00:25
<tantek>
yikes - rel-logo was apparently added to the rel-registry 2011-11-24. snuck-in on last year's US Thanksgiving holiday.
00:36
<tantek>
I've written up the reasons why rel=logo is unnecessary and moved it to the "rejected" table in existing-rel-values: http://microformats.org/wiki/existing-rel-values#rejected
00:37
<tantek>
interesting that the notion of linking to a logo has been *twice* reinvented (pavatar, logo)
01:38
<Druide_>
lol
01:38
<Druide_>
[02:37] <josiah> What is the best way to generate a secret for a SF2 application?
01:38
<Druide_>
[02:39] <wiistriker> drop face on keyboard
03:31
<MikeSmith>
roc: you around? I wanted to ask you about something related to the HTML WG DRM video stuff
03:31
<roc>
yes
03:34
<MikeSmith>
roc: the chairs are planning to have a weekly telcon to discuss the DRM proposal and the other for enabling adaptive-streaming use cases
03:34
<MikeSmith>
what I wanted to know is, would you be interested in calling in
03:35
<roc>
my immediate reaction is "urk"
03:35
<MikeSmith>
if so, there's a survey about the call time which I hope can respond to
03:36
<roc>
I don't feel like I have anything to contribute beyond the policy issues we've already discussed on the list
03:36
<MikeSmith>
because otherwise, the call time is likely going to be pretty bad for NZ and Australia too
03:36
<MikeSmith>
heh
03:36
<MikeSmith>
yeah
03:36
<MikeSmith>
well, the survey is here: https://www.w3.org/2002/09/wbs/1/html-media/
03:36
<MikeSmith>
in the current results, Tuesday, 1500 UTC is the time that's leading
03:36
<MikeSmith>
OK
03:37
<roc>
it's basically impossible to schedule a telecon that works for both Europe and me
03:38
<roc>
so if I don't absolutely need to be on it, better that I'm not
03:38
<MikeSmith>
I see
03:40
<MikeSmith>
2100 UTC would work if Europe participants were willing to call in at 11pm
03:40
<roc>
(it is possible, but it's never fun)
03:40
<MikeSmith>
the thing is, what always ends up happening if we do have a telcon is that the New Zealand and Australia and Asia people always get the short end of the stick as far as call times
03:40
<roc>
yep
03:40
<MikeSmith>
yeah, I hear you
03:40
<roc>
small price to pay for the privilege of living here, I say :-)
03:40
<MikeSmith>
indeed
03:41
<MikeSmith>
I'd love to visit there some time
03:41
<roc>
if people wonder whether some specific proposal would meet with my approval, of course I can answer that via email or even dial in to a particular meeting if necessary
03:42
<MikeSmith>
OK
03:44
<MikeSmith>
I'll make an effort personally to give you a heads-up if there's anything on the calls that seems worth alerting you about
03:45
<MikeSmith>
also, I guess the call time is likely to end up amenable to hsivonen being able to call in -- though I'm not sure he'll want to
03:46
<MikeSmith>
I know I don't want to..
03:48
<roc>
heh
06:20
<zcorpan>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17151 isn't this something for the encoding standard?
06:29
<othermaciej>
zcorpan: maybe - which spec defines other types of UTF-16 encoding errors, such as unpaired surrogates, nulls, etc?
06:30
<othermaciej>
oh, this is the case of unpaired surrogate
06:36
<zcorpan>
seems the encoding standard already does "emit a decoder error" for lone surrogates, which for HTML means U+FFFD
06:36
<zcorpan>
(and for XML means fatal error)
06:53
<zcorpan>
TabAtkins: should css3-syntax U+FFFD surrogates in escapes?
07:24
<annevk>
https://twitter.com/awbjs/status/204962858113564672 didn't [tm] do that
07:24
<annevk>
?
07:24
<annevk>
also, cool how twitter now lets you copy permanent URLs and they don't include the hash bang nonsense
07:26
<zcorpan>
nice
07:28
<annevk>
so the differences is that it's a normative copy http://ecma-international.org/ecma-262/5.1/
07:28
<smaug____>
rafaelw_: since I don't know how to update chromium on linux, can't test this right now, but does webkit handle mutationobserver correctly when using documentfragments?
07:28
<annevk>
but apparently has less information on bug reports and such when compared to http://es5.github.com/ which seems logical
07:29
<annevk>
oh "The PDF version is the definitive specification."
07:29
<annevk>
meh
07:30
<zcorpan>
The D in PDF stands for Definitive
07:32
<smaug____>
interesting, I got chromium into a state where it doesn't load anything
07:34
<annevk>
http://w3cmemes.tumblr.com/post/23559522267/http-w3cmemes-tumblr-com how is this foul bachelor frog?
