09:08
<zcorpan>
jgraham_: the chocolate was very nice. haven't had anything quite like it before. much taste. such flavour. or as i said to jenni, bra fräs
09:09
Ms2ger
wonders which
09:10
<zcorpan>
http://marouchocolate.com/post/58397164815/products 80%
09:12
<jgraham_>
zcorpan: Excellent :)
13:48
<SimonSapin>
http://w3c.github.io/webcomponents/spec/custom/#concepts links for NCName to a 1999 spec that’s been updated twice since, and the new definition looks different. Is this deliberate?
13:57
<zewt>
heh, someone posting to mongodb-user with an <img> containing a gigantic data: url
13:58
<zewt>
i wonder if he got that from "copy image location" on a canvas, that's what firefox did last i checked
14:47
<gsnedders>
SimonSapin: Oh, sorry, it was 3e that updated NCName with a semantic change. I forgot the timing of the changes. :)
14:49
<SimonSapin>
gsnedders: so, is webcomponents linking to 1e a good idea?
14:50
<gsnedders>
SimonSapin: Is anyone using XML1.05e+XMLnames1.0e3? AFAIK Everyone still targets XML1.0e4 + XMLnames1.0e2
14:50
<SimonSapin>
I have no idea
14:50
<gsnedders>
AFAIK no browser has wanted to impl the changes in XML1.05e
14:50
<Ms2ger>
Maybe Servo accidentally
14:51
<jgraham>
gsnedders: annevk suggested that libxml is 5th edition and Chrome blindly imports it
14:51
<gsnedders>
Even though everyone constantly updates the list of what names are valid in JS identifiers
14:51
<SimonSapin>
gsnedders: How can I and other people working on Servo can find out about "everyone agrees" besides asking you?
14:51
<jgraham>
https://bugzilla.mozilla.org/show_bug.cgi?id=501837
14:51
<gsnedders>
jgraham: libxml2 is 5th ed, yes, but has a flag for 4th ed, dunno if Chrome sets it
14:53
<gsnedders>
I don't get why people are willing to change the JS syntax through Unicode updates but not XML.
14:53
<Ms2ger>
Other people
14:54
<gsnedders>
I know. But the people objecting to XML changes typically concern themselves with all of the web platform, inc. JS.
14:55
<Ms2ger>
Fighting tc39 is hard
14:55
<gsnedders>
Sure, but why follow TC39 and ignore XML Core WG?
14:56
<gsnedders>
Difficulty of fighting is irrelevant when you're willing to ignore specs.
14:57
<Ms2ger>
I guess presence of JS engineers who are willing to follow tc39
14:57
<zcorpan>
i think upgrading to 5ed presented difficulties because different set of tools make assumptions, so you need to get all of them upgraded which is hard. js doesn't have that so much
14:58
zcorpan
now escapes
16:01
<Domenic_>
why is XML relevant? I assume this is about HTML syntax?
16:06
<jgraham>
Domenic_: Because web components defines something in terms of xml-names
16:07
<gsnedders>
But *why*?
16:07
<gsnedders>
Because the DOM's creators limit themselves to XML-names?
16:07
<jgraham>
Oh one level deeper?
16:07
<jgraham>
No idea
16:08
<gsnedders>
That's the level I'm curious at.
16:08
<jgraham>
Doesn't it match HTML?
16:08
<gsnedders>
No, it doesn't.
16:09
<gsnedders>
<foo:bar:lol> is perfectly fine in HTML. Also even the local name "foo:bar".
16:09
<Domenic_>
seems bad
16:09
<Hixie>
("perfectly fine" except for being non-conforming of course)
16:09
<gsnedders>
Hixie: At a parser level or just an unknown element level?
16:10
<jgraham>
Sure, I know the parser can deal
16:10
<gsnedders>
Because the latter doesn't really apply in the web components case, surely?
16:10
<jgraham>
I guess HTML has a closed list of elements
16:10
<Hixie>
("foo:bar:lol" is not a known element, though i expect there's a cascade of other errors that would turn up at the same time)
16:10
<jgraham>
So that's not very relevant
16:10
<SamB>
and nobody is ever going to define an element of that name, either
16:10
<Hixie>
yeah, i'm not really sure what we should do for conformance of web components
16:10
<gsnedders>
Who cares that it isn't a known element when we're in the web components case, though?
