00:18
<zewt>
wow, cancelling mousedown in chrome is being really strange
00:18
<zewt>
causes a click to be registered if the mouseup happens while the mouse is over the element; no "minimum drag distance" click cancels happen at all
00:19
<zewt>
new one for me
00:20
<tantek>
Thanks hober.
00:20
tantek
smiles at https://www.w3.org/Bugs/Public/show_bug.cgi?id=18231#c1
00:22
<tantek>
Hixie, the What you can do write-up seems reasonable, given your preferred methods of working/communication.
00:22
<tantek>
The only thing I'd change is to move this sentence to the start: "The best place to start is to join us on IRC and ask what you can do. "
00:34
<zewt>
eh? it's happening in firefox too--even more bizarre
00:35
<zewt>
maybe I've just never implemented something that's both drag-scrollable and clickable before to never have noticed this, but it's nonsensical behavior
00:55
<Hixie>
tantek: heh, i'm amused at the difference in how he handled it vs how i handled it. :-) https://www.w3.org/Bugs/Public/show_bug.cgi?id=17499
00:56
<Hixie>
tantek: i thought about doing that but i think the best place to start for suggesting a new feature is to describe the problem and post it to the list, not join IRC. joining IRC is more for just participating in existing discussions, which is what the last few paragraphs are about
01:10
<tantek>
Hixie, asking about a general problem area on IRC may be faster than spending time writing up a more thorough description (what people tend to do on email) of the problem and posting it on the list.
01:11
<tantek>
and if it's not already known problem area, a bit of back/forth on IRC from the "regulars" may help a new person better frame/describe their problem in a writeup for the list (or wiki), which should increase efficiency for everyone involved.
01:13
<Hixie>
fair enough
01:14
Hixie
looks at how to phrase it
01:14
<Hixie>
i definitely don't want to encourage people to come to irc before they're thought about what their problem is
01:14
<Hixie>
so it can't be the first step
01:24
<Hixie>
i don't really see how to do it
01:24
<zewt>
Hixie: better than posting to the list before they've thought about it :)
01:28
<Hixie>
zewt: well, the article suggests very strongly that they don't do that, no?
01:29
<zewt>
Hixie: but putting irc above or below the mailing list shouldn't change that very much
01:30
<Hixie>
i just don't see how to phrase it in a way that doesn't hurt the flow
01:30
<zewt>
(and if people are going to come in with half-formed ideas, better it be on IRC than the list, IMO)
01:30
<tantek>
Hixie, something like, Optional: considering pinging briefly on IRC about your general problem area to see if the folks there might suggest a quick solution, or perhaps offer suggestions for how to better phrase/writeup your problem description for the list (or wiki)
01:31
<tantek>
on that zewt, I have to agree.
01:31
<zewt>
(noise on IRC based on an idea that hasn't been formulated well enough is less likely to trigger a two-week ill-formed discussion than on the list)
01:31
<tantek>
= more efficient
03:07
<Hixie>
i shall have to look into it in the morning
07:14
<Smylers>
From Hixie's mammoth src-set mail: “While the page is loading, the user's roommate starts watching an HD show on Hulu while his neighbour, also using the wife, starts downloading a 4GB file on Bittorrent.”
07:14
<Smylers>
I'd've thought that if his neighbour is using his wife, he's got more to worry about than the speed at which his images are downloading ...
07:14
<jgraham>
MikeSmith: Did you have a chance to look at Web Perf -> github yet?
07:15
<jgraham>
Smylers: Different people have different domestic arragements :)
07:27
<MikeSmith>
jgraham: not yet but will do it later today
07:30
<jgraham>
MikeSmith: Thanks :)
09:18
<odinho>
Hixie: [The how to improve HTML article] That bug link is kinda ugly, yes, but still, is it better to place it behind an url shortener?
