00:28
<smaug____>
anyone want to interpret a sentence to me?
00:28
<smaug____>
Domenic: perhaps you're awake
00:28
<TabAtkins>
heycam: Yo, PR on WebIDL.
00:28
<smaug____>
https://html.spec.whatwg.org/multipage/forms.html#dom-cva-reportvalidity
00:28
<TabAtkins>
smaug____: I'm also available if it's in my wheelhouse.
00:28
<heycam>
TabAtkins: looking
00:29
<TabAtkins>
smaug____: What needs to be interpreted about that?
00:29
<smaug____>
TabAtkins: so, "if the element is a candidate for constraint validation and does not satisfy its constraints" ... "return false", "Otherwise, it must only return true without doing anything else."
00:29
<smaug____>
or is the "Otherwise, it must only return true without doing anything else." bound somehow differently
00:30
<smaug____>
like to "if that event is not canceled"
00:30
<TabAtkins>
Unless this sentence is perverse, your first interpretation is right - the condition covers the rest of that sentence, and the following sentence gives the "else" branch for the conditional.
00:31
<MikeSmith>
I concur with TabAtkins’s assessment
00:31
<smaug____>
ok, thanks
00:31
<smaug____>
so, return true if element is valid effectively
00:32
<TabAtkins>
yeah
00:33
<TabAtkins>
I prefer wording my conditionals so that the simpler branch is first, so the binding is clearer. So I'd have reversed this if I were writing it.
00:33
<TabAtkins>
I see why it's the way it is, tho - the condition is an AND of two conditions, and negating that makes it less clear to express.
00:34
<TabAtkins>
"If the element is not a candidate for constraint validation, or is a candidate and satisfies its constraints..."
00:34
<TabAtkins>
Definitely a little fiddlier.
00:35
<TabAtkins>
I might just abandon English and go for a <dt class=switch>, then. Those are great when you've got conditionals with long branches.
00:35
<TabAtkins>
s/dt/dl/
09:28
<annevk>
hayato: not sure if you're still awake, but would greatly appreciate your input on the issues around events
09:30
<zcorpan>
<p>If the <var>current media resource</var> is a raw data stream (e.g. from a
09:30
<zcorpan>
<code>File</code> object), then to determine the format of the <span>media resource</span>,
09:30
<zcorpan>
can it be a File object?
09:30
<annevk>
zcorpan: MediaProvider lists Blob, so yes
10:30
<zcorpan>
can someone test http://software.hixie.ch/utilities/js/live-dom-viewer/saved/4085 in Edge?
10:30
<zcorpan>
(interested in whether body fills the viewport or not)
10:45
<MikeSmith>
apologies to any Star Wars fans who follow me in twitter and may have been offended by my disrespect
10:45
<MikeSmith>
at the time I tweeted I was filled up with a lot of cheap airplane Chardonnay washed down with Goosehead IPA
10:47
<AutomatedTester>
MikeSmith: were those tweets about the latest one?
10:47
<annevk>
haha
10:48
<annevk>
MikeSmith: was already thinking of putting on Star Wars movies while you're here
10:51
<MikeSmith>
annevk: haha
10:52
<MikeSmith>
Star Wars marathon
10:52
<MikeSmith>
AutomatedTester: yeah the latest one
10:53
<MikeSmith>
anyway was I drifting in and out of sleep so in hindsight I am not sure how much of it I remember was the the actual movie or instead something I was dreaming
11:09
<annevk>
Wait per https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit/Source/core/events/EventTarget.cpp&sq=package:chromium&l=274 stuff dispatched through dispatchEvent() doesn't go through various phases in Chrome?
11:13
<annevk>
I guess I'm missing something
11:13
<annevk>
<script>document.documentElement.addEventListener("x", w);document.body.dispatchEvent(new Event("x", { bubbles: true }))</script> works fine
11:14
<ondras>
annevk: I would say that it goes only through AT_TARGET and bubbling
11:14
<Ms2ger>
JS calls https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit/Source/core/events/EventTarget.cpp&sq=package:chromium&l=252&rcl=1461122240
11:14
<ondras>
or maybe not
11:17
<annevk>
I wonder why they have all that different code
11:17
<annevk>
It's certainly not very clear
11:19
<annevk>
Ms2ger: but that also seems to call dispatchEventInternal
11:20
<Ms2ger>
Yeah, dunno
11:23
<annevk>
I was trying to figure out the shadow tree logic by reading the code, but ...
11:58
<MikeSmith>
so anybody ever had to block somebody from being able to post comments to a github issue tracker?
11:58
<MikeSmith>
have reached the end of my patience with this clown: https://github.com/w3c/push-api/issues/189#issuecomment-212377594
11:58
<annevk>
I blocked the entire <keygen> issue...
