSmith> | after a one-line change, 5.5 minutes to 6 minutes mostly just spent linking |
| 03:29 | <Domenic> | domfarolino: so this is an instance where Web IDL is not expressive enough (currently). A dictionary with no required members must be marked optional per Web IDL. MutationObserverInit has no required members... technically. At least one member must be present, but no single member is marked as required (i.e. always required). |
| 04:00 | <domfarolino> | Domenic: ahh ok. Interesting. Thanks for that. So the logic of the API requires at least one to be present, but since we don't actually need any one of these to be "required"...we see a little oddity? |
| 04:00 | <Domenic> | Yeah |
| 04:01 | <Domenic> | People have suggested not requiring that such dictionaries be marked optional, but my opinion is that hurts the 99% case |
| 04:02 | <Domenic> | (i.e. dictionaries with no formally-required members really should be marked optional, and it's best if tools complain about that if you forget) |
| 04:03 | <domfarolino> | Domenic: Gotcha. Hm so is there presently work being done on WebIDL to get around this (you alluded to this with "currently"?). Like expressing in the logic of an API that even though there are no required members, there are requirements present that must mandate its presence anyways? |
| 04:04 | <Domenic> | Not really, just debate about maybe removing the requirement about optional. I think it's so small of a case and it's really about confusing documentation, not confusing processing model, so it's IMO not worth it |
| 04:13 | <domfarolino> | Domenic: Ok that makes sense. Yeah I assume if we forced dictionaries with no formally-required members to be not optional that would probably what, lead to a bunch of apis looking like this: `something('somethingElse', {})` all the time? |
| 04:13 | <domfarolino> | where we'd rather not pass in that empty object or something? (when it has to members) |
| 04:13 | <Domenic> | Maybe not a bunch, but some, where the spec editors didn't realize their mistake and roaming reviewers didn't stop by to correct them, yeah. |
| 04:15 | <domfarolino> | Ah yes. |
| 04:24 | <domfarolino> | Is there a reason there is not a mutation-observer directory in wpt? Has nobody gotten to it yet or something? |
| 04:24 | <Domenic> | It'd be under dom |
| 04:30 | <domfarolino> | ah |
| 04:34 | <MikeSmith> | is foreign fetch dead? |
| 04:35 | <MikeSmith> | specifically is https://github.com/w3c/ServiceWorker/issues/1188 basically already decided? |
| 04:53 | <Domenic> | That is my impression, just from reading public stuff |
| 05:06 | <MikeSmith> | OK |
| 06:54 | <annevk> | MikeSmith: yup, is dead |
| 06:55 | <annevk> | MikeSmith: is not compatible with double keying cookies and storage, which browsers don't all do now, but would be better so is a goal |
| 07:26 | <zcorpan> | Anyone want to review https://github.com/w3c/web-platform-tests/pull/6982 Update IDL tests for CSSOM and CSSOM View |
| 07:35 | <MikeSmith> | annevk: thanks |
| 07:36 | <MikeSmith> | annevk: so, given that though, it seems there is the matter of open browser implementation bugs to at least post comments to |
| 07:36 | <MikeSmith> | e.g., the Chrome feature bug for it |
| 07:37 | <MikeSmith> | not sure if there are also bugs for other UAs that have been opened, but I know there’s at least the Chrome one |
| 07:50 | <annevk> | MikeSmith: yeah I suppose, service workers issue is not closed yet |
| 08:14 | <MikeSmith> | annevk: I guess I meant specifically https://bugs.chromium.org/p/chromium/issues/detail?id=540509 |
| 08:14 | <MikeSmith> | as far as posting comment or changing the status of that issue |
