08:08
Ms2ger
hates css
08:28
<zcorpan>
MikeSmith: https://www.w3.org/Bugs/Public/show_bug.cgi?id=29563 maybe kill that component?
08:46
<MikeSmith>
zcorpan: thanks for the heads-up
08:47
<MikeSmith>
have disabled it for new bugs
10:27
<gsnedders>
Ms2ger: what's CSS done now?
11:02
<Ms2ger>
gsnedders, I was trying to start my day with something simple: add a few spec links to our text-decoration code
11:02
<Ms2ger>
Unfortunately, the spec doesn't have any requirements, let alone anchors to link to
11:03
<jgraham>
You sound surprised
11:03
<jgraham>
Any chance I can interest you in some PR reviews?
11:03
<Ms2ger>
You could try
11:04
<jgraham>
https://github.com/w3c/wptrunner/pull/179 or https://github.com/w3c/wpt-tools/pull/61 perhaps
11:06
<Ms2ger>
I wish wpt-tools had tests
11:06
<jgraham>
I wrote a test in that patch!
11:07
<Ms2ger>
!
11:07
<Ms2ger>
How does one run it?
11:07
<jgraham>
Using py.test
11:08
<Ms2ger>
, he says, expecting the other party to understand what that means in practice :)
11:13
<gsnedders>
Ms2ger: pip install pytest; py.test
11:13
Ms2ger
pokes around
11:13
<jgraham>
In general I agree that the lack of tests is sadness
11:14
<jgraham>
Code that is mostly reading files is kind of hard to test without the tests spiralling out of control
11:15
<jgraham>
I suppose I could try writing something that actually creates a git repo with some content and then makes local changes to that…
11:16
<jgraham>
In other news, all the blink-derived, mozilla-modified, web-platorm-tests for service workers just got upstreamed
11:17
<jgraham>
So hurrah to wanderview and others for that
11:17
<MikeSmith>
yippee
11:22
<Ms2ger>
Hmm, it discovered pywebsocket tests
11:22
<jgraham>
Hah
11:22
<jgraham>
You might want to run it under the manifest directory
11:22
<jgraham>
Or pass in that directory on the command line
11:34
<Ms2ger>
ImportError: No module named html5lib
11:37
<jgraham>
Ms2ger: Ah, you might have to create a virtualenv and install that, or something
11:37
<annevk>
Hmm, GitHub should have some feature where the author changes depending on who made the most changes in a squashed changeset
14:04
<annevk>
Interesting tactic by the W3C: https://github.com/w3c/webcomponents/commit/9d349eed0182c4a2c04e9da1676af2b2c41fdb9c
14:05
<annevk>
You can't be an editor but we'll take your edits anyway
14:11
<ato>
annevk: When a browsing context navigates to i.e. file:///, does the spec still consider that the same browsing context?
14:12
<ato>
annevk: May sound like an odd question, but it looks like Gecko replaces the content window with a chrome window, and I’m finding it difficult to grasp the terminology here.
14:13
<wanderview>
jgraham: thanks for uplifting it!
14:14
<jgraham>
wanderview: np
14:14
<jgraham>
I was super-pleased with the number of patches to uplift this time
14:16
<wanderview>
meaning lots of new wpt tests?
14:16
<jgraham>
Well I think lots of fixes from the e10s work too
14:16
<wanderview>
cool
14:17
<annevk>
ato: I think it would be the same browsing context, yes
14:17
<annevk>
ato: browsing context : window is 1:m
14:17
<annevk>
ato: but each browsing context does have a single WindowProxy, iirc, which makes this a bit confusing
14:19
<ato>
annevk: Okay, this is fine. Just needed some confirmation my tests aren’t completely off.
14:20
<Ms2ger>
browsing context : windowproxy is 1:1, yes
14:20
<annevk>
jgraham: sounds like a couple of years of work are starting to have some impact 😃
15:34
<zcorpan>
Domenic: do you want https://github.com/whatwg/html/pull/994 as one commit or as three?
15:39
<zcorpan>
i suppose one
15:47
<Domenic>
zcorpan: sorry. Could go either way, one is fine.
15:47
<Domenic>
Interesting, I can no longer close or assign bugs on w3c/webcomponents.
16:03
<annevk>
That is weird, MikeSmith? ^^
16:04
<annevk>
I wonder if I got the same treatment, can't check right now, but that seems super petty
16:04
<annevk>
Especially after all the work they just copied from you without asking
16:58
<annevk>
Domenic: seems they only took it out on you, I can still do those kind of things, though I can't assign anything to you
18:54
<smaug____>
trying to recall... DOMString may include '\0', right?
18:55
<jsbell>
smaug____: yes
18:56
<jsbell>
and doesn't need to be valid UTF-16, can include non-characters, yadda yadda
18:57
<smaug____>
\u0000 seems to work
19:03
<jsbell>
Frequent source of implementation bugs with APIs that take DOMStrings, of course.
19:56
<TabAtkins>
Yeah, DOMString is in the WTF-16 encoding https://simonsapin.github.io/wtf-8/#wtf-16
20:46
<Sebmaster>
annevk: does http://unicode.org/reports/tr46/#Validity_Criteria disallow urls like https://r3---sn-p5qlsnz6.googlevideo.com/, or am i misinterpreting somewhere?
21:38
<SimonSapin>
Sebmaster: it does but implementations apparently don’t follow that rule: https://github.com/whatwg/url/issues/53#issuecomment-181528158
21:40
<Sebmaster>
Sweet, thanks, missed that response for some reason
22:03
<TabAtkins>
Domenic: Mind screenshotting that date input and sending to me?
22:04
<jyasskin>
MikeSmith: Can you switch https://github.com/w3c/permissions to only allow squash merges? (Or give me admin over that repo so I can do it?)
22:04
<Domenic>
TabAtkins: I just did a google image search https://www.google.com/search?q=ios+input+type+date&source=lnms&tbm=isch&sa=X&ved=0ahUKEwij06vBhYDMAhXLez4KHVyOA_YQ_AUICCgC&biw=2301&bih=1315
22:04
<Domenic>
I'll edit the post to include that link
22:09
<jyasskin>
MikeSmith: (https://help.github.com/articles/configuring-pull-request-merge-squashing/)
22:11
<MikeSmith>
jyasskin: Yeah will do now
22:11
<jyasskin>
MikeSmith: Thanks!
22:11
<MikeSmith>
we have done so already for other repos
22:12
<MikeSmith>
well, where we = whatwg
22:19
<MikeSmith>
jyasskin: done now
22:19
<MikeSmith>
and you are now an admin for it as well
22:19
<jyasskin>
:)