| 00:00 | <zewt> | heh i just went "where's ms2ger" about 30 minutes ago as I went to give a partial answer to his question |
| 00:01 | <zewt> | (Array.map) |
| 08:42 | <foolip> | roc, my inbox is aware of the problem :) |
| 08:46 | <Ms2ger> | zewt, sure, but that's pure JS, how about DOM callbacks? :) |
| 08:57 | <zcorpan> | hsivonen: http://simon.html5.org/test/css/css-charset/page-utf16-css-bomless-utf16be.html - seems chrome inherits utf-16 even when the stylesheet is bomless utf-16 big endian, but opera and firefox detect it as big endian and apply the style sheet |
| 08:58 | <hsivonen> | :-( BOMless UTF-16. I wonder how much of that exists |
| 08:59 | <hsivonen> | should BOMless UTF-16 detection happen right after BOM detection in the prededence order |
| 09:00 | <zcorpan> | get we get away with it not happening? |
| 09:01 | <zcorpan> | i guess i'll check the tests in to csswg repo now and send an email to www-style |
| 09:01 | <hsivonen> | maybe we could. the magnitude of the BOMless UTF-16 problem is all hearsay lore to me |
| 09:02 | <hsivonen> | and I'm generally very skeptical of lore of the type "weird things happen in Japan" |
| 09:03 | <Ms2ger> | Weird things do happen in Japan, have you seen their game shows? |
| 09:06 | <tantek> | greetings - anyone else here in Lyon? |
| 09:13 | <SimonSapin> | tantek: yes, and 2 seats on your left :) |
| 09:23 | <annevk> | I'm loving https://o.twimg.com/1/proxy.jpg?t=FQQVBBg9aHR0cDovLzI1Lm1lZGlhLnR1bWJsci5jb20vdHVtYmxyX21jcDdrclVXbG0xcnZzYmg5bzFfNDAwLmpwZxQCFgASAA&s=SEoPugLBeIQN508LZx8yZdpKfW-5LTFsMuzqsQ2JgLs |
| 09:29 | <annevk> | without context you instantly know it's about the CSS WG |
| 09:30 | <annevk> | although maybe that's because nobody else really cares about "RESOLUTION" is much as they do |
| 09:31 | <hober> | annevk: :) |
| 09:31 | <hober> | annevk: you could come upstairs and soak up the meme material |
| 09:32 | <MikeSmith> | become a mutation observer |
| 10:05 | <jgraham> | MikeSmith: Do you remember which websockets servers we have on w3c-test.org |
| 10:05 | <jgraham> | ? |
| 10:06 | <jgraham> | We have some pywebsockets stuff that we want to submit |
| 10:40 | <tantek> | hello SimonSapin! |
| 10:48 | <divya> | yeah why dont u join the csswg as a joke annevk |
| 10:48 | <divya> | it would be fun. |
| 10:48 | <divya> | or a world war. |
| 10:48 | <divya> | either way entertainment would be primary |
| 10:48 | <annevk> | I'm in the CSS WG by proxy |
| 10:49 | <divya> | who is your proxy |
| 10:49 | <divya> | hober? |
| 10:49 | <annevk> | my latest spec is http://html5.org/temp/cssencoding.txt and it already has a test suite and a "CSS spec" |
| 10:49 | <divya> | UH OH |
| 10:49 | <divya> | paging glazou |
| 10:49 | <annevk> | o_O |
| 10:50 | <divya> | (you missed yesterday's battle with macej) |
| 10:51 | <annevk> | also the Member-only email (although I guess not really, since it leaked somehow to me) |
| 10:51 | <annevk> | I'm happy to miss out on these things and amuse myself by reading w3cmemes |
| 10:54 | <othermaciej> | what Member-only email? |
| 10:55 | <othermaciej> | divya: I still don't understand why there was a battle in the first place there |
| 10:56 | <divya> | some people felt slighted othermaciej |
| 10:56 | <divya> | thats all. |
| 10:56 | <divya> | but i dunno, i am essentially a passive observer. |
| 10:56 | <annevk> | this WebApps discussion is such a waste of time |
