| 00:07 | <smaug> | if window.open(..., "noopener"); is used for about:blank, what is the origin of the about:blank ? |
| 00:07 | <smaug> | same as opener's or opaque ? |
| 00:08 | smaug | is too lazy to read the spec |
| 04:06 | <annevk> | ajeffrey_: they return null when the browsing context is discarded? |
| 09:08 | <zcorpan> | i learned that Optimus have keyboards with a display in each key |
| 09:15 | <annevk> | MikeSmith: seems he got a response for once |
| 09:15 | <annevk> | ajeffrey_: I found out the answer btw, thanks for filing |
| 11:39 | <MikeSmith> | annevk: https://www.w3.org/TR/streams-api/ has been changed to Note (instead of WD) |
| 11:40 | <annevk> | Very cool |
| 11:41 | <MikeSmith> | hmm the warning thing at the bottom got dropped though |
| 11:41 | <MikeSmith> | will get it re-added |
| 11:41 | <MikeSmith> | I mean like it had at https://www.w3.org/TR/2014/WD-streams-api-20141023/ |
| 11:43 | <MikeSmith> | annevk: hey also btw at https://github.com/w3c/web-platform-tests/pull/4261#issuecomment-263612913 you’ll notice that the test filenames in the stability-checker comments are now hyperlinks to the http://w3c-test.org/submissions copies |
| 11:43 | <MikeSmith> | I think you had asked for that a few weeks back |
| 11:45 | <annevk> | I noticed that, but forgot to thank someone, great |
| 11:48 | <MikeSmith> | annevk: next change needed is for cases like https://github.com/w3c/web-platform-tests/pull/4247 and https://github.com/w3c/web-platform-tests/pull/4104 where the change is to the .js or .json source for a test but not to the .html test file |
| 11:49 | <annevk> | That seems a little harder |
| 11:49 | <MikeSmith> | in those cases currently the stability checker does not run |
| 11:49 | <MikeSmith> | yeah harder but working on it |
| 11:49 | <MikeSmith> | https://github.com/w3c/web-platform-tests/pull/4100 is hte PR for that |
| 11:50 | <MikeSmith> | should be able to get it done by early next week |
| 11:53 | <annevk> | MikeSmith: can we do https://www.w3.org/TR/url/ and https://www.w3.org/TR/url-1/ too you think? |
| 11:55 | <annevk> | MikeSmith: updated https://wiki.whatwg.org/wiki/Fork_tracking btw |
| 12:23 | <MikeSmith> | annevk: yep will ask for those to get moved to Note too |
| 12:53 | <annevk> | \o/ |
| 13:28 | <zcorpan> | mathiasbynens: "(Things like window.opener.document are subject to CORS though.)" https://mathiasbynens.github.io/rel-noopener/ - is not correct; CORS doesn't enable accessing the DOM (unless you use XHR) |
| 13:28 | <zcorpan> | mathiasbynens: window.open and <a> can't do a CORS request |
| 13:28 | <zcorpan> | or iframe etc |
| 13:28 | <smaug> | annevk: jgraham: is there some list of APIs missing wpt tests? |
| 13:28 | <annevk> | Not that I know about |
| 13:29 | <smaug> | or should one file github issue about missing tests or something? |
| 13:29 | <jgraham> | smaug: File an issue I guess |
| 13:29 | <jgraham> | I think there is a label to indicate missing coverage |
| 13:29 | <annevk> | Filing an issue is a start, perhaps against the API and WPT |
| 13:31 | <smaug> | this is about BroadcastChannel |
| 13:34 | <annevk> | smaug: BroadcastChannel has at least one open PR for a test |
| 13:34 | <smaug> | ah, good |
| 13:37 | <smaug> | annevk: which spec defines noopener? |
| 13:37 | <smaug> | oh, html |
| 13:38 | <ajeffrey_> | annevk: np |
| 13:38 | <smaug> | ok fine. noopener doesn't affect to origin of window.open'ed about:blank or javascript or anything |
| 13:39 | <annevk> | smaug: nah, just reachability |
| 13:55 | <mathiasbynens> | zcorpan: care to send a PR? https://github.com/mathiasbynens/rel-noopener/blob/gh-pages/index.html |
| 14:01 | <zcorpan> | mathiasbynens: https://github.com/mathiasbynens/rel-noopener/pull/6 |
| 14:12 | <mathiasbynens> | zcorpan: you da man |
| 17:57 | <zcorpan> | what should OAuth do instead, if rel=noopener was used (or it was the default)? in that case postMessage doesn't work either. maybe iframe + popup + localStorate + "storage" event? |
| 18:43 | <Domenic> | mkwst: not sure you want to wade into the conspiracy theories but https://github.com/whatwg/html/issues/2122 is an ... opportunity to educate people. |
