03:41
<MikeSmith>
maybe somebody else needs to take over maintenance of developers.whatwg.org
06:11
<Hixie>
MikeSmith: yeah it's supposed to be me but i've been slacking
08:40
<foolip>
which phantomjs
08:40
<foolip>
oops :)
08:41
<Ms2ger>
Good morning foolip
08:41
<Ms2ger>
Want to do some test reviews? :)
09:19
Ms2ger
guesses that's a no
09:43
<foolip>
Manishearth: heh :)
09:43
<foolip>
bah, tab completion, and Ms2ger is gone
16:13
<TabAtkins>
foolip: Never rely on tab completion with less than three letters.
16:18
<jgraham>
Well there's always a case where it doesn't work for N letters. I think "read what tab completed to" is better advice :)
16:18
<jgraham>
(e.g. in #ateam on irc.mozilla.org there are two people matching jgr*)
16:24
<Domenic>
irccloud at least is smart and tries to tab-complete the last person you were talking to or something similar
16:31
<rektide>
if anyone has guidance for helping me equivocate the need for Push to actually relate to the web, i feel like i direly need it.
16:31
<rektide>
Push: Not a Friend of SPDY/HTTP, lists.w3.org/Archives/Public/public-webapps/2014OctDec/0257.html
16:31
<rektide>
(sorry) http://lists.w3.org/Archives/Public/public-webapps/2014OctDec/0257.html
16:32
<rektide>
it's absolutely reviling to me people are making yet another high level protocol that will happen to use http and yet not actually expose what's in http
16:32
<rektide>
and calling it a web spec! who the fuck are these people?
16:33
<rektide>
already said too much; i'm here for context setting to understand, and guidance to direct, perhaps if people want to chime in in list that'd be great. please disregard my venting right here.
16:33
<MikeSmith>
rektide: what spec are you commenting on?
16:33
<MikeSmith>
the IETF protocol spec?
16:34
<MikeSmith>
anyway regardless I think Martin Thomson is probably who you want to talk to
16:34
<rektide>
this is on www.w3.org/TR/2014/WD-push-api-20141007/ , but i'm here because i think ya'll are probably more aligned with my pov
16:35
<rektide>
again Chrome with the lack of http:// in copy-paste, thanks obama. http://www.w3.org/TR/2014/WD-push-api-20141007/
16:35
<MikeSmith>
I think nobody who's usually around here is following the Push stuff all that closely
16:35
<MikeSmith>
maybe Domenic has a bit
16:35
<MikeSmith>
or slightlyoff
16:36
<Domenic>
i don't really want to get involved in that conversation; the OP's vulgar attitude makes me think engaging will not be constructive.
16:36
<MikeSmith>
rektide: have you looked at the related IETF spec yet
16:36
<rektide>
thanks Domenic
16:36
<rektide>
MikeSmith: I have not, was not aware that was in progress
16:37
<MikeSmith>
it's called WebPush
16:37
MikeSmith
looks for a URL
16:37
<rektide>
http://http://tools.ietf.org/html/draft-thomson-webpush-http2-00
16:38
<MikeSmith>
rektide: btw as much as I personally love your rhetorical style I hope you can see it's not going to help you much as far as getting people who know about the background on this to feel like paying attention
16:41
<annevk>
rektide: that you equate the Push API with SPDY Push shows you need to a do a little more reading
16:43
<rektide>
annevk: no it doesn't
16:44
<annevk>
well, enjoy, I guess
16:44
<rektide>
i haven't equated, i've indicated that the new fabricated Push API should serve to expose SPDY
16:44
<rektide>
yeah thanks
16:45
<zcorpan_>
Hixie: https://www.w3.org/Bugs/Public/show_bug.cgi?id=27073 - html/dom hasn't been developed for 3 decades though
16:46
<annevk>
rektide: here's a hint, the Push API is about incoming messages from origin A when there's no open connection to origin A
16:48
<annevk>
zcorpan_: arguably it's a continuum of what came before in the form of SGML and such
16:48
<rektide>
annevk: i get that. i tried to state emphatically that i love the idea that the user-agent could become a server.