| 08:15 | <MikeSmith> | or maybe that’s something that yhirano__ himself should be the one to take action one |
| 08:15 | <MikeSmith> | *on |
| 08:16 | <MikeSmith> | or kochi |
| 08:16 | <MikeSmith> | or Mek :) |
| 08:20 | <kochi> | I'm not working on ServiceWorker or fetch :) |
| 08:21 | <kochi> | ah, on that bug I just removed false accidental dependency |
| 08:24 | <MikeSmith> | ah ok |
| 08:24 | <MikeSmith> | oh Kenji is the service workers PM? |
| 08:25 | MikeSmith | tries to remember, and tries to re-visualize the faces around the table at the f2f SW meeting we had at Facebook on Tokyo |
| 09:14 | <yhirano__> | I also am not working on foreign fetch. |
| 09:39 | <MikeSmith> | yeah, sorry I think when I pinged y'all I was looking at two different bugs (the foreign-fetch one and another one) and I conflated the two |
| 09:40 | <MikeSmith> | holding a sleeping baby in one arm while browsing chrome bugs and typing with one hand is a bit error-prone it seems |
| 11:53 | <annevk> | JakeA: FYI, I'm editing https://github.com/whatwg/html/pull/3038 |
| 11:58 | <ondras> | are css custom properties supposed to work with generated content? |
| 11:58 | <ondras> | as in, "content: var(--mystuff)" ? |
| 12:05 | <gsnedders> | ondras: yes |
| 12:05 | <gsnedders> | ondras: 'The var() function can be used in place of any part of a value in any property on an element.' |
| 12:06 | <ondras> | gsnedders: thanks. it was not working for me, but apparently because I was using *numeric* custom properties. |
| 12:06 | <ondras> | gsnedders: strings are working just fine. |
| 12:12 | <ondras> | gsnedders: it still behaves in a weird way: https://jsfiddle.net/z83ug7dd/ |
| 12:13 | <ondras> | (both firefox and chrome show nothing when modified by JS) |
| 12:15 | gsnedders | shrugs :) |
| 12:15 | <hasather> | ondras: document.body.style.setProperty("--test", "'clicked'"); |
| 12:15 | <hasather> | ondras: i.e., it should be a string |
| 12:16 | <ondras> | hasather: thanks a lot. makes sense. |
| 12:16 | <gsnedders> | hasather: shouldn't it be converted at the WebIDL level? |
| 12:16 | <hasather> | gsnedders: you probably know that better than me :) No idea |
| 12:17 | <gsnedders> | hasather: (what are you even doing nowadays? you were at finn, no?) |
| 12:17 | <hasather> | gsnedders: still there |
| 12:19 | <ondras> | I see. It works with strings, but these cannot be then used in calc() expressions |
| 12:19 | <ondras> | dammit. |
| 12:20 | <JakeA> | annevk: cheers! I don't have anything else to push |
| 12:20 | <ondras> | my case: passing a current/total number of slides to be #1 displayed in a footer, #2 making the footer's width calc(current/total)) |
| 12:20 | <ondras> | apparently not possible with a single css custom property |
| 12:23 | <Ms2ger> | gsnedders, I'm guessing it's parsed as css |
| 12:29 | <gsnedders> | Ms2ger: oooh, that'd make sens |
| 12:29 | <Ms2ger> | Yeah, so that's probably not it :) |
| 12:30 | <gsnedders> | idk, it seems the easiest way to explain behaviour :) |
| 12:50 | <JakeA> | annevk: are you adding that "white true" note or shall I? |
| 12:51 | <annevk> | JakeA: feel free, I was gonna wait for another round from Domenic before addressing more nits |
| 13:44 | <wanderview> | JakeA: annevk: did anything change much in the abortable fetch in the last round of reviews? |
| 13:44 | <wanderview> | AFAIK we are planning to ship in FF57 |
| 13:45 | <JakeA> | I don't think so, it was mostly editorial |
| 13:49 | <wanderview> | JakeA: any word on when chrome will implement? |
| 13:50 | <annevk> | wanderview: not sure when we'll ship tbh |
| 13:50 | <annevk> | wanderview: did it land? |