12:04
<Ms2ger>
Woot
12:04
<Ms2ger>
FileAPI has an editor
12:17
<MikeSmith>
yeah
12:18
<MikeSmith>
was there any announcement anywhere?
12:18
<MikeSmith>
Ms2ger: ⬆
12:18
<Ms2ger>
Not that I saw
12:23
<MikeSmith>
OK yeah would like to avoid somebody noticing who thinks we need to get formal approval
12:24
<MikeSmith>
and instead just get the facts on the ground from actual editing before they do
12:25
<MikeSmith>
not that I can imagine anybody would find this controversial but they do find surprising and novel ways to step in the way of productive working actually getting done
13:09
<annevk>
smaug____: \o/
13:10
<annevk>
smaug____: you're awake, do you also have time?
13:51
<MikeSmith>
I did manage to figure out how to perma-ban the https://github.com/w3c/push-api/issues/189#issuecomment-212377594 troll org-wide
13:53
<MikeSmith>
it’s at https://github.com/organizations/FOO/settings/user_blocks where FOO is the org name
14:00
<MikeSmith>
in other news, among the long list of ugly consequences the semweb crowd have inflicted on themselves with bad (over)designs, apparently since RDF vocabularies and “JSON-LD context files” are hard-wired in some way to depend on dereferenceable namespaces, they have dug themselves into being stuck with http as a result
14:01
<MikeSmith>
and can’t move those resources to https URLs ever
14:02
<annevk>
Can't they use owl:sameAs?
14:02
<MikeSmith>
I guess this was noted in the objections of some TAG members to the consensus that the rest of the universe has about moving to HTTPS everywhere, but I wasn’t paying attention
14:03
<MikeSmith>
annevk: dunno but if not that maybe the can layer even some more extra shit on top of it
14:04
<MikeSmith>
no matter what they would have a hard time making it fundamentally worse
14:04
<annevk>
That's exactly what I understood owl:sameAs to be
14:04
<annevk>
Some way to paper over shit that is other people's identifiers
14:04
<MikeSmith>
nobody has mentioned that yet that I recall
14:05
<MikeSmith>
but again I’m not really paying attention more than just to read stuff and shake my head in disbelief
14:06
<annevk>
MikeSmith: you don't read the stuff W3C publishes? This is from back in the days of "The Workshop": https://www.w3.org/TR/owl-ref/#sameAs-def
14:06
<MikeSmith>
the context for why I notice this today is that (surprise surprise) they are lobbying against efforts to get the W3C site moved all over to TLS
14:06
MikeSmith
looks at https://www.w3.org/TR/owl-ref/#sameAs-def
14:07
<MikeSmith>
shrug I guess they have no way to hook that into namespace resolution
14:08
<MikeSmith>
but boy look at the list of worthies who authored that doc
14:08
<MikeSmith>
really heavy-hitters in Web technologies
14:09
<MikeSmith>
seriously this is what happens when you have people designing things for that Web who don’t actually understand how the Web actually works
14:10
<annevk>
There's lots that applies though
14:10
<MikeSmith>
sure
14:10
<annevk>
SMIL must be one of my more favorite examples of designing gone wrong: https://annevankesteren.nl/2006/03/smil
14:11
MikeSmith
reads https://annevankesteren.nl/2006/03/smil
14:12
<MikeSmith>
god ten years ago, old man
14:12
<MikeSmith>
and still nobody listens to you!
14:13
<MikeSmith>
haha “Ten namespaces more than HTML element names in HTML 2.0, 3.2, 4.01 and XHTML 1.1 combined” nice
14:13
<MikeSmith>
the Schreiber guy who co-edited that doc at least went on to become Wolverine’s brother and fight back to back together when him against evil Dead Pool, which is arguably a bit more than you have accomplished
14:13
<MikeSmith>
but you still have time
14:15
<Ms2ger>
"W3C isn't really consistent"
14:15
<Ms2ger>
- annevk, 17 December 2003
14:21
<wanderview>
annevk: is there a big difference between a Response with an empty url list and a Response with a single url that is the empty string?
14:21
<wanderview>
or a significant difference, I mean
14:21
<annevk>
wanderview: a URL cannot be the empty string
14:21
<annevk>
wanderview: at least, a URL record cannot
14:22
<wanderview>
annevk: but Response only exposes urls as strings
14:23
<wanderview>
annevk: is it safe to safe Request must have at least one url in its url list, but Response may have no urls in its url list?
14:23
<annevk>
wanderview: that sounds correct
15:01
<annevk>
gsnedders: happy b-day young man 🎉🎂
15:50
<gsnedders>
annevk: I SAID YOUNG MAN WHEN YOU WALK IN THE DOOR
15:51
<annevk>
gsnedders: loud party? 😛
17:16
<Domenic>
TabAtkins: should Bikeshed grow an "extended attribute" definition type (for Web IDL), or should I just stick with "dfn"?