16:10
<jgraham>
Anyway, making web-component names that you can't create through DOM APIs seems bad
16:11
<gsnedders>
jgraham: Yeah, I presume this is the reason for the restriction
16:11
<gsnedders>
dglazkov: yt?
16:12
<Hixie>
it would be interesting if we could get something useful out of validators even when you use custom components
16:13
<gsnedders>
Indeed.
16:13
<gsnedders>
jgraham: But yeah, elements in HTML are closed set, attributes are open set.
16:13
SamB
wonders what about 5e would make it important to upgrade all software at once -- does just unpinning Unicode really cause this?
16:13
<Hixie>
though really once you're using custom components you've pretty much sacrificed the ability for the web to work without css and js, especially if you don't have a fallback element...
16:29
<SimonSapin>
http://dom.spec.whatwg.org/#dom-document-createelementns and other parts of the same spec link to http://www.w3.org/TR/xml/#NT-Name and http://www.w3.org/TR/xml-names/#NT-QName . Should that change if everyone agrees the latest editions are bad?
16:29
<dglazkov>
good morning, whatwg!
16:29
<Ms2ger>
Hmm, did someone change it...
16:29
<dglazkov>
SimonSapin: I just did what annevk said to do. Let me find the bug
16:32
<dglazkov>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=20973
17:04
<mathiasbynens>
zcorpan (if you read the logs): https://twitter.com/kizmarh/status/453215891908722689
17:22
<Hixie>
hsivonen: yt?
17:30
<annevk>
gsnedders: Chrome implements XML 1.0 5th edition afaict
22:25
<Hixie>
what's happened to live dom viewer on firefox?
22:30
<smaug____>
live dom viewer?
22:32
<Hixie>
http://software.hixie.ch/utilities/js/live-dom-viewer/
22:32
<Hixie>
the rendered view no longer renders?
22:32
<smaug____>
works fine here
22:33
<Hixie>
i'm on 29, beta
22:33
<Hixie>
the rendered view just gives a 0-height iframe
22:33
<miketaylr>
wfm on nightly
22:33
<Hixie>
weird
22:34
<smaug____>
wfm on beta
22:38
<Hixie>
i wonder what's up with my build then
23:04
<gsnedders>
jgraham: req html5lib #146/147 speedy review, blocks Travis CI passing anything
23:25
<zewt>
not sure i've seen anything in a version control system that kills my workflow instantly more than git's auto-gc
23:25
<SamB>
zewt: you could turn that off and set up a cronjob?
23:25
<gsnedders>
You can disable it :P
23:26
<zewt>
right now i'm twiddling my thumbs since it's taking several minutes
23:26
<zewt>
also i don't want expensive cronjobs on my laptop, heh
23:28
<SamB>
excuses!
23:28
<SamB>
I guess you could try to set the threshold so high you could just take a break and watch an episode of something while you wait?
23:31
<zewt>
doesn't help when it happens while I'm at work an hour late trying to finish something :P
23:32
<SamB>
well, you have a choice here: have while you work, or at some other point
23:34
<Hixie>
that's a fallacy of an excluded middle
23:35
<Hixie>
there's a third option: have git not have any synchronous or blocking maintenance tasks.
23:35
<zewt>
i tend to expect database-like systems to not require periodic downtime to retain performance these days
23:36
<SamB>
anyway, patches seem welcome
23:36
<Hixie>
or just use svn :-)
23:36
<SamB>
crazies
23:37
<Hixie>
anyone here managed to get requestAutocomplete() to work?
23:51
<annevk>
Hixie: JakeA might have
23:51
<annevk>
Hixie: sicking expressed a concern with requestAutocomplete() about universal address input not being sorted yet
23:52
<Hixie>
the address-level3 stuff? that's blocked awaiting mozilla input...
23:52
<Hixie>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=25235
23:53
<Hixie>
backstory in http://lists.w3.org/Archives/Public/public-whatwg-archive/2014Apr/0010.html
23:57
<Hixie>
annevk: ^
23:58
<annevk>
hmm
23:59
<annevk>
maybe I should become project manager
23:59
<annevk>
feels like I do stuff like that at least 30% of my time
23:59
<Hixie>
heh