09:54
<matjas>
odinho, Hixie: perhaps a more descriptive short URL would be better, e.g. http://html5.org/bug
09:55
<odinho>
matjas: Yea, and one that is controlled by the project itself. -- Not that I don't trust Google or anything, it's just ... not good having single point gatekeepers :]
09:55
matjas
sets up http://mths.be/htmlbug
09:56
<matjas>
(for personal use)
09:56
<matjas>
odinho: exactly
09:56
<odinho>
I even have my own s0.no for short urls when I really needed it. But I stopped using it. Got it when short urls were all the rage back in 2007 or when it was.
09:58
<matjas>
hipster!
09:58
<odinho>
:P
10:02
<jgraham>
odinho: Do you also ride a bike with only one gear?
10:03
<odinho>
jgraham: I looked for one, actually! :D lol
10:03
<odinho>
jgraham: Because it's easier tech, so it might not be too prone to breakage. And easy to fix.
10:03
<odinho>
Oh man.
10:03
odinho
hides
14:22
<Wilto>
Hixie: I’m happy to pass that post along to the rest of the ALA team, whenever you’d like.
14:43
<tobie>
Hixie: holding a small informal meet up on AppCache woes next week (yeah, I know, it's a pet peeve of mine). What kind of output would be most useful to you, if any?
14:47
<Ms2ger>
w3c: 'XForms 2.0' and 'XForms 2.0: XPath expression module' Drafts Published http://t.co/QnC7c4Mw
15:35
<dglazkov>
good morning, Whatwg!
15:38
<odinho>
good evening! :D
15:46
<garciawebdev>
good morning
15:46
<garciawebdev>
are PUT and DELETE HTTP methods supported by most modern browsers and IE7+ ?
17:24
<hober>
I take it at this point that the Anolis repo at hg.gsnedders.com is not the most recent version that people actually use?
17:24
<Ms2ger>
Nope
17:24
<Ms2ger>
https://bitbucket.org/ms2ger/anolis
17:25
<hober>
word.
17:26
<Ms2ger>
There's some more documentation somewhere on wiki.whatwg.org
18:10
Ms2ger
waves at Geoffrey Sneddon (New to Bugzilla)
18:56
<gsnedders>
Ms2ger: I'm new to Bugzilla?
18:56
<Ms2ger>
Apparently :)
18:57
<gsnedders>
Interesting, given my account is almost 8 years old. :)
19:41
<Hixie>
odinho, matjas: good point, will change that url
19:41
<Hixie>
Wilto: zeldman gave me an e-mail address to use, so i'm good. but thanks.
19:41
<Hixie>
tobie: use case descriptions that don't mention appcache :-)
19:48
<Hixie>
that wifi/wife typo is really weird -- i and e aren't near each other on any keyboards that i used to type that, and it's not like i type the word "wife", like, ever (certainly not as often as wifi), so i don't see why there'd be muscle memory!
19:48
<Hixie>
funny though
19:49
<Ms2ger>
Freudian slip?
19:51
<Hixie>
that i want to use the neighbour's wife? my neighbour isn't married, as far as i know.
19:54
<Hixie>
tantek: i added an earlier mention of IRC at http://wiki.whatwg.org/wiki/What_you_can_do#How_you_can_improve_HTML -- what do you think?
20:05
<Ms2ger>
Bah
20:06
<Ms2ger>
This microsoft test uses video.startTime
20:06
<Ms2ger>
Which apparently doesn't exist
20:37
<jgraham>
Hixie: It reads like you mentioned IRC twice even though you didn't quite
20:38
<jgraham>
Also, you should probably mention that real-world examples of people working around the lack of a feature are useful
20:38
<jgraham>
i.e. people addressing the use case using existing technologies
20:58
<Yuhong>
http://news.ycombinator.com/item?id=4336924
20:58
<Yuhong>
On the HTML5 buzzword.
21:07
<jarek>
it's not a buzzword, it's a marketing term
21:09
<Yuhong>
And a misleading one at that.
21:09
<jarek>
http://en.wikipedia.org/wiki/Buzzword
21:09
<jarek>
actually, I think you are right :)
21:20
<tantek>
Hixie, that's pretty good.
