| 00:39 | <yhirano__> | domfarolino: yt? |
| 00:39 | <domfarolino> | yhirano__: I am |
| 00:40 | <yhirano__> | domfarolino: what do you think about having a VC about priority API work? Or if you are coming to BlinkOn we can chat there. |
| 00:47 | <domfarolino> | yhirano__: VC sounds great! Don't think I will be able to make BlinkOn (finals week is right around there for me + also I had no idea it was open to anyone?) |
| 01:07 | <domfarolino> | yhirano__: Let me know when works for you, next monday or tuesday sound alright? |
| 01:09 | <yhirano__> | domfarolino: Thanks, I'll send an e-mail. I'll invite Yoav and a few others too. |
| 01:10 | <domfarolino> | yhirano__ Kinuko might be nice to include too, she seems to know about how prioritized requests are queued and I have a few questions that I haven't gotten to writing out yet |
| 01:11 | <domfarolino> | mostly about how effective Priority Hints can really be in Chrome right now by *just* adjusting computed priorities |
| 06:09 | <MikeSmith> | JakeA: https://stackoverflow.com/questions/49674193/cache-control-directive-only-if-cached-changed-by-dev-tools |
| 06:10 | <JakeA> | MikeSmith: I'll reply. Cheers! |
| 06:12 | <MikeSmith> | thanks~ |
| 09:19 | <nox> | foolip: Ping? |
| 09:19 | <nox> | Something weird happened to all IDL tests. |
| 09:20 | <nox> | Now I have blanket "Untitled" and "IDL check of WebCrypto" failures. |
| 09:23 | <nox> | https://github.com/servo/servo/pull/20569/files#diff-2167d5fa6196ac128d0d626f87e20c8c |
| 09:23 | <nox> | Err, https://github.com/servo/servo/pull/20569/files#diff-a7b0bfde18cbf33a664497d0241111cd |
| 09:23 | <nox> | all those failures seem spurious to me, |
| 09:24 | <nox> | what's the point of having a failure for the entire suite when there are already the subtest failures that are more informative? |
| 09:29 | <nox> | https://github.com/w3c/web-platform-tests/issues/10343 |
| 09:38 | <Ms2ger> | nox, did you find the code change? |
| 09:38 | <nox> | Ms2ger: There are many. |
| 09:39 | <nox> | Ms2ger: E.g. https://github.com/w3c/web-platform-tests/commit/13effc35311c62c60df4c2f34fa62b70f3947934 |
| 09:39 | <nox> | which adds a "compat interfaces." failure. |
| 09:50 | <Ms2ger> | nox, I don't understand. That's a new test, and it fails. So what? |
| 09:50 | <nox> | What's the point of this new test, especially when it's not even named? |
| 09:51 | <nox> | It will only fail if another test already failed, I just think that's a curious way to test things. |
| 09:52 | <annevk> | nox: another test? |
| 09:53 | <nox> | Oh damn, |
| 09:53 | <nox> | I think I see what happened. |
| 09:53 | <annevk> | nox: the point of that test is testing IDL interfaces, fwiw |
| 09:53 | <annevk> | which is very good(tm) |
| 09:53 | <Ms2ger> | Though it seems to be missing Global annotations |
| 09:53 | <nox> | annevk: Ok, let's back up. |
| 09:53 | <nox> | annevk: We are already running those tests in Servo, |
| 09:53 | <nox> | annevk: I know what those tests are for, |
| 09:54 | <annevk> | Ms2ger: hmm yeah |
| 09:54 | <nox> | annevk: https://github.com/servo-wpt-sync/servo/blob/68d62eb13908dd0bebd929c7eb3903138ad348fc/tests/wpt/metadata/FileAPI/idlharness.html.ini#L18 |
| 09:54 | <annevk> | Ms2ger: and Exposed |
| 09:54 | <nox> | annevk: In which case can this fail? |
| 09:54 | <Ms2ger> | annevk, that's the one I meant :) |
| 09:55 | <annevk> | nox: if you don't implement bindings correctly or if you don't implement the feature to begin with |
| 09:55 | <nox> | annevk: I mean specifically the line I pointed at. |
| 09:55 | <nox> | annevk: It's an *additional* test around the actual WebIDL tests. |
| 09:55 | <Ms2ger> | nox, pretty hard to say without the test change... |
| 09:56 | <nox> | https://github.com/w3c/web-platform-tests/commit/13effc35311c62c60df4c2f34fa62b70f3947934#diff-73ed5bdf7c7b2944893d61c4b158e3b3R15 <- This tests the WebIDL stuff, and emits a failure per missing operation/missing attribute/other IDL issue/whatever |
