00:30
<caitp>
does anyone know where that spec about same-origin security went? I thought it got put on github on the whatwg org, but I don't see it
00:33
<caitp>
whatever it was called, that thing that used to be just an etherpad
00:37
<caitp>
god why is it impossible to find frigging anything :<
00:42
<caitp>
https://github.com/annevk/html-cross-origin-objects this thing --- y'all should move that to an actual org sometime so it's easier to find @_@
01:26
<Domenic>
caitp: it's in the actual spec now
01:26
<Domenic>
annevk: you should gut that repo
01:27
<caitp>
eh, it's even harder to find things in the html spec, I dunno how you guys deal with it
01:27
<Domenic>
caitp: https://html.spec.whatwg.org/#cross-origin-objects + https://html.spec.whatwg.org/#the-windowproxy-exotic-object + https://html.spec.whatwg.org/#the-location-interface
01:27
<Domenic>
singlepage plus Ctrl+F gets you pretty far
01:27
<caitp>
on a laptop? takes forever :p
01:27
<caitp>
even worse on a phone
01:39
<caitp>
hmm, so you guys made Window not array-like by moving the indexed property access stuff to WindowProxy to escape the rules of webidl? o_o
01:46
<Domenic>
I believe so that Window could become a non exotic object
01:46
<Domenic>
Also because of something about navigation and cross origin indexed access
08:32
<annevk>
Domenic: it says right at the top that it's obsolete
08:42
<mkwst>
annevk: How does Anolis resolve references to documents? Like, what do I need to poke to get the credential management thing into a list that it can resolve?
08:42
<mkwst>
(Also, good morning!)
08:42
<annevk>
mkwst: morning, you need to have https://github.com/whatwg/xref/ in parallel dir
08:43
<mkwst>
annevk: I do. I guess I need to send a PR there, thanks!
08:44
<annevk>
mkwst: I can give you commit access
08:44
<mkwst>
Ok. I'd still ask you to review a PR. :)
08:44
<annevk>
ok
08:44
<mkwst>
just to make sure I understand how it works.
08:47
Ms2ger
sticks his head up
08:48
<Ms2ger>
Who's still using anolis?
08:48
Ms2ger
is starting to understand why Windows XP is still everywhere
08:53
<zcorpan>
Ms2ger: so how do i get anolis to run on XP??
08:53
<Ms2ger>
You'd better send me chocolate...
08:53
<zcorpan>
:-)
08:55
<annevk>
Given your respective living locations, I'd expect chocolates to move in the other direction
08:56
<Ms2ger>
And they historically have
08:58
<annevk>
Must be a great many Anolis on Windows XP users for that to have changed
09:23
<zcorpan>
why doesn't https://html.spec.whatwg.org/#dom-option-text just use the child Text nodes??
09:25
<annevk>
zcorpan: topic?
09:26
<zcorpan>
annevk: i'm introducing the "child text content" concept for <title>.text, <script>, <style>, <time>, etc., and i found <option>.text doing this weird thing (which seems to be implemented in webkit/blink/gecko)
09:35
<Ms2ger>
zcorpan, is that the skipping script thing?
09:42
<annevk>
zcorpan: heh
09:48
<zcorpan>
Ms2ger: yeah. now found https://www.w3.org/Bugs/Public/show_bug.cgi?id=19549 https://github.com/whatwg/html/commit/b1adc35035eff05e6177d11cc5e564934a0cc977
09:49
<Ms2ger>
On the plus side, there's a bunch of tests
09:53
<zcorpan>
yeah, i'm not going to try to change this one
15:17
<annevk>
Domenic: philipj_: MikeSmith: shall we adopt an "Editorial:" policy? That commits that are editorial start with that in their title
15:17
<annevk>
Makes it easier to filter the log
15:17
<annevk>
TC39 does this and it seems rather nice (not all is perfect there btw, they also use the repo for the spec for a ton of other things)
15:24
<annevk>
I'm glad there's at least one group maintaining HTML that does seek outside input since this is rather dreadful: https://github.com/w3c/html/pull/144
15:25
<annevk>
I thought the whole pitch was that the W3C had "due process"
15:36
<MikeSmith>
annevk: OK by me but I wonder if we could abbreviate it, given that git wants to limit the first like of the description to 50 chars
15:40
<annevk>
MikeSmith: maybe, it seems better to be able to tell it's editorial quickly and if it's a big change you can elaborate in the body
15:40
<MikeSmith>
ok
15:40
<MikeSmith>
yeah, makes sense
15:41
<gsnedders>
Ed: foo?