| 10:56 | <annevk> | almost asked for the mike to say WHATWG |
| 10:56 | <divya> | plz dont say that annevk now someone will tweet this. |
| 10:56 | <annevk> | but fixing URLs seems a better use of time |
| 10:57 | <divya> | "WHATWG CLIQUE HATES WEBAPPS" |
| 10:57 | <annevk> | people tweet a lot of things |
| 10:57 | <divya> | NEWS AT 11 |
| 10:57 | <annevk> | it's Flm at 11 |
| 10:57 | <annevk> | Film* |
| 10:57 | <annevk> | !11! |
| 10:57 | divya | adopts a roman orator pose for argument |
| 10:57 | <othermaciej> | divya: I do not understand the part where you solve a cross-wg communication problem by refusing to communicate and then ambushing the chairs of the other wg |
| 10:58 | <divya> | well you all walked into the trap othermaciej :))) |
| 10:58 | <divya> | it should have been a htmlxcss meeting like it happens with svgxcss |
| 10:59 | <hober> | othermaciej: that sounds like meme fodder right there... |
| 10:59 | <othermaciej> | hober: its a traaaapppp |
| 11:00 | <divya> | ¡11! |
| 11:00 | <othermaciej> | divya: yeah, next time I'm asked to make a surprise appearance of another WG I am probably going to say no |
| 11:00 | <divya> | i think you should ask for a combined meeting othermaciej i thought it worked well for css and svg |
| 11:01 | <divya> | omg the memes keep pouring |
| 11:03 | <divya> | https://twitter.com/sgalineau/status/263234033985679362 |
| 11:06 | <zcorpan> | ok i've submitted the css tests and sent an email. http://lists.w3.org/Archives/Public/www-style/2012Oct/0833.html |
| 11:06 | <divya> | othermaciej: http://w3cmemes.tumblr.com/image/34630866378 |
| 11:07 | <divya> | good work all these w3cmeme creators. |
| 11:07 | <othermaciej> | divya: true |
| 11:07 | <hsivonen> | would you rather be ambushed by ambiguity or by CSS chairs? |
| 11:08 | hober | notes that or is not xor: one of the css chairs is a tag member :) |
| 11:14 | <annevk> | zcorpan: note that step 5 was already changed yesterday |
| 11:14 | <annevk> | zcorpan: per discussion between hsivonen and I |
| 11:15 | <zcorpan> | ah. i hadn't refreshed. |
| 11:51 | <MikeSmith> | jgraham: (when your back) we have mod_pywebsccket on w3c-test.org |
| 11:58 | <AryehGregor> | annevk, I study full-time and work like five hours a week average. No chance I'm going to any conferences. |
| 13:03 | <divya> | wow it is othermaciej time on w3cmemes |
| 13:42 | <karlcow> | hmm replies to digest… sad panda |
| 14:12 | <zewt> | notherems2: there should be no inherent difference between callback-based APIs that happen to be defined by JS or that happen to be defined by DOM APIs |
| 14:20 | <sigwhite> | Is there an API to access the device's network speed (e.g. 3G, 4G, LTS) yet? |
| 14:20 | <jgraham> | No |
| 14:20 | <sigwhite> | Thanks. You rock. |
| 14:20 | <jgraham> | What would you use that for? |
| 14:20 | <sigwhite> | No idea. Someone asked me. I figured I might as well go to the source. |
| 14:20 | <jgraham> | OK |
| 14:26 | <GPHemsley> | annevk: Don't forget to handle these schemes: http://tools.ietf.org/html/rfc2324#section-3 |
| 14:28 | <zewt> | (not that "3G" actually tells you what the speed is...) |
| 14:40 | <zewt> | RIM's inspiring appearance on whatwg suggests bright, innovative future |
| 14:41 | <GPHemsley> | Just found this (from 2010): http://www.w3.org/2010/11/TPAC/HTMLnext-perspectives.pdf |