16:55
<annevk>
rektide: that's now how any push architecture works, though
16:55
<annevk>
not*
17:02
<rektide>
annevk: i've implemented Pushchannel, a module for node which monkeypatches the SPDY module to create a list of all resources which are SPDY pushed
17:02
<rektide>
annevk: the client when it requests a real resource gets an X-Associated-Content header which contains the list of all pushed resources
17:03
<annevk>
I don't see how that's related to the push concept this draft is discussing
17:10
<annevk>
jgraham: why does Critic not remember I logged in?
17:15
<darobin_>
annevk: you mean after a while you're logged out or it doesn't remember you at all?
17:16
<annevk>
darobin_: logged out
17:16
<annevk>
darobin_: I only have to click "sign in"
17:16
<darobin_>
annevk: because annoyingly the cookie doesn't live long enough
17:16
<annevk>
darobin_: but I don't like clicking that
17:16
<darobin_>
annevk: I asked jgraham and he said the duration is likely hardcoded
17:16
<darobin_>
annevk: I know, I reported the same bug
17:17
<annevk>
okay
17:17
<annevk>
I guess I'll keep clicking then and occasionally bug jgraham because I forgot
17:17
<darobin>
I took a quick look at the Critic source but it wasn't immediately obvious so I didn't look further
17:17
<jgraham>
It's a session cookie I think, and I don't think that's intentional
17:17
<darobin>
yeah, let's just bug jgraham now and then
17:18
<jgraham>
I'm not sure I like this plan :p
17:20
<gsnedders>
jgraham: your fault
17:21
<jgraham>
I know :(
17:32
<Hixie>
zcorpan_: 25 years is "more than 2 decades but not many" :-)
17:33
<jgraham>
Context?
17:39
<zcorpan_>
jgraham: https://www.w3.org/Bugs/Public/show_bug.cgi?id=27073
17:40
<zcorpan_>
Hixie: i think that's stretching it but i guess the statement will be true in 5 years
17:41
<jgraham>
So it should be a promise for a statement? :)
17:43
<jgraham>
If you just drop the word "several" it becomes entirely true
17:44
<jgraham>
and makes the sentence stronger, I think
18:17
<SteveF_>
Domenic: did your email to pf ever get through? if not will follow up
18:17
<Domenic>
SteveF_: I don't think so :(
18:18
<SteveF_>
Domenic: if you want you can send it to me i can forward to get it in and sort out issue
18:18
<MikeSmith>
Domenic, SteveF_ I'll check on it right now
18:19
<SteveF_>
MikeSmith: thanks
18:20
<MikeSmith>
SteveF_: which list is this exactly?
18:21
<SteveF_>
MikeSmith: public-pfwg⊙wo
18:21
<MikeSmith>
k
18:23
<MikeSmith>
SteveF_: so I see nothing in the moderation queue. I suggest pinging Michael Cooper and asking him to approve it. He's the owner of the list and he gets copies of all messages that get held for moderation and he can forward them
18:23
<SteveF_>
MikeSmith: will do
18:47
<SteveF_>
Domenic: email sent to m cooper, ccd mike
18:47
<MikeSmith>
thanks
18:47
<Domenic>
SteveF_: thanks. I can definitely re-send if necessary.
19:01
<Domenic>
oh boy. so re-send to -comments?
20:00
<SteveF_>
Domenic: yeah that's BS but hey...
20:01
<SteveF_>
MikeSmith: what are other WG's policies on mails from non members?
20:05
<Hixie>
annevk: yeah, i still get a 500 on the twitter script. Not that I'm especially happy with Twitter these days so *shrug*.
20:10
<SteveF_>
Domenic: sorry mispelt your name in my response to MC's email
20:10
<Domenic>
SteveF_: no problem! thanks for the help.
20:17
<MikeSmith>
SteveF_: other groups have a policy called "common sense"
20:18
<MikeSmith>
but this group as usual apparently has some separate policy of its sown
20:19
<SteveF_>
MikeSmith: right
20:19
SteveF_
needs to chnage
20:20
<jgraham>
MikeSmith: Now people are probably wondering why I'm giggling
20:20
<MikeSmith>
heh
20:46
<annevk>
Hixie: bah, I wonder what is going wrong
20:50
<Domenic>
for a method declared f(DOMString x, DOMString y, DOMString z), what is f(undefined, undefined, undefined) supposed to do?