07:34
<othermaciej>
doesn't make sense to me either
07:34
<othermaciej>
maybe Lazy College Senior but even that does quite fit
07:35
<othermaciej>
maybe even a Good Guy Greg
07:35
<annevk>
yeah, GGG would be it I think
07:35
<annevk>
who doesn't want a more readable spec
07:46
<othermaciej>
I guess there is implied laziness to not trying harder to understand it
07:46
<othermaciej>
depends on if the rewrite is easier to understand for everyone, or only the author
07:48
<annevk>
true
08:00
<annevk>
via rniwa http://discuss.joelonsoftware.com/default.asp?joel.3.219431.12
08:00
<annevk>
pretty excellent
08:00
<rniwa>
annevk: yeah. it's via othermaciej
08:01
<annevk>
:)
08:02
<othermaciej>
that pieces is pretty well known, I can't take much credit for knowing about it
08:02
<othermaciej>
but it is funny
08:02
<othermaciej>
and explains why DOM2/3 sucks as bad as it does
08:02
<rniwa>
annevk: and then othermaciej posted http://www.facebook.com/l.php?u=http%3A%2F%2Fcode.google.com%2Fp%2Faxis-ssl%2Fsource%2Fbrowse%2Ftrunk%2Fsrc%2Forg%2Fapache%2Faxis%2Fcomponents%2Fnet%2FSocketFactoryFactory.java&h=VAQGMbw_NAQEG5nvKnT2Shq9fjmmeHTHVR7k9nQHoV4gjRA&enc=AZP-XNPofm9WLTtopsES2fGSbWkXS_7jVMv6HKWGt2J3fiOJmWxlBjPCaK_-IvXSrcOVvK1DdqgYPA3peVIltXdEVoL2ATD4W8hLzQfhe1E5og
08:02
<rniwa>
annevk: ugh... i mean http://code.google.com/p/axis-ssl/source/browse/trunk/src/org/apache/axis/components/net/SocketFactoryFactory.java
08:03
<annevk>
ah thanks
08:03
<othermaciej>
if you do a Google search for Java FactoryFactory you will find lots of real cases of factory factory classes
08:03
<annevk>
I'm afraid of following facebook.com URLs in case they never heard about GET not having side effects
08:04
<annevk>
I'm glad we don't have factories
08:04
<heycam>
the "though you may also have to deploy a few configuration files to make it all work" is the funniest part
08:05
<annevk>
the follow up is great too
08:05
<annevk>
"Luckily, 70% of the workers in the Tool-Oriented Metafactory Union are certified against this version of the spec."
08:06
<heycam>
that is why I write all my web apps in bash
08:43
<annevk>
kind of sad even new drafts such as http://people.mozilla.org/~bbirtles/web-animations/web-animations.html format everything per the old DOM standards...
08:44
<annevk>
and thereby define things in a crappy way, e.g. http://people.mozilla.org/~bbirtles/web-animations/web-animations.html#widl-TimedTemplate-animate-TimedItem-Element-target-float-startTime
08:45
<annevk>
oh, it doesn't even use RFC 2119
08:45
<annevk>
*sigh*
08:48
<othermaciej>
hmmm
08:48
<othermaciej>
it has "must" in notes
08:48
<othermaciej>
"must" in at least one conformance requirement
08:48
<othermaciej>
and then a colored small-caps MUST in another place
08:48
<jgraham>
Tell this to the people writing the spec, perhaps :)
08:49
<othermaciej>
so it's actually much worse than not using RFC2119
08:50
<othermaciej>
so far, the proposed API seems kinda bad too, not just the spec
08:50
<othermaciej>
so not yet worth investing energy in
08:50
<zcorpan>
do we have a wiki with guidelines about how to write conformance requirements correctly and how to format specs in a good way?
08:50
<zcorpan>
s//page/
08:50
<zcorpan>
it seems this comes up time and time again with "new" editors
08:51
<zcorpan>
having new editors is awesome; having specs that suck is not :-)
08:51
<othermaciej>
this spec has at least a couple of non-new editors
08:51
<othermaciej>
http://wiki.whatwg.org/wiki/How_to_write_a_spec
08:52
<jgraham>
We have http://ln.hixie.ch/?start=1140242962&count=1 ofc
08:53
<othermaciej>
oh, that one is also good, too bad it's not in that whatwg wiki page
08:55
<annevk>
I always wonder why people haven't read existing specs...
08:55
<annevk>
e.g. if you have read HTML, XMLHttpRequest, DOM, ... you should know the problems
08:56
<annevk>
but I guess my expectations don't meet reality, so we should indeed have some kind of wiki or document explaining that
08:56
<annevk>
and tell everyone "you're awesome, but take a look at <URL>"
08:57
zcorpan
expanded a bit on rfc2119 usage
08:58
<smaug____>
annevk: "haven't read existing specs"? well, depends on which specs
08:58
<smaug____>
there are tons of different specs
08:58
<smaug____>
HTML, XHR, DOM use similar style
08:59
<smaug____>
and have their problems
08:59
<annevk>
smaug____: everyone writing APIs should be at least familiar with DOM/HTML/IDL imo
08:59
<smaug____>
DOM/IDL yes,
08:59
<smaug____>
and some HTML
08:59
<smaug____>
reading and understanding full HTML...