| 09:56 | <annevk> | nox: the line you pointed to is an ini file, which isn't part of web-platform-tests |
| 09:56 | <nox> | https://github.com/w3c/web-platform-tests/commit/13effc35311c62c60df4c2f34fa62b70f3947934#diff-73ed5bdf7c7b2944893d61c4b158e3b3R17 <- This tests that the WebIDL tests succeeded. |
| 09:56 | <nox> | Am I right? |
| 09:57 | <Ms2ger> | The outer test is mainly to catch any exceptions in the setup |
| 09:57 | <nox> | Ok, so my last question, |
| 09:57 | <annevk> | Ooh I see what you mean, yeah; that's the setup test |
| 09:57 | <nox> | if there are subtest failures, |
| 09:57 | <nox> | does the setup test fail too? |
| 09:58 | <nox> | Or is it failing in Servo because you ended up using more modern stuff in those, |
| 09:58 | <Ms2ger> | Probably not |
| 09:58 | <annevk> | That very much depends on the setup (drumroll) |
| 09:58 | <nox> | and thus we aren't actually running any of them anymore? |
| 09:58 | <Ms2ger> | You're not even running the compat test! |
| 09:58 | <nox> | That sounds super sad for Servo. :/ |
| 09:58 | <nox> | Holy fu |
| 09:58 | <nox> | So we can't update WPT anymore until we have a smup. |
| 09:58 | <nox> | ajeffrey_: ^ |
| 09:58 | <nox> | Super sad news on this sunny Friday. |
| 09:58 | <Ms2ger> | Can you start talking about tests you actually run and how they actually fail? |
| 10:01 | <nox> | What? |
| 10:03 | <nox> | I linked to https://github.com/servo/servo/pull/20569/files#diff-a7b0bfde18cbf33a664497d0241111cd at the start. |
| 10:04 | <Ms2ger> | Servo doesn't run compat/, so why do you keep pointing at a test in compat/? |
| 10:04 | <nox> | I linked to https://github.com/servo/servo/pull/20569/files#diff-a7b0bfde18cbf33a664497d0241111cd at the start. |
| 10:04 | <nox> | Which is definitely not compat/, but FileAPI. |
| 10:05 | <nox> | And that's a link to an expectation in Servo, which, as you probably know, means we are running that test. |
| 10:05 | <Ms2ger> | Okay, and where is the test change that caused it? How does it fail? |
| 10:05 | <nox> | I linked to compat/ because all those changes are similar. |
| 10:05 | <Ms2ger> | I can't find any similar changes in FileAPI |
| 10:06 | <nox> | https://github.com/w3c/web-platform-tests/commit/2f451fb5da7865cc45c7c8eb803072ae7e0649e0#diff-433cab07c3c6ec1525c907a283582cef |
| 10:06 | <Ms2ger> | > merged 1 commit into master from idl-in-interfaces 29 days ago |
| 10:06 | <Ms2ger> | That's not it |
| 10:07 | <nox> | Well, |
| 10:07 | <nox> | then I don't know where something changed, but something changed since yesterday as I mentioned earlier, |
| 10:07 | <nox> | if you don't know what changed, I don't know either. |
| 10:08 | <Ms2ger> | Can you run the test locally, and tell me what the error message is? |
| 10:08 | <annevk> | JakeA: so the CORB 206 blocking is not exactly a subset of what you propose, right? |
| 10:13 | <JakeA> | annevk: They'll block a media element reading an HTML response, which is outside of what I'm working on. Their proposal prevents Attack 4 for particular content types, but my PR blocks it for all content types. I could write up a little comparison if that would help? |
| 10:13 | <nox> | Ms2ger: Ok. |
| 10:15 | <nox> | Ms2ger: So I was entirely wrong about thinking that the test was just failing because subtests were failing, |
| 10:15 | <nox> | Ms2ger: and entirely correct when I pinged Alan with Sad News(tm). |
| 10:15 | <nox> | 0:05.99 pid:29923 ERROR 2018-04-06T10:14:24Z: script::dom::bindings::error: Error at http://web-platform.test:8000/resources/idlharness.js:687:15 missing = in const declaration |
| 10:15 | nox | cries. |
| 10:15 | <annevk> | JakeA: yeah it would; don’t want to end up blocking something twice |
| 10:16 | <Ms2ger> | nox, so it's https://github.com/w3c/web-platform-tests/pull/10285 |
| 10:17 | <Ms2ger> | nox, go ahead and make that use `var`; I'll review |