| 14:41 | <GPHemsley> | Can't tell if the colors are supposed to be meaningful |
| 14:48 | <GPHemsley> | abarth: Do you recall if the order of the list of type patterns is significant in mimesniff? |
| 14:49 | <GPHemsley> | abarth: (There doesn't appear to be much overlap between the patterns that could result in a precedence conflict, IMO.) |
| 15:07 | <annevk> | I don't really get brendan on public-script-coord |
| 15:07 | <annevk> | but I guess I'll let heycam solve it now he's at least okay with having it |
| 15:08 | <annevk> | AryehGregor: I had forgotten you were studying, my bad |
| 15:09 | <annevk> | GPHemsley: a lot of those URL schemes can never be URL schemes |
| 15:09 | <annevk> | GPHemsley: the person that wrote that RFC has no understanding of the URI/IRI syntax rules |
| 15:09 | <annevk> | GPHemsley: poor joke :p |
| 15:10 | GPHemsley | points to the author of http://tools.ietf.org/html/rfc1738 |
| 15:12 | <annevk> | heh |
| 15:14 | <annevk> | http://w3cmemes.tumblr.com/post/34633601085/grumpy-old-maciej-has-a-question-about-your-spec oh god |
| 15:14 | <annevk> | w3cmemes is on a roll |
| 15:17 | <zcorpan> | the only case where the websocket api now gives an 'error' event, afaict, is: "If the user agent was required to fail the WebSocket connection or the WebSocket connection is closed with prejudice, fire a simple event named error at the WebSocket object. [WSP]" |
| 15:17 | <zcorpan> | though i guess that gets invoked by various things |
| 15:19 | <zcorpan> | yeah |
| 15:26 | <GPHemsley> | I'm still bothered by the fact that people think that the WHATWG has somehow "subverted the standards process"... |
| 15:31 | <zewt> | what's wrong with subverting a damaged and damaging process and replacing it with one that works? |
| 15:31 | <GPHemsley> | Hmm... this log somehow got duplicated: http://logbot.glob.com.au/?c=freenode%23whatwg&s=13+Mar+2012&e=13+Mar+2012 |
| 15:55 | <MikeSmith> | tantek: are you in the webapps meeting room? |
| 15:55 | <MikeSmith> | Greg Billock is talking about Web Intents |
| 15:56 | <MikeSmith> | ah you are |
| 15:56 | <MikeSmith> | nm |
| 15:56 | <hober> | MikeSmith: no, tantek is in css and trolling webapps over irc :) |
| 15:57 | <MikeSmith> | ah |
| 15:57 | <zewt> | why is it 2012 and it's still hard to find http libraries that even have the word "pipelining" in the documentation |
| 16:00 | <annevk> | hober: did CSS actually "resolve" to not use ASCII case-insensitive matching? |
| 16:00 | <annevk> | ooh |
| 16:00 | <annevk> | I just thought of something |
| 16:01 | <hober> | no, i think we are waiting on i18n wg feedback |
| 16:01 | <hober> | bert wants something more complicated than ascii case-insensitive matching |
| 16:01 | <annevk> | ffs |
| 16:01 | <annevk> | so what is CSS internally? DOMString? |
| 16:01 | <annevk> | because of the CSSOM? |
| 16:01 | <annevk> | because then using Unicode will be even weirder, what's gonna happen to lone surrogates? |
| 16:02 | <annevk> | will somebody please think of the lone surrogates |
| 16:02 | <hober> | "lone surrogates" was the name of my high school country-western band |
| 16:03 | annevk | updated http://wiki.whatwg.org/wiki/Band_names |
| 16:04 | <divya> | is that a w3c meme annevk |
| 16:04 | <divya> | it appears to be so |
| 16:05 | divya | coughs at w3cmeme creators |
| 16:05 | <annevk> | I learned about the ASCII thing because of the meme :) |