08:59
<smaug____>
no one does that :)
08:59
<othermaciej>
annevk: are you expecting people to just be generally familiar with those technologies, or to understand in depth how they are specced, why they are specced that way, what problems you could run into building on their specs, etc?
09:01
<annevk>
othermaciej: I expect them to see that the way they define methods is radically different (and has holes); I expect them to look for patterns in existing methods with respect to exception handling, event dispatching, etc.
09:01
<annevk>
othermaciej: my expectations however seem unrealistic
09:01
<annevk>
e.g. quite some specs that clearly require use of the event loop don't use it
09:02
<annevk>
(some started using it after it being pointed out)
09:02
<zcorpan>
http://wiki.whatwg.org/wiki/How_to_write_a_spec doesn't mention the event loop
09:03
<othermaciej>
many people writing web platform specs are not expects on other web platform specs or necessarily even the other relevant technologies at all
09:10
<roc>
or in some cases, not even familiar with what a spec is
09:11
<zcorpan>
maybe the respec tool should have a "read this before use" sticker
09:12
<roc>
annevk: email your gripes to Brian, I'm sure he'll take them on board
09:13
<annevk>
zcorpan: I guess we should expand that wiki page with "Patterns" of some kind; examples of how to do various things
09:13
<annevk>
zcorpan: including dispatching events asynchronously
09:14
<zcorpan>
annevk: sounds like a good idea
09:14
<zcorpan>
annevk: also explicitly say that the old DOM specs is a bad pattern and why it is
09:15
<zcorpan>
annevk: it could point to real-world examples in actual specs and discuss what it does
09:42
<roc>
hmm, why is the From-Origin spec stalled?
09:43
<annevk>
roc: I updated it yesterday to include a reason
09:44
<annevk>
roc: it doesn't mean much though, if people want to implement it, I'll keep updating it and we'll go back to REC-track
09:44
<roc>
is the reason here? http://www.w3.org/TR/from-origin/ I can't see it
09:44
<annevk>
roc: http://dvcs.w3.org/hg/from-origin/raw-file/tip/Overview.html
09:45
<annevk>
roc: TR/ is never up-to-date
09:45
<roc>
sorry, I couldn't find the real version :-)
09:45
<roc>
I wonder if I can teach google to simply not show me TRs
09:46
<annevk>
TR/ will be updated next week with http://dvcs.w3.org/hg/from-origin/raw-file/tip/TR.html fwiw
09:46
<annevk>
I created that yesterday too per request from Art
09:49
<othermaciej>
are canvas gradients new?
09:49
<othermaciej>
I have a hard time remembering which are the new bits
09:50
<annevk>
they're not new
09:50
<annevk>
see http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2012-March/035239.html for what's new
09:51
<othermaciej>
thanks!
09:52
<jgraham>
So, when's Hixie back? (I assume he is away)
09:55
<othermaciej>
he's on vacation, I forgot when back
09:57
<jgraham>
It is curiosuly unsettling to be trapping ina buggy part of the spec without the Dear Leader around for guidance
09:57
<zcorpan>
jgraham: about <script src>, does it await a stable state before fetching src (when setting src), in case the script also sets crossorigin? (like <img>)
09:58
<zcorpan>
the spec doesn't have crossorigin on script yet, it seems
09:58
<annevk>
jgraham: would be more fun if Hixie took a three-month break :)
09:59
<zcorpan>
OMG WE NEED A NEW EDITOR THAT DOESN'T DO VACATIONS
09:59
<annevk>
jgraham: I guess we should count our blessings, it used to be so much worse: http://ln.hixie.ch/?start=1172653243&count=1
10:00
<jgraham>
zcorpan: We should jsut be glad he relocated to the US where no one does vacations (almost :p)
10:01
<annevk>
so my non-n network seems to perform better outside
10:02
<jgraham>
annevk: I guess with a three month break you would quickly adjust your expectations
10:02
<jgraham>
s/you/one/ ^c.
10:15
<annevk>
in wiki markup, what's the best way to markup some kind of example?
10:19
<gsnedders>
annevk: Woah, that was five years ago.
10:19
<othermaciej>
is Selectors API level 2 still being worked on?
10:19
<othermaciej>
last draft looks old
10:19
<zcorpan>
annevk: <blockquote>? :-)
10:21
<zcorpan>
or <div class="..."> maybe; are there any classes with default styling?
10:22
<annevk>
how does http://wiki.whatwg.org/wiki/How_to_write_a_spec#Patterns look?