21:20
<tantek>
I'd more strongly emphasize key adjectives in the opening "First, " sentence
21:20
<tantek>
e.g. right now you have: "First, write a description of the problem"
21:20
<tantek>
I'd say something like: "First, write a brief user-centric description of the problem."
21:20
<tantek>
I think what you're calling "high-level", I'd call user-centric
21:21
<tantek>
for example, "need a native triple-store" reads like a high-level description of a problem
21:21
<tantek>
but has nothing to do with the user
21:21
<tantek>
hence I prefer the term user-centric rather than high-level for these things
21:22
<tantek>
this is certainly very good: "0. Discuss the topic on our IRC channel (#whatwg on Freenode) to see if you've missed anything obvious "
21:24
<tantek>
interestingly enough, that page seems to assume people already have some experience with HTML - which is a reasonable assumption
21:25
<tantek>
though I bet a lot of folks coming up with feature requests don't realize there's already something in HTML (or CSS, or JS, or WebAPIs) that satisfies their use case
21:25
<tantek>
with microformats, we've had to make that implicit step explicit: http://microformats.org/wiki/process#Get_some_experience_first
21:26
<gavinc>
Perhaps a link to a web interface to IRC?
21:26
<gavinc>
as otherwise step 0 may be a large one
21:27
<tantek>
the subsequent section, "Why?" http://microformats.org/wiki/process#Why.3F is similar to the start of "How you can improve HTML "
21:27
<tantek>
gavinc - excellent idea - feel free to improve http://wiki.whatwg.org/wiki/Irc accordingly
21:28
<tantek>
(now linked from that step 0)
21:28
<Wilto>
Would ALA editor feedback be helpful at this point?
21:29
<tantek>
gavinc - also, feel free to copy from http://microformats.org/wiki/irc#Quick_Start which provides some minimal setup/web interface info
21:36
<Yuhong>
tantek: Noticed I mentioned your suggestion to modularize HTML on HN.
21:36
<tantek>
Yuhong - thanks! Hopefully we'll get some additional discussion.
21:37
<Yuhong>
tantek: Based on the fact that the "HTML5" buzzword will not disappear tomorrow.
21:37
<Yuhong>
http://news.ycombinator.com/item?id=4336924
21:37
<tantek>
so we might as well take advantage of that
21:37
<tantek>
Yuhong, also, since you're here often enough, add yourself to http://wiki.whatwg.org/wiki/IRC
21:44
<gavinc>
tantek: will do
21:54
<tobie>
Hixie: that's exactly what I had in mind.
22:19
<Hixie>
tantek, jgraham: updated, thanks.
22:19
<Hixie>
tobie: cool
22:19
<Hixie>
Wilto: sure
23:03
<Hixie>
wait, the HTML test suite isn't MIT licensed?
23:07
<zewt>
heh <zewt> jgraham: are there any requirements on test licensing to make sure the tests can be reused? (eg. by whatwg)
23:08
<Hixie>
forget whatwg, what about browser vendors??
23:09
<zewt>
they don't necessarily even need a license to be able to run the tests, but you definitely need one to modify them
23:09
<Hixie>
browser vendors import the tests into their repos (which means they need a license to copy them) and modify them to fit their needs (so, they need to modify them)
23:10
<Hixie>
ironically, we'd need less of a license since we'd only need to describe diffs
23:10
<Hixie>
which don't necessarily require anything beyond fair use
23:10
<Hixie>
if that
23:11
<zewt>
a widely-debated topic, from what i recall
23:11
<zewt>
(recalling good old DJB and his "no modifications allowed, not even patches" licensing)
23:12
<Hixie>
many things are widely debated
23:12
<zewt>
that is so
23:12
<Hixie>
e.g. whether humans affect the climate
23:12
<Hixie>
whether species evolve from each other
23:12
<Hixie>
doesn't mean there's any actual question as to the matter :-)
23:13
<zewt>
you're aware "fair use" doesn't even exist in many countries, right? :)
23:13
<Hixie>
i am
23:17
<tantek>
the test suites should be placed in the public domain (per CC0) just like everything else.