15:42
<MikeSmith>
“✏️ Fixed typo in previous commit”
15:43
<MikeSmith>
anyway I now realize I really liked the way that the old commit-log viewer grayed out all the editorial commits
16:28
<wanderview>
JakeA: do you have an opinion about my Cache API questions here? https://github.com/w3c/webappsec-credential-management/issues/11
16:32
<JakeA>
wanderview: thanks for the heads up, will give it a proper read
16:32
<wanderview>
thanks
16:42
<wanderview>
annevk: does the websocket integration with fetch allow any interception by service workers? #lazyirc
16:44
<annevk>
wanderview: it sets the skip-service-worker flag
16:45
<wanderview>
annevk: ok, thanks!
17:07
<annevk>
wanderview: if we ever did WebSocket it would have its own event I think
17:07
<annevk>
wanderview: WebSocket in service workers, that is
17:08
<wanderview>
cool
17:08
<annevk>
wanderview: but I don't think there's a real need for that given postMessage() protocols and such
17:08
<wanderview>
yea, WebSocket and service workers needs thought... just wanted to make sure I hadn't missed it in case I needed to write an "implement this" bug on gecko
17:10
<annevk>
The main reason this is done is so that WebSocket doesn't miss out on all the security hooks that Fetch has
17:10
<annevk>
As a bonus it also defines cookie handling better
17:11
<Domenic>
annevk: sure but I think just gutting it like I did https://github.com/whatwg/meta-theme-color would be better. Old commits are still there.
17:12
<Domenic>
annevk: "Editorial: " prefix seems reasonable to me, although the line can be pretty fuzzy sometimes. I guess we should add a tag that we tag PRs with so people remember.
17:15
<annevk>
Domenic: removed more content from it
17:21
<Domenic>
👍
18:37
<jsbell>
MikeSmith: while I'm not (yet) volunteering to edit FileAPI (still trying to find someone), can you anoint me ('inexorabletash') with appropriate powers to triage/label the open issues?
18:46
<MikeSmith>
jsbell: will do
18:46
<MikeSmith>
gimme a moment
18:49
<MikeSmith>
k
18:49
<MikeSmith>
jsbell: should be done now
18:51
<jsbell>
MikeSmith: thx. Hasn't appeared yet but perhaps it takes a bit.
18:54
<MikeSmith>
no should be immediate
18:54
MikeSmith
re-checks
18:55
<MikeSmith>
oops
18:55
<MikeSmith>
yeah my mistake
18:55
<MikeSmith>
fixed now
18:55
<MikeSmith>
jsbell:
18:56
<MikeSmith>
jsbell: ⬆
18:56
<jsbell>
MikeSmith: confirmed - thx!
18:57
<MikeSmith>
super
20:18
<TabAtkins>
Domenic: Watching your WebDir talk, and I *cannot stop* hearing you saying "async Perl".
22:39
<rniwa>
annevk, smaug____: W3C tests for rootNode at https://github.com/w3c/web-platform-tests/pull/2665
22:39
<smaug____>
jgraham: you're going to merge those soon, right ;)
22:40
<smaug____>
I mean to mozilla-inbound
22:42
<jgraham>
smaug____: Well if someone reviews them and merges them to wpt, sure
22:42
<gsnedders>
http://news.mit.edu/2016/system-loads-web%20pages-34-percent-faster-0309
22:46
<jgraham>
I read that whole article and I still have no idea what they did
22:53
<gsnedders>
jgraham: yeah. I hope more gets said at conference and published thereafter
22:54
<MikeSmith>
what conference?
22:57
<tantek>
the Web Hypertext Application Technologies conference?
23:01
<gsnedders>
MikeSmith: "he will present at this week’s USENIX Symposium on Networked Systems Design and Implementation"
23:01
<MikeSmith>
ah ok
23:02
<MikeSmith>
well, big deal :) nobody’s ever heard of that
23:03
<MikeSmith>
seriously it looks promising if true
23:03
<MikeSmith>
tantek: the Web Hypertext Application Technologies conference is what we got here every day 😆
23:03
<MikeSmith>
the longest-running conference on the Web runtime and security model and feature set
23:03
<tantek>
nah, we have more like a Web Hypertext Application Technologies webinar, watched through Lynx
23:04
<MikeSmith>
hahah
23:04
<MikeSmith>
fair enough
23:04
<tantek>
seriously would be amazing for there to be a Web Hypertext Application Technologies conference, especially if we could convince Hixie to keynote