10:23
<annevk>
gsnedders: you're getting old man
10:23
<annevk>
the Patterns I wrote down now is a start btw, I plan on adding the exception stuff and such
10:24
<zcorpan>
<</var>
10:24
<annevk>
thanks, though not quite the feedback I was looking for
10:24
<zcorpan>
heh :-) looks good
10:25
<zcorpan>
maybe bake in id="" pattern recommendation
10:28
<smaug____>
does anyone know where to download chromium for linux
10:28
<smaug____>
the latest snapshot
10:28
<smaug____>
(whatever it is called)
10:29
<annevk>
smaug____: http://www.chromium.org/getting-involved/dev-channel#TOC-Linux
10:30
smaug____
was hoping .tar.* packages
10:31
<smaug____>
or .zip
10:35
<MikeSmith>
david_carlisle: I just pushed the rest of the MathML3 updates to http://validator.w3.org/nu/
10:35
<david_carlisle>
MikeSmith: ohh thanks will try
10:36
<MikeSmith>
including that openmath fix
10:36
<MikeSmith>
but really I would rather we just rip out all of the openmath stuff altogether
10:36
<david_carlisle>
yo! one valid mathml spec:-) (could you chnage the identification string on the schema Using the schema for HTML5 + SVG 1.1 + MathML 2.0 + to say 3.0
10:36
<MikeSmith>
ah yeah
10:37
<MikeSmith>
will change that now
10:37
<annevk>
is the validator not supporting post-HTML5 things?
10:37
<MikeSmith>
hmm, I wonder of hsivonen wants it specified or not
10:37
<annevk>
it seems better to have "valid HTML/SVG/MathML" or just "this content is valid"
10:38
<MikeSmith>
yeah
10:38
<MikeSmith>
that's what I meant
10:38
<MikeSmith>
I think hsivonen might now prefer that we don't specify the versions at all
10:39
<david_carlisle>
MikeSmith: Really I would rather there were not any openmath special cases in the schema. I'd prefer that the attribute value be open as it says in the spec, but if you/heri think that's too risky and just want to allow mathml/svg/html in teh annotations I'll just grumble and not inline the openmath bits in that version of the spec
10:39
<MikeSmith>
but in the case of SVG, we have it that way to make sure the people realize we're not supporting SVG 1.2
10:39
<smaug____>
hmm, am I confident enough that MutationObserver is a good API and that the implementation is good enough for un-prefixing...
10:39
<david_carlisle>
apply usual typo correction, especially to Henri's name
10:40
<MikeSmith>
david_carlisle: yeah, I think it's better not to make it open and I think Henri still thinks so as well. We don't leave anything open anywhere else like that.
10:40
<annevk>
smaug____: no, but if you don't start unprefixing we'll never get rid of the much worse thing
10:41
<smaug____>
true
10:41
<david_carlisle>
MikeSmith: I think it;'s a mistake but I'm not going to argue just grumble:-) but as I said either way it shouldn't special case openmath
10:42
<MikeSmith>
OK
10:43
<david_carlisle>
because people who want to use ChemML (say) should be able to look what I did for OpenMath and I don't want them to find that what you have to do is mail you and Henri and ask to be whitelisted
10:46
<MikeSmith>
david_carlisle: ah yeah
10:46
<MikeSmith>
so I will yank the openmath stuff
10:48
<david_carlisle>
MikeSmith: Thanks for all the work on this, appreciated.
10:48
<MikeSmith>
no problem
10:48
<MikeSmith>
sorry for taking so long
11:41
<annevk>
rename proposal "Howto_spec"
11:42
<annevk>
reasoning: easier to remember and search for
11:50
<zcorpan>
what does a Test Facilitator do?
11:50
<annevk>
I think the idea is that similar to an editor, it's the person responsible for making sure the test suite is adequate
11:51
<annevk>
it's kind of lame name though
11:51
<annevk>
but then so is "editor"
11:51
<zcorpan>
Test Suite Adequate Making Surer
11:52
<zcorpan>
+Is
11:52
<gsnedders>
And dealing with test case approval etc.
12:01
<annevk>
http://wiki.whatwg.org/wiki/Howto_spec#Patterns added attributes
12:09
<jgraham>
The way that gecko implements alert is funny
12:09
<jgraham>
I particularly like the way you can make the background get darker and darker as you layer alerts one on top of the other
12:09
<annevk>
is "Legacy DOM" a good name for the wiki page describing the problems with respec's default output?
12:10
<annevk>
or should it be Legacy DOM-style which is kind of annoying to remember and has a dash
12:13
<annevk>
mah I'll just make it a separate heading at the end of "howto spec"
12:21
<annevk>
http://wiki.whatwg.org/wiki/Howto_spec#Legacy_DOM-style
12:44
<matjas>
“ES5 claims the global scope "this" is the same as the global object, which is not always true in HTML5.” — http://wiki.whatwg.org/wiki/Web_ECMAScript#Global_scope
12:44
<matjas>
can haz more info on this? ^
12:45
<annevk>
see Window versus WindowProxy
12:49
<zcorpan>
annevk: awesome
12:51
matjas
tries to understand
12:52
<matjas>
hmm, so when is global this != window?