17:17
<TabAtkins>
Domenic: Currently I sprout specific types for the few extended attributes I care about. Lay out some use-cases for me that you want to hit and I'll see what the right thing should be.
17:18
<Domenic>
TabAtkins: well I realized I defined https://html.spec.whatwg.org/multipage/scripting.html#cereactions yesterday but forgot to export it so I can use it in DOM etc. So now I want to correct that but I'm not sure how I should mark it up.
17:19
<annevk>
smaug____: could you make time today to explain how Gecko does retargating (I'm assuming we have code for that) in https://github.com/w3c/webcomponents/issues/485?
17:20
<annevk>
smaug____: I have to go in a bit, but that would help a lot with figuring out the right way to integrate the monkey patching of events from Shadow DOM
17:20
<annevk>
smaug____: it's rather unclear to me at the moment what the right design should be, since the existing specification is rather unclear on the particulars, at least to me
17:20
<TabAtkins>
Right now "dfn" is the right type. A new type *might* be useful, so I could autolink IDL to the definition of the extended attribute, but that would also clash with the autolinking of [Constructor]. I'll have to think about this.
17:21
<Domenic>
Yeah, I was going to ask about autolinking. But for now I think it works to just add [<a>CEReactions</a>]
17:21
<smaug____>
annevk: Gecko's relatedTarget retargeting may not be the good one. it happens during event handling
17:22
<annevk>
smaug____: I'm just talking about target retargeting for now
17:22
<TabAtkins>
Domenic: It does not, no, because Bikeshed reparses the IDL and throws out any markup in it. ^_^
17:23
<smaug____>
annevk: but let me read that bug again once I've dealt with these housing cooperative papers
17:23
<Domenic>
uh oh
17:23
<TabAtkins>
(It'll work for HTML where y'all still mark up all your IDL manually.)
17:23
<smaug____>
in 10 mins or so
17:23
<annevk>
smaug____: enjoy and thanks
17:23
<Domenic>
So I guess I'll file a bug on Bikeshed to figure this out, and in the meantime DOM will just not have clickable annotations
17:24
<TabAtkins>
It might be fine to clash with Constructor linking - the definition of a Constructor is just the constructor method, which is super-obvious.
17:24
<TabAtkins>
All the rest can link to their concept definition.
18:01
<wanderview>
JakeA: ping?
19:22
<Domenic>
annevk: is ShadowRoot's innerHTML defined anywhere?
19:23
<annevk>
Domenic: that's why I'm considering taking over parts of P&S…
19:23
<Domenic>
annevk: ah OK...
19:26
<Domenic>
So many ReSpec specs.....
19:43
<zcorpan>
font-family: monospace, Droid Sans Fallback, Helvetica Neue, sans-serif; https://resources.whatwg.org/standard.css
19:43
<zcorpan>
that... makes little sense
19:46
<smaug____>
contradiction in the spec: " In all cases, the input event comes before the corresponding change event (if any)." ... "the user agent must queue a task to fire a simple event that bubbles named input at the input element. The corresponding change event, if any, will be fired when the control loses focus."
19:47
smaug____
files a bug
20:11
<smaug____>
mounir: so what would a web page do with memorypressure?
20:12
<smaug____>
I mean, it could sure try to not keep certain stuff alive, but it would be still up to the GC/CC/whatever to actually release the stuff
20:12
<Domenic>
release blobs or canvases maybe
20:13
<smaug____>
mounir: oh, I see, UA is expected to GC afterwards
20:15
<smaug____>
maybe this could work
20:18
<smaug____>
mounir: though, do we really need any levels?
20:36
<JonathanNeal>
I’m rewriting flex in JS, and I’m trying to group (row with row-reverse) and (column with column-reverse). I’m thinking the former means the container has a mainAxis of inline and later means the container has a mainAxis of block. Or should I just call them horizontal and vertical?
21:14
<TabAtkins>
Correct - row means inline, column means block
21:14
<TabAtkins>
JonathanNeal: ^^^
21:15
<JonathanNeal>
Awesome. I thought I had caught that. And the label for that is `axis`?
21:15
<TabAtkins>
What do you mean by "label"?
21:17
JonathanNeal
hopes to explain shortly.
21:33
<JonathanNeal>
TabAtkins: something like this https://gist.github.com/jonathantneal/500a650ce01ddea12d555e127cbb49ba#flex-container- I’m trying to define what happens as I calculate flexbox.
21:39
<JonathanNeal>
I’m describing an `axis` property on the flex container. It is either `inline` or `block` based on `flex-direction`.
22:38
<TabAtkins>
JonathanNeal: Then yes.