| 10:17 | <nox> | Ms2ger: Oh, ok. |
| 10:18 | <Ms2ger> | nox, well, and test in Servo first :) |
| 10:18 | <nox> | Ms2ger: Heh. |
| 10:19 | <nox> | Ms2ger: It works. |
| 10:20 | <nox> | ajeffrey_: https://github.com/w3c/web-platform-tests/issues/10343#issuecomment-379211826 <- This is scary, we really need a smup. |
| 10:20 | <Ms2ger> | So is a smup already in the goals for Q2? |
| 10:23 | <nox> | Ms2ger: It's one of the things Alan is working on, |
| 10:23 | <nox> | not even for Q2 but for his mental health in general. |
| 10:24 | <nox> | We are pretty tired of living in the past. |
| 10:24 | <Ms2ger> | Good :) |
| 10:24 | <nox> | Ms2ger: r? https://github.com/w3c/web-platform-tests/pull/10344 |
| 10:27 | <nox> | Ms2ger: I would have just kept crying if the changes had been more invasive, I'm glad it was that easy of a workaround. |
| 10:27 | <Ms2ger> | nox, r+, but don't ask me again in 2019 ;) |
| 10:27 | <nox> | Ms2ger: Yeah, totally fair. |
| 10:27 | <nox> | Ms2ger: Reminds of 2 years ago, when the same thing happened but this time because we didn't have promises. |
| 10:28 | <nox> | (And where "just remove use of promises" was pretty damn bigger than "just avoid that one for const loop". :P) |
| 10:28 | <Ms2ger> | Indeed |
| 10:28 | Ms2ger | wanders off for a baguette |
| 10:29 | <nox> | Hah. |
| 10:30 | <nox> | It's unfortunate that the expectation update tool doesn't remove expectations of tests that weren't run at all, |
| 10:30 | <nox> | because I almost didn't realise we weren't running any of those tests anymore. |
| 11:07 | <nox> | I feel like there is something broken in our update-wpt tool. :( |
| 11:10 | <nox> | In WPT, are the tools/, resources/ folders and whatnot supposed to be listed in MANIFEST.json? |
| 11:12 | <Ms2ger> | nox, I wouldn't be surprised by either answer :) |
| 11:13 | <nox> | Ms2ger: Heh. |
| 11:13 | <nox> | Ms2ger: So I ran update-wpt which introduced suspiciously-unrelated files in the manifest, |
| 11:13 | <nox> | ran update-manifest afterwards, and now they are gone. |
| 11:13 | <nox> | I decided I'm not paid enough to try to understand what just happened and moved on. :P |
| 11:14 | <nox> | tests/wpt/metadata/MANIFEST.json | 2163 +++++++++++++++++++++++++++------- |
| 11:14 | <nox> | Seems small enough. |
| 11:15 | <Ms2ger> | I guess you'd have to ask jgraham :) |
| 11:25 | <foolip> | nox: did you ping me here? I can't find the notification that I saw :/ |
| 11:25 | <nox> | foolip: Don't worry. :) |
| 11:25 | <nox> | foolip: I confused myself. |
| 11:30 | <JakeA> | annevk: https://github.com/whatwg/fetch/issues/681#issuecomment-379226334 |
| 12:18 | <jgraham> | nox: Huh, I reckon that's a regression |
| 12:18 | <nox> | jgraham: The stuff being dumped inside? |
| 12:19 | <jgraham> | Yeah |
| 12:22 | <nox> | jgraham: Was it fixed recently? |
| 12:22 | <jgraham> | nox: I'm not sure; I'm investigating :) |
| 12:34 | <annevk> | hmm, window.frames is a data property in Chrome?! |
| 13:00 | <jgraham> | nox: I think there's a mismatch between the worktree and the git-based behaviour |
| 13:09 | <ajeffrey_> | nox: we need an SMup for many many reasons. |
| 13:10 | <nox> | ajeffrey_: True. :) |
| 13:34 | <Ms2ger> | ajeffrey_, that's been true for years :) |
| 13:42 | <nox> | Ms2ger: 2 |
| 13:42 | <Ms2ger> | nox: o |
| 13:43 | <nox> | Did last one Summer 2016. |
| 13:48 | <Ms2ger> | Did I comment out loud back then that we shouldn't wait long with the next one? |
| 14:26 | <nox> | Ms2ger: It was not supposed to go like this. |
| 14:26 | Ms2ger | cackles, scaring the neighbourhood's children |
| 14:38 | <nox> | Ms2ger: This is my life now: https://github.com/servo/servo/labels/A-content%2Fwebgl |
| 14:39 | <Ms2ger> | My condolences |
| 14:39 | <nox> | :D |
| 15:34 | <JakeA> | annevk: Who should I get to review changes to test_harness? |
| 15:36 | <annevk> | JakeA: jgraham or gsnedders |
| 15:37 | <JakeA> | Thanks |