12:52
<annevk>
matjas: prolly when navigating
12:53
<matjas>
heh
12:53
<annevk>
ah yeah, the spec says so to
12:53
<annevk>
o
12:54
<annevk>
yeah so each Document has a Window
12:54
<annevk>
but each browsing context (e.g. <iframe>) has a WindowProxy
12:56
<jgraham>
zcorpan: So, do you want to be Web Workers test facilitator?
12:56
<zcorpan>
jgraham: i guess i could
12:57
<jgraham>
Excellent
12:57
<jgraham>
I think that makes more sense than me doing it :)
12:58
<zcorpan>
i just need to find that email again so i can reply to it
12:59
<annevk>
if you do work you can break the rules
12:59
<annevk>
I mean, create a new thread :)
13:00
<zcorpan>
""ext James Graham" <jgraham⊙oc>"?
13:00
<jgraham>
??
13:00
<annevk>
some email software inserts "ext"
13:01
<annevk>
prolly an enterprise feature
13:01
<annevk>
sorry, "feature"
13:01
<zcorpan>
i'm as confused as you are. the email i'm replying to has "James Graham" in the To field, but when i was going to reply to it, it became that
13:01
<jgraham>
Maybe M2 became an enterprise mail client :)
13:01
<annevk>
or your stored contacts have been garbled by someone else
13:02
<zcorpan>
oh wait, the email i'm replying to actually had To: "ext James Graham" <jgraham⊙oc>
13:02
<annevk>
from Art?
13:02
<zcorpan>
it's just opera's UI uses the contact name rather than what the email header said
13:03
<zcorpan>
yeah
13:09
<Ms2ger>
Yeah, Art's got lots of "ext"s
13:12
<zcorpan>
should <label><a href=x> be allowed? it seems to be allowed in the spec currently
13:52
<matjas>
annevk: ok, so it’s still safe to rely on global `this === window`, as it will always be the case; be it through a Window or a WindowProxy
13:52
<matjas>
phew
13:53
<zcorpan>
this === windows is always true, but window and this don't always refer to the ES global object (instead they both refer to the WindowProxy)
13:55
<gsnedders>
zcorpan: When do they refer to different things?
13:56
<annevk>
gsnedders: navigation
13:57
<gsnedders>
But when do you have a global existing after a navigation?
13:58
<annevk>
not an ES global, but you can still have a reference to a WindowProxy object
13:58
<annevk>
it just proxies to a different ES global
13:58
<zcorpan>
say you have an iframe, the parent saves a reference to its window, then it navigates
13:59
<smaug____>
hmm, is this===window when doing document.open()
13:59
smaug____
tests
14:00
<zcorpan>
smaug____: document.open() should reuse the Window per spec, IIRC
14:00
<Ms2ger>
Parsing scientific notation in CSS is still undefined?
14:00
<smaug____>
zcorpan: really
14:00
<smaug____>
hmm
14:00
<zcorpan>
Ms2ger: see css3-syntax
14:01
zcorpan
has gotta go
14:01
<Ms2ger>
I don't want to
14:02
<smaug____>
window===this is true even in the case I was thinking
14:02
<smaug____>
ofc
14:02
<smaug____>
because of outer window
14:02
<smaug____>
silly me
14:26
<hsivonen>
webben: You can follow along the Maven stuff in http://bugzilla.validator.nu/show_bug.cgi?id=921 . I landed the OSGi stuff from http://bugzilla.validator.nu/show_bug.cgi?id=919 .
15:27
<webben>
hsivonen: ta
15:39
<jgraham>
Oh look! A new mailing list!
15:41
<Ms2ger>
Needs ski instructor meme
15:42
<dglazkov>
good morning, Whatwg!
15:42
<Ms2ger>
Good, you
15:47
<jgraham>
Ms2ger: +God + again?
15:47
<Ms2ger>
No comment
16:09
<Ms2ger>
dglazkov, sure, as long as we decide in favour of Anolis :)
16:22
<jgraham>
I propose that we settle it by a fight to the death. rberjon vs gsnedders.
16:23
<jgraham>
Although rberjon would win
16:23
<jgraham>
So Ms2ger would be unhappy
16:23
<Ms2ger>
OTOH, no more gsnedders...
16:24
<Ms2ger>
That would be doubleplusungood
16:28
gsnedders
wonders what he's done around #whatwg for, well, years
16:28
<Ms2ger>
py3!
16:28
<Workshiva>
Talk
16:28
<gsnedders>
Ms2ger: Oh yeah.
16:28
<Ms2ger>
And accept one of fantasai's patches after like three years :)
16:29
<jgraham>
dglazkov: Anyway afaict the tradeoff between respec and anolis is that people prefer writing specs using respec, presumably due to the lack of a compile step, but it encourages a worse style of spec writing and has a worse user experience (due to scripts running when the document loads)
16:30
<gsnedders>
jgraham: You could always set up respec to build for each commit
16:30
<gsnedders>
Instead of only doing it for TR
16:31
<jgraham>
gsnedders: Then it would have a compile step
16:31
<gsnedders>
jgraham: But eliminating that for actual work before you commit is probably the big gain anyway.