20:50
<Domenic>
Stringify to "undefined", or throw?
20:50
<jgraham>
Yes :)
20:50
<jgraham>
I'll look
20:51
<jgraham>
(pretty sure it stringifies, but like I say)
20:51
<Domenic>
I think just stringify to undefined but unsure if the treatment of undefined as args is special...
20:52
<jgraham>
It defers to ES ToString
20:52
<Domenic>
yeah so "undefined"
20:52
<jgraham>
Right
20:52
<jgraham>
(the only case that doesn't is with [TreatNullAs=EmptyString] and null)
20:53
<Domenic>
right, and undefined will trigger defaults I think, but doesn't count as "missing"
21:11
<annevk>
Domenic: you need to use the optional keyword for missing
21:12
<annevk>
Domenic: otherwise it's the same as JS grabbing a reference to the second argument and invoking ToString on it regardless of what it's value is
21:13
<Domenic>
annevk: yeah, just a little weird that f(x) is not the same as f(x, undefined, undefined)
21:13
<Domenic>
JakeA: which is your best service worker video do you think?
21:17
<The-Compiler>
Not sure if this is the right place to ask - does anyone here know about a standard for javascript bookmarklets, i.e. how a browser should interpret them/expect the URL to be encoded?
21:19
<JakeA>
Domenic: http://youtu.be/SmZ9XcTpMS4
21:19
<Domenic>
I saw that one :D
21:24
<The-Compiler>
I mean, it looks like scheme:path, but in the wild it looks like people use javascript:... with ... just being unescaped javascript code, maybe %20 instead of spaces, maybe not, but certainly slashes and all
21:24
<jamesr_>
annevk: should fetch handle the javascript: scheme?
21:24
<jamesr_>
The-Compiler: yeah the handling of the javascript: scheme is very complicated
21:24
<jamesr_>
not sure who's supposed to define it
21:25
<jamesr_>
ah, think it's in HTML
21:25
<jamesr_>
The-Compiler: https://html.spec.whatwg.org/multipage/browsers.html#navigating-across-documents see step 15
21:25
<The-Compiler>
I'm writing a browser and got a feature request to implement javascript bookmarklets, and I wonder what the most sensible way to handle them is :)
21:27
<The-Compiler>
jamesr_: perfect, that's what I was searching for
21:28
<jamesr_>
The-Compiler: plug your nose and do *exactly* what that says
21:28
<The-Compiler>
I love you guys for readable standards which read like pseudocode :)
21:29
<JakeA>
Domenic: haha yeah, I'm afraid that's my best
21:29
<JakeA>
Might do a better one on Friday. It's certainly longer.
21:30
<Domenic>
:)
21:30
<Domenic>
short is nice to point people to
21:32
<zcorpan>
The-Compiler: just avoid letting users run javascript: from the address bar since that's a social engineering attack :-)
21:35
<The-Compiler>
zcorpan: works in Chrome/Firefox though :P
21:35
<The-Compiler>
(okay, it strips the javascript: when pasting)
21:37
<caitp>
would a browser follow a 302 to a javascript: url?
21:37
<caitp>
that would be scary
21:37
<caitp>
i mean it would be hard for it to do any real harm
21:37
<The-Compiler>
caitp: why? How's that different from a page just executing the javascript directly?
21:38
<caitp>
because you could go to your bank's host, which you trust
21:38
<The-Compiler>
(okay, if scripts are blocked somehow that shouldn't be a loophole, but except of that)
21:38
<caitp>
but you've got a weird dns record which takes you somewhere hostile, wheresomeone plans to open infinite modal dialogs just to annoy you
21:39
<caitp>
hey I'm not saying it's a very compelling threat model
21:42
<Domenic>
How do I unsubscribe my (old) email from whatwg@. Can I assume whatwg-unsubscribe⊙wo works?
22:20
<zewt>
where did this fad of not showing mouse cursor spinners in browsers come from
22:21
<annevk>
jamesr_: the idea is that the navigate algorithm handles it
22:21
<zcorpan>
The-Compiler: yeah the attack is copy/paste into address bar
22:21
<annevk>
jamesr_: if javascript ever reaches Fetch it'll just fail
22:22
<annevk>
jamesr_: mind you, implementations might not have this strategy yet