| 05:59 | <Domenic> | annevk: https://infra.spec.whatwg.org/#byte-sequence is "space separated" correct here? |
| 05:59 | <Domenic> | Oh, nevermind |
| 07:02 | foolip | flies to the homeland of annevk to smell (?) the tulips |
| 07:03 | <annevk> | foolip: let me know what that's like and if it's something for kids; I don't recall ever going to those fields |
| 07:04 | <annevk> | foolip: I don't know what your kid(s?) are into, but Madurodam was a big hit with Oscar |
| 07:05 | <foolip> | annevk: we're actually leaving Edison (only kid) with my parents |
| 07:06 | <annevk> | Ah, that sounds rather relaxing 😊 |
| 07:10 | <annevk> | Enjoy! |
| 07:12 | <Ms2ger> | I've never noticed tulips to smell, particularly |
| 08:33 | <ato> | annevk: Looking at pictures, I think that would be a big hit with me as well (-: |
| 08:34 | <annevk> | ato: heh, yeah, it's good |
| 09:14 | <foolip> | PSA: check https://github.com/orgs/web-platform-tests/invitation if you don't read all your GitHub email very carefully |
| 09:14 | <foolip> | nox: ^ |
| 09:15 | <nox> | foolip: Do we have a logo? :P |
| 09:17 | <foolip> | nox: see https://bit.ly/wpt-transition, being discussed semi-seriously |
| 09:18 | <nox> | Nice! Now we just need a html5lib logo. |
| 09:18 | <nox> | https://irccloud.mozilla.com/file/TrCgZ5RO/Capture%20d%E2%80%99e%CC%81cran%202018-04-19%20a%CC%80%2011.15.43.png |
| 09:18 | <nox> | :D |
| 09:18 | <foolip> | nox: we don't have any perfect candidate yet, so we'll see :) |
| 09:21 | <annevk> | Seems to have gotten a bit contentious since last time |
| 09:22 | <annevk> | Even negative points |
| 10:54 | <annevk> | Oooh fun times; the Content-Type Firefox extracts for navigation is different from the one it extracts for Blob objects |
| 10:54 | <annevk> | Chrome is better, but drops stuff on the floor |
| 11:01 | <Zegnat> | Hmm. I was reading about DOMTokenList and I don’t really understand the use of `supported tokens`. It doesn’t look like the validation steps are ever used when tokens are added/toggled/replaced to the list. So a DOMTokenList may contain unsupported tokens no matter how the list was defined? Or is there some implied moment when the validation steps must be executed? |
| 11:15 | <Ms2ger> | Zegnat, looks like it's only used for supports(), yeah |
| 11:16 | <Zegnat> | Yep. That surprised me. Any idea why the list isn’t limited to only supported tokens, e.g. by doing validation on add()? |
| 11:19 | <Ms2ger> | The platform doesn't enforce what values an attribute can take at all, so it'd be weird to start doing it here, without even getting a guarantee (because setAttribute / plain HTML) |
| 11:20 | <Ms2ger> | Also, it's probably better not to block new values in old browsers that don't support them yet |
| 11:22 | <Zegnat> | That makes sense, thanks! |
| 11:26 | <Ms2ger> | My pleasure |
| 11:26 | <nox> | foolip: Voted. |
| 11:36 | <Ms2ger> | nox, you don't get a vote, sorry :) |
| 11:36 | <nox> | Why? ( ._.) |
| 13:08 | <annevk> | jgraham: so part of the reason https://travis-ci.org/w3c/web-platform-tests/jobs/368620442 is failing seems to be that I'm using non-ASCII in a test title |
| 13:09 | <annevk> | jgraham: encoded = msg.encode("utf8", "backslashreplace").decode("utf8") |
| 13:38 | <jgraham> | annevk: Looks like we're trying to double-encode there |
| 13:38 | <jgraham> | Silly Python |
| 13:39 | <annevk> | And there's no tests? |
| 13:39 | <annevk> | I guess it's specific to the Firefox bot |
| 14:36 | <jgraham> | annevk: Specific to the travis integration |
| 14:40 | <jgraham> | annevk: https://pastebin.mozilla.org/9083418 looks like a possible patch. Mind trying it out? |
| 14:41 | <annevk> | jgraham: locally? I'm not sure I even have all the dependencies |
| 14:49 | <jgraham> | annevk: I was thinking you could push it to the PR and see if travis passes |
| 14:49 | <annevk> | aah okay, will do later |
| 14:49 | <DDK_> | A quick curious question.. Does anyone know of a way to use inspect element to legitimately change the value of an online shopping cart from "x" amount to "0" |
| 14:49 | <DDK_> | Just curious to see if it works |
| 17:14 | <annevk> | I guess Fetch is just gonna define parsing of all HTTP headers… |
| 17:55 | <MikeSmith> | annevk: ? |
| 17:56 | <annevk> | MikeSmith: see mimesniff issue on Content-Type |
| 17:57 | <annevk> | MikeSmith: have been contemplating figuring out parsers for other core header values too, e.g., Content-Length |
| 17:57 | <annevk> | Content-Length: 42,42 and such |
| 17:58 | <tantek> | The HTTP spec doesn't define that? |
| 17:58 | <annevk> | tantek: in theory, but not implemented that way |
| 17:58 | <tantek> | oh boy |
| 17:58 | <MikeSmith> | good times |
| 17:59 | tantek | braces himself for the sequel to URL / IRI spec battles |
| 17:59 | <annevk> | tantek: I have filed upstream issues and ping mnot now and then |
| 17:59 | <tantek> | that is good to hear :) |
| 18:00 | <tantek> | I suppose this was inevitable, cleaning up the original TimBL trifecta as it were (HTML, URL, now HTTP) |
| 18:03 | <annevk> | All the primitives belong to us (a handful) |
| 19:10 | <jyasskin> | The HTTPWG is working on a revision of the core specification, so this is a good time to see if they have the appetite to do it right. (Probably not, but still.) |
| 19:10 | <annevk> | jyasskin: yeah am aware, and that’s where my issues are filed |
| 19:11 | <jyasskin> | \o/ |
| 19:11 | <jyasskin> | e.g. https://github.com/httpwg/http-core/issues/58 |
| 19:11 | <annevk> | jyasskin: but even \r as header separator did not make it last time around so… |
| 19:15 | <annevk> | (I filed most issues when that repo was called http11bis or some such.) |
| 19:26 | <railgun> | annevk: a bit off-topic, but would any changes to the spec now imply an HTTP/2.1? I've been out of the loop for a while now and only recently learned about HTTP/2... |
| 19:35 | <annevk> | railgun: it’s a little more complicated; header value parsing is shared across HTTP versions for instance; the \r thing I mentioned only affects HTTP/1, etc. |
| 19:37 | <zcorpan> | hsivonen: re https://github.com/w3c/web-platform-tests/issues/10208 , the html parser waits for 1024 bytes or encoding declaration? |
| 19:52 | <wanderview> | zcorpan: while trying to get around the parser wait in firefox I also had to make sure the response had a Content-Type header... not sure if that impacts you |
| 19:52 | <wanderview> | Content-Type: text/html |
| 19:54 | <zcorpan> | thanks, added that to the issue |