16:32
<gsnedders>
You can automate it through commit hooks to get all the gain for no effort.
16:33
<jgraham>
gsnedders: For what?
17:18
<TabAtkins>
Man, zcorpan always disappears just as I come in for the day. It's hard to remember to continue a discussion in the evening with him. :/
17:19
<Ms2ger>
Get up earlier :)
17:19
<TabAtkins>
Getting to work at 8am should be early enough for anybody.
17:19
<Ms2ger>
Relocate
17:19
<TabAtkins>
Already did. It made the problem worse.
17:19
<Ms2ger>
Relocate in the right direction
17:20
<TabAtkins>
Man, nothing makes you happy.
17:20
<Ms2ger>
No
17:23
<jwalden>
TabAtkins: you must be new here :-)
17:28
<nights>
TabAtkins: lots of your demos doesnt work.
17:29
<nights>
and if they do work they only work in Chrome...
17:29
<nights>
the live image processing is cool though
17:35
<TabAtkins>
nights: It's very possible that they might require Chrome, because I was lazy and authored them solely for my own use in a presentation, where I know I'll be using Chrome.
17:41
<nights>
I see
17:41
<nights>
:)
17:41
<TabAtkins>
All of the demos on xanthir.com/demos work, though - I jsut tried thema ll.
17:48
<nights>
ok cool! :)
17:48
<nights>
its nasty when things doesnt work in a presentation
17:51
<Ms2ger>
I find it fascinating that other glenn claims to have a hybrid spec tool for cssom, yet hasn't actually ever edited cssom
17:56
<zcorpan>
TabAtkins: you summoned?
18:01
<tantek>
Ms2ger, perhaps you can ask him in irc://irc.w3.org:6665/css - looks like he's even there now.
18:04
<TabAtkins>
zcorpan: Oh, I was just going to say that your suggestion sounds like a good idea.
18:04
<TabAtkins>
Ms2ger: Yes he has. What makes you think otherwise?
18:05
<Ms2ger>
Apart from putting his name on it
18:05
<nights>
TabAtkins: you work at google?
18:05
<TabAtkins>
I dont' know whether he's commited the relevant changes yet, but I know for a fact that he and Shane are working on it.
18:05
<TabAtkins>
nights: Yeah.
18:06
<nights>
TabAtkins: Cool...i actually rejected an offer from them once :)
18:06
<TabAtkins>
Bad move. ^_^
18:06
<Ms2ger>
Good move
18:06
<nights>
now im cofounder of a company making USD -10000 every month
18:06
<Ms2ger>
Come join us at The Rebellion
18:06
<nights>
hehe
18:06
<Ms2ger>
Er, Mozilla
18:07
<TabAtkins>
I can't make much revenue directly from doing what I love, which is working on standards.
18:07
<Ms2ger>
TabAtkins, http://dev.w3.org/csswg/cssom/ <-- this is cssom. He's never actually edited it
18:08
<nights>
TabAtkins: i wrote minus 10000 usd
18:08
<nights>
so im sure you make more than me ;)
18:08
<TabAtkins>
Oh, hahahaha
18:08
<TabAtkins>
I read that as a ~
18:08
<nights>
:)
18:08
<nights>
that time
18:08
<nights>
i started working for an another american company
18:08
<nights>
i regret that move very much :)
18:11
<nights>
google was a few floors down
18:11
<nights>
their office here in tokyo is very cool
18:12
<zcorpan>
TabAtkins: what suggestion?
18:12
<TabAtkins>
zcorpan: To make the parser replace escapes of surrogates with U+fffd
18:12
<TabAtkins>
nights: Agreed! I've been to the tokyo office before.
18:13
<zcorpan>
ah
18:14
<nights>
TabAtkins: the entrance looks like a bath house :)
18:14
<nights>
a few friends of mine work there at the android department
18:14
<nights>
once my company bleed up all our investment i might give it another shot. hope they have no hard feelings :)
18:15
<TabAtkins>
We're generally blind to previous attempts, and encourage people to try again.
18:15
<nights>
actually i didnt respond
18:16
<nights>
so maybe the offer is still hot
18:16
<nights>
:p
18:16
<nights>
been two years
18:16
<TabAtkins>
Then there's definitely no hard feelings. Go for it. ^_^
18:16
<nights>
what do you guys think about still running a business when applying?
18:17
<nights>
maybe that wouldnt look good?
18:17
<TabAtkins>
I have no idea what our policies might be about that.
18:17
<TabAtkins>
I suspect it's problematic to be *working* at another company at the same time. But just working/running one while applying is probably fine?
18:18
<nights>
its essentially bootstrapped
18:18
<smaug____>
nights: there are other (less evil ;) ) companies than the big G.
18:19
<nights>
hehe
18:19
<TabAtkins>
LIES
18:19
<nights>
i worked for the most evil company in the world
18:19
<nights>
so my perception is quite damaged
18:19
<Ms2ger>
Facebook?
18:19
<TabAtkins>
Sony?