| 13:50 | <annevk> | wanderview: and yeah, editorial shuffling only |
| 13:50 | <wanderview> | annevk: it did land and is pref'd on... baku sent mail to dev-platform |
| 13:51 | <annevk> | 🎉 |
| 13:51 | <annevk> | Missed/forgot that |
| 13:51 | <wanderview> | I find it interesting that chrome is implementing fetch keepalive before aborting |
| 13:51 | <annevk> | Many unrelated things going on |
| 13:53 | <JakeA> | wanderview: haven't heard anything in terms of implementation, but I'll give them a nudge. Also I'll write about it when it's in Firefox beta unflagged, or is that already the case? |
| 13:53 | <wanderview> | is fetch keepalive high priority for some particular use case? |
| 13:53 | <wanderview> | JakeA: it will be in beta in a couple weeks |
| 13:54 | <JakeA> | wanderview: I'll get an article out on https://developers.google.com/web/updates/ |
| 13:57 | <wanderview> | JakeA: thanks... in theory MDN will be updated in month or so |
| 13:58 | <JakeA> | (I kinda like the idea of shouting about a new Firefox feature on Google web updates) |
| 13:59 | <wanderview> | JakeA: well, edge implemented first! |
| 13:59 | <JakeA> | wanderview: true, but it's first to release that counts :D |
| 14:07 | <wanderview> | JakeA: also, is there a still compelling use case for link-element service worker registration? I'm not sure we will implement without foreign fetch |
| 14:10 | <JakeA> | wanderview: some folks like the declarative thing, but yeah, I don't think there's a strong case for it |
| 14:11 | <wanderview> | JakeA: I guess we should find out if edge or webkit are going to implement it |
| 16:20 | <annevk> | wanderview: I don't think we should do it |
| 16:21 | <annevk> | wanderview: I think we should wait for "origin policy" and if that's really fast somehow see if we can link service workers from there, which would make them available from first run... |
| 16:21 | <annevk> | wanderview: just lazily fetching the primary service worker doesn't seem like enough of a benefit |
| 18:17 | <foolip> | annevk: something's wrong with the header of https://dom.spec.whatwg.org/ |
| 18:17 | <foolip> | just opening (, maybe a bikeshed regression? |
| 18:20 | <wanderview> | JakeA: do you know who might work on the ft.com service worker? |
| 18:21 | <annevk> | foolip: yeah must be I think since I don't think anyone touched that recently, TabAtkins ^^ |
| 18:21 | <foolip> | annevk: it does say Sep 14 though |
| 18:22 | <TabAtkins> | ...huh. I absolutely tested the scenario that's failing here. |
| 18:22 | <TabAtkins> | Sorry about that, will fix immediately. |
| 18:22 | <annevk> | foolip: I mean nobody touched the header recently, but DOM itself was updated today |
| 18:22 | <annevk> | TabAtkins: thanks |
| 18:28 | <TabAtkins> | ...somehow, the precise thing I tested works perfectly, but almost any variant I'm trying breaks. wtaf |
| 18:29 | <wanderview> | JakeA: nm, I'm spamming twitter |
| 18:54 | <TabAtkins> | annevk: Sorry, fix pushed now. (I was accidentally doing something different for custom metadata specified in Metadata Order and ones that weren't, and my testcase was only doing the one. It's now testing both.) |
| 22:43 | <TimothyGu> | MikeSmith: thanks :). Sorry about the non-reply, I was (and am) doing my university orientation |
| 23:00 | <MikeSmith> | TimothyGu: no worries about reply latency, I am used to that here :) |
| 23:01 | <MikeSmith> | and also in general thanks for all you have been contributing to the WHATWG work over the previous months |
| 23:01 | <MikeSmith> | super glad to have you involved |
| 23:03 | <MikeSmith> | from what I have seen, you bring in a ton of insight toward making some important improvements in various places |