| 19:04 | <Mek> | smaug: I upstreamed the chrome BroadcastChannel tests, which landed upstream in wpt yesterday iirc |
| 19:04 | <smaug> | Mek: ah, great thanks |
| 19:05 | <smaug> | Mek: probably not tests for using BroadcastChannel in contexts with opaque origin |
| 19:05 | <smaug> | (since that fails in FF and Chrome, but FF is about to be fixed) |
| 19:06 | <Mek> | actually I did add a test for that too (currently testing that it should fail, since that seemed to match the current implementations; and tehre is an open spec issue about it) |
| 19:07 | <Mek> | smaug: https://github.com/whatwg/html/issues/1319 |
| 19:07 | <Mek> | I guess I should possibly have tried to get that issue resolved before landing the test though... |
| 19:09 | <smaug> | Mek: you landed tests against the spec? |
| 19:09 | <smaug> | fwiw, I think that spec bug is definitely invalid |
| 19:09 | <smaug> | no good reason to add special cases there |
| 19:10 | <smaug> | tests should test spec, not implementations |
| 19:10 | <Mek> | yeah, possibly I shouldn't have included that particular test |
| 19:14 | <Mek> | and I agree that implementation limitations shouldn't be the guiding force for what the spec should say about things. But BCChannel in opaque origins also seems of limited usefulness |
| 19:15 | <smaug> | Mek: how so |
| 19:16 | <smaug> | BC can be used for random in-page communication too, and special casing it for opaque origins would just break such scripts |
| 19:16 | <smaug> | without any reason |
| 19:16 | <Mek> | well, at least in chrome's implementation of opaque origins it's unlikely that I'll be able to implement anything more than letting you talk to yourself. But yeah, I suppose that's true. |
| 19:17 | <Domenic> | oh i see the conversation already happened here :) |
| 19:22 | <Mek> | probably better to keep the conversation on the bug though :) |
| 19:29 | <Mek> | and my original chrome implementation for BCChannel actually did more or less support it in opaque origins. But code reviewers pushed back heavily against the extra code complexity... At least bringing back that complexity will make it more performant... |
| 21:56 | <gsnedders> | https://github.com/html5lib/html5lib-python/issues/310 — I… don't know how to respond really. |
| 22:02 | <Mek> | hah |
| 22:08 | <caitp> | it's understandable to want to be able to do that |
| 22:09 | <gsnedders> | yeah, I agree |
| 22:09 | <caitp> | even if you didn't want to copy/paste, it's just easier to read, which is the most important thing in software |
| 22:09 | <gsnedders> | but it's p. awkward |
| 22:12 | <gsnedders> | and I can't remember if there aren't concerns with legacy browsers about that |
| 22:12 | <gsnedders> | though that's not such a worry any more |
| 22:15 | <TabAtkins> | It's definitely necessary to encode the ampersand for links in general - otherwise you get things like "example.com?foo¢er=true" being interpreted as being "foo¢er". |
| 22:16 | <TabAtkins> | I don't think there are any collisions between recognized mailto query params and the optional-semicolon named entities, but shrug. |
| 22:17 | <TabAtkins> | (I know that ¢ thing is still happening because I recently had to fix one my personal apps due to it.) |
| 22:19 | <gsnedders> | TabAtkins: that's not true in attributes |
| 22:19 | <TabAtkins> | whaaaaaat? |
| 22:20 | <gsnedders> | TabAtkins: only entities ending in ";" are supported in attributes |
| 22:20 | <gsnedders> | TabAtkins: the ";"-less ones are only in element content |
| 22:20 | <TabAtkins> | Ahhhh. |
| 22:21 | <gsnedders> | TabAtkins: the ambiguity is only with &[a-zA-Z]+; which might be entities in future |
| 22:21 | <TabAtkins> | Then, hm. Are we still allowing ; as the alternate param separator? |
| 22:21 | <gsnedders> | idk |
| 22:33 | <darobin> | using ; as an alternate separator is completely up to server-side implementations, and they vary a lot |
| 23:04 | <7YUAAQM6D> | this was mailto: |
| 23:05 | <zcorpan> | testing |
| 23:05 | <zcorpan> | wtf happened to my nick |
| 23:05 | <gsnedders> | zcorpan: you got 7UAA'd |
| 23:06 | <TabAtkins> | hehehe |