18:19
<nights>
actually ive worked for sony
18:19
smaug____
has also background in an evil empire
18:19
<nights>
but that wasnt the company i had in mind
18:19
<nights>
an american bank
18:19
<Ms2ger>
Mordor?
18:20
<smaug____>
Nah, never visited Mordor, though Morgoth was a nice boss
18:20
<nights>
mordor
18:20
<nights>
that must be much nicer than where i went
18:20
<smaug____>
:)
18:20
<nights>
:)
18:21
<nights>
the dark lord at that company basically had a will to dominate all life
18:22
<nights>
so there was similarities
18:22
<nights>
over me he poured his cruelty, his malice and a shitty bonus
18:35
<annevk>
am I missing something or is http://daringfireball.net/2012/05/bigger_display_iphone_thing_wwdc not discussing landscape mode weird?
18:45
<annevk>
about that email problem I had earlier
18:45
<annevk>
just learned of http://sparrowmailapp.com/mac.php via http://www.metafilter.com/116164/Comin-like-a-ghost-town#4358181
18:45
<annevk>
apparently it has support for the Gmail label concept
18:48
<smaug____>
rafaelw_: hey, does the webkit implementation of MutationObserver support DocumentFragment?
18:48
<Ms2ger>
A test!
18:49
<smaug____>
I tested, but only using C19
18:49
<smaug____>
haven't figured out where to download newer build for Linux (.zip/tar.* package, not rpm)
18:49
<smaug____>
C19 didn't pass the test
18:57
<smaug____>
TabAtkins: do you happen to know if there are .zip builds of chromium available ?
18:57
<smaug____>
I do have a .zip for C19
18:57
<smaug____>
but don't remember where I downloaded it
18:57
<smaug____>
linux builds
19:00
<Wilto>
Not gonna lie, guys: I’m pretty pumped about Florian’s compromise proposal.
19:01
<Wilto>
Also: “Florian’s Compromise” would be a great name for either a bar or a scientific theory.
19:01
<Wilto>
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2012-May/036162.html
19:01
<Wilto>
Seems like it covers everyone’s bases.
19:09
<sicking>
annevk: why aren't we adding the "chunked-text" and "chunked-arraybuffer" responsetypes to XHR?
19:10
<annevk>
pretty sure I emailed about that
19:12
<annevk>
so Sparrow does not do bottom posting
19:12
<annevk>
other than that it's pretty neat... :/
19:26
<TabAtkins>
smaug____: No clue.
19:26
<TabAtkins>
annevk: "Does not", or "does not automatically"?
19:27
<annevk>
does not as far as I can tell
19:28
<annevk>
I could not find an option that inserts the signature at the bottom by default and leaves no space at the top
19:28
<TabAtkins>
Oh, okay. That counts as "not automatically" as far as I'm concerned. You can always put your signature down there yourself.
19:28
<annevk>
initially I was afraid plain text was also doomed, but that appears to be a simple setting (which is not accessible via the preferences panel, but whatever)
19:28
<TabAtkins>
Benefit of having a trivial signature.
19:31
<annevk>
yeah sure, I could type the whole response myself including all the quotes, too
19:31
<TabAtkins>
That's a silly extrapolation.
19:40
<dglazkov>
jgraham, Ms2ger: I actually prefer writing specs in just HTML.
19:41
<Ms2ger>
So do I
19:44
<TabAtkins>
I write my specs in a combination of HTML and Spanish.
19:46
<annevk>
dglazkov: you don't use any scripts to generate table of contents and such?
19:46
<annevk>
dglazkov: or cross-specification cross-references?
19:46
<annevk>
dglazkov: because I prefer HTML too and I use Anolis for some of the boring stuff
19:47
<dglazkov>
annevk: I wrote some simple scripts for the most annoying stuff.
19:47
<annevk>
dglazkov: what language did you use?
19:48
<dglazkov>
annevk: Jabba Script
19:48
<annevk>
dglazkov: I might be interested in that, maybe we don't need all of Anolis
19:48
<annevk>
dglazkov: heh
19:49
<dglazkov>
:)
19:53
<dglazkov>
annevk: I need to look at Anolis too. the Pythoney smell took me back at first.
19:54
<dglazkov>
another thing I've been experimenting with is using c9.io for spec-editing.
19:54
<dglazkov>
but it's still pretty wonky
19:55
<annevk>
so currently Shadow DOM is still using the legacy DOM-style
19:55
<dglazkov>
who is bjoern hoehrmann?
19:55
<annevk>
I take it that's because of Jabba Script?
19:55
<annevk>
dglazkov: http://thebjoernhoehrmannproject.org/
19:55
<dglazkov>
no, that's because I stole the style from respec
19:56
<annevk>
dglazkov: student/ex-student? from Germany; not much else is known :)
19:56
<annevk>
a Ms2ger with an actual name and slightly different location within Europe
19:57
<Ms2ger>
Also, I don't care for Frisian
19:57
<annevk>
and when you troll it's never subtle :p
19:57
<Ms2ger>
Oh, no
19:57
<TabAtkins>
dglazkov: I don't know, but on CSS issues he's a troll.