23:17
<tantek>
(in standards)
23:17
<Hixie>
that's what i had assumed was the case, but recent e-mails suggest it may not be
23:18
<Hixie>
tantek: btw i updated the wiki page, let me know if you have further comments. Regarding your comment about it sounding like it was aimed at web devs rather than users: it is; specifically, it's intended for a list apart.
23:24
<tantek>
Hixie - looks good
23:24
<tantek>
any further iteration can happen lazily / as-needed
23:27
<tantek>
is there an ALA article in the works about "How to improve HTML?"
23:28
<Hixie>
yes, that wiki section
23:29
<Hixie>
heycam|away: ping (re: making the concept of 'global environment' broader than just for JS in webidl, so that I don't have to be JS-specific in defining the three kinds of global environments in the web)
23:36
<tantek>
Hixie, re: PD(CC0) test cases, perhaps we could make such an encouragement (to make test cases all explicilty PD/CC0) on the wiki page: http://wiki.whatwg.org/wiki/Test_cases
23:37
<Hixie>
i doubt that wiki page would have any impact on the relevant folk :-)
23:37
<Hixie>
but feel free to do so
23:38
<tantek>
and http://wiki.whatwg.org/wiki/Testsuite
23:38
<tantek>
i find it hard at times to guess a priori who the "relevant folk" are
23:38
<tantek>
by posting things on a wiki page, it increases the chance that new folks will pick up the suggestions and incorporate them
23:39
<tantek>
ok
23:39
<Hixie>
relevant folk in this case is whoever at the w3c is in charge of picking a license
23:40
<tantek>
I'd rather that individuals just post PD/CC0 test cases to the web, and then anyone can aggregate/curate them, no need for W3C to be a bottleneck
23:41
<Hixie>
the gap between what i'd rather be happening here and what is actually happening here is a vast chasm that rivals some of the most impressive natural canyons known to man
23:41
<Hixie>
but yes, in principle i completely agree
23:41
<tantek>
sounds good, I'll make a note of it
23:44
<tantek>
Hixie, all the CSSWG tests are available under 3-clause BSD in addition to whatever bespoke W3C license: http://www.w3.org/Style/CSS/Test/Overview.en.html
23:44
<tantek>
so there is that
23:45
<Hixie>
i thought the html tests were under the mit license
23:45
<Hixie>
glad at least one group is doing things right
23:45
<tantek>
I suppose I could push to switch from BSD to PD/CC0 and see what happens
23:45
<Hixie>
there's essentially no important difference between pd/cc0, bsd, and mit
23:46
<Hixie>
i prefer mit because it's shorter, personally
23:46
<tantek>
I prefer CC0 because it's more internationalized
23:47
<tantek>
and shorter (by reference)
23:47
<Hixie>
"shorter by reference"?
23:51
<tantek>
yes, it's a one sentence declaration that incorporates the CC0 declaration by hyperlink (with rel=license)
23:51
<tantek>
no need to include all the mumbo jumbo in every document.
23:52
<Hixie>
well you can do the same for the MIT license, that's a non-issue. In fact you don't even need to mention the MIT license, so long as it's in the relevant package somewhere.
23:52
<zewt>
(personally I always avoid linking licenses, since I can't prove that the thing it points to will never change, or always be there)
23:54
<tantek>
zewt - seems to have worked ok for Creative Commons for 8+ years. let me know when you hear of an actual (rather than theoretical) problem of that occurring.
23:54
<zewt>
uh huh, because sites never go down
23:55
<zewt>
be sloppy with licensing matters if you like
23:55
<tantek>
just saying in practice it has not posed a problem for CC
23:55
<tantek>
so let me know when you hear of an actual problem
23:55
<tantek>
rather than handwaving about how problems might happen all the time
23:58
<zewt>
(sorry, I find you consistently rude and talking to you makes me irritable, so I'm going to avoid doing so)
23:59
<tantek>
zewt I'm concisely pointing out flaws with your reasoning, if you think that's rude, perhaps avoid debate altogether.