19:58
<dglazkov>
GMail needs a troll hat feature.
19:59
<Ms2ger>
He's the only person who ever got a W3C spec rescinded, afaik
19:59
<TabAtkins>
Wow, which one?
19:59
<TabAtkins>
Also: how?
19:59
<Ms2ger>
"Spiderman and the XHTML Kindergarten" / http://lists.w3.org/Archives/Public/www-archive/2009May/0029.html
20:00
<Ms2ger>
They republished anyway soon after, but still
20:01
<annevk>
he also once submitted 300-500 emails about issues in SVG
20:02
<Ms2ger>
I also rather enjoyed http://lists.w3.org/Archives/Public/www-archive/2012Apr/0072.html
20:02
<annevk>
oh sweet
20:02
<annevk>
smaug____ went ahead
20:02
<annevk>
http://hg.mozilla.org/mozilla-central/rev/e54a85233701
20:02
<annevk>
yay smaug____
20:03
<TabAtkins>
smaug____++
20:03
<dglazkov>
smaug____++ indeed
20:03
<smaug____>
waiting for webkit to unprefix too :)
20:05
<annevk>
Ms2ger: agreed, updated http://thebjoernhoehrmannproject.org/
20:05
<Ms2ger>
\o/
20:05
<Ms2ger>
Now I really accomplished something :)
20:08
<annevk>
hmm
20:09
<annevk>
maybe I should use some prefixes in that XML document
20:09
<annevk>
Google seems to badly cheat
20:09
<gsnedders>
Ideally in such a way as to expose browser bugs.
20:09
<annevk>
gsnedders: are there still browser bugs with prefixes?
20:09
<annevk>
gsnedders: other than Opera applying html:class or some such?
20:12
<gsnedders>
annevk: I dunno.
20:12
<gsnedders>
annevk: That Opera bug is the only one I know of off hand, but I know our bugs better than anyone else's.
20:12
<annevk>
gsnedders: don't you learn anything at uni? :p
20:12
<gsnedders>
annevk: Pretty much, no.
20:14
<annevk>
btw
20:14
<annevk>
the stuff about layout
20:14
<annevk>
that does not actually apply to media queries afaik
20:14
<annevk>
the whole point about media queries is that they don't depend on layout
20:14
<annevk>
the font-size preference from the user (defaulting to 16px) is known way before layout
20:15
<annevk>
and that's the only font-size 'em' in media queries depends upon
20:15
<annevk>
same for 'rem'
20:15
<annevk>
and 'ex', etc.
20:35
<jgraham>
dglazkov: I write specs in blood myself
20:35
<jgraham>
Tony Wilson style
20:36
<gsnedders>
Next you'll be documenting how to open the chamber of secrets…
20:36
<dglazkov>
jgraham: that's why they are always called First Public Working Draw
20:41
Ms2ger
hisses "Open"
20:43
<jgraham>
Ms2ger is a parselmouth?
20:43
<Ms2ger>
There have been precedents of non-parselmouths opening the chamber of secrets
20:44
<dglazkov>
hey, Harry Potter is a fellow standards wonk! He has specs!
20:45
<TabAtkins>
Boooooo
20:45
<dglazkov>
:P
20:46
<jgraham>
Ms2ger: Indeed. Notably Ron Weasley.
20:46
<Ms2ger>
dglazkov, wow
20:46
<Ms2ger>
That was pretty bad
20:46
<dglazkov>
so bad it's good?
20:46
<jgraham>
The other kind
20:46
<Ms2ger>
No
20:47
<Ms2ger>
Afraid not
20:47
<dglazkov>
:D
20:51
<WeirdAl>
oh, Harry Potter. :|
20:52
<dglazkov>
speaking of Ron Weasley, how do we define HTML interfaces from JavaScript perspective? For example, what is the distinction between HTMLElement and any object that implements this interface?
20:52
<dglazkov>
magic hosted objects aside
20:52
<WeirdAl>
testability?
20:53
<dglazkov>
both document.createElement('div') and window.HTMLDivElement are just objects that have HTMLElement in their prototype chain
20:53
<TabAtkins>
document.createElement('div') creates an element with HTMLDivElement as its prototype.
20:53
<dglazkov>
window.HTMLDivElement.prototype that is
20:53
<TabAtkins>
HTMLDivElement has HTMLElement in its prototype.
20:55
<dglazkov>
right. But from JS perspective, they are really not special in any way. There's no such thing as an "interface" in JS. So how do we get around this in DOM specs?
20:56
<Ms2ger>
What do you mean, get around this in DOM specs?
20:56
<Ms2ger>
What are you trying to do?
20:57
<dglazkov>
darn it, gotta run
20:57
<dglazkov>
bbiab
21:04
<TabAtkins>
dglazkov: Something is an "interface" if it doesn't have a constructor.