| 16:05 | <annevk> | divya: you want to submit memes? |
| 16:05 | divya | is on quickmemes |
| 16:06 | <hober> | what happened to the wiki? i can't log in anymore. :( |
| 16:07 | <divya> | you are no longer member of whatwg hober |
| 16:07 | <divya> | given you are now editor in w3 |
| 16:07 | <divya> | c |
| 16:07 | <hober> | nvm, just forgot my username |
| 16:07 | divya | is disappointed there is no conspiracy |
| 16:07 | <odinho_> | divya: I liked your version better |
| 16:08 | <hober> | divya should know by now that there is no vast, browser-wing conspiracy |
| 16:08 | <divya> | there is only a small WG conspiracy |
| 16:08 | divya | is joking in case people want to quote me out of context |
| 16:08 | divya | awaits twitter drama |
| 16:13 | <annevk> | http://w3cmemes.tumblr.com/post/34640099897/unicode-operations-on-a-domstring-are-not-a-good |
| 16:14 | <divya> | +++++ |
| 16:23 | <hober> | "user bang important rule" was the second album of which band? |
| 16:25 | <hsivonen> | what kind of <intent> element came up while I away from the webapps room? |
| 16:27 | <annevk> | dunno, I'm trying to fix URLs |
| 16:29 | <annevk> | zcorpan: willing to test ftp/gopher as well? |
| 16:29 | <annevk> | zcorpan: per spec only a few schemes can have a query component: http://url.spec.whatwg.org/#relative-scheme |
| 16:34 | <annevk> | zcorpan: in Gecko I get page-encoding for e.g. ws://test/?† |
| 16:34 | <annevk> | zcorpan: did you test inside <a>? |
| 16:34 | <annevk> | zcorpan: because for APIs the encoding flag might just be set to utf-8, as it is in e.g. XMLHttpRequest and CSS |
| 16:34 | <zcorpan> | annevk: my test case tests WebSocket(...) and what gets sent to the server |
| 16:35 | <annevk> | yeah that's different |
| 16:40 | <zcorpan> | annevk: http://simon.html5.org/test/url/url-encoding.html |
| 16:41 | <annevk> | add file:/// |
| 16:41 | <annevk> | and thanks! |
| 16:42 | <zewt> | thanks!:/// |
| 16:42 | <annevk> | bot detected |
| 16:44 | <hsivonen> | so presumably dinner is only after the pointer events meeting? |
| 16:45 | <zcorpan> | annevk: file: added |
| 16:45 | <zcorpan> | with two slashes :-) |
| 16:49 | <zcorpan> | what's the procedure for subscribing to whatwg bugs in w3c bugzilla? |
| 16:49 | <zcorpan> | i stalked Hixie for a while but realized that i received some bug emails twice, which is uncool |
| 16:50 | <MikeSmith> | what is subscribing? |
| 16:50 | <zcorpan> | well |
| 16:50 | <zcorpan> | i want to be notified about the bugs |
| 16:50 | <MikeSmith> | ah |
| 16:51 | <MikeSmith> | you got notification for the clone and bug both I guess you mean |
| 16:51 | <zcorpan> | no, i got two emails for the same bug # |
| 16:51 | <zcorpan> | iirc |
| 16:51 | <MikeSmith> | oh |
| 16:52 | <zcorpan> | how is one supposed to set up the prefs to get notified about bugs for specs one cares about? |
| 16:52 | <MikeSmith> | can't |
| 16:52 | <MikeSmith> | as far as I know |
| 16:52 | <zcorpan> | :-( |
| 16:53 | <MikeSmith> | with our bugzilla |
| 16:53 | <MikeSmith> | currently |
| 16:53 | <MikeSmith> | there is some feature for watching a component |
| 16:53 | <MikeSmith> | in some recent version of bugzilla |
| 16:53 | <MikeSmith> | but not in the version we're running now |
| 16:53 | <Hixie> | zcorpan: dude i get like 4 copies of each bugmail, it's insane |
| 16:53 | <zcorpan> | isn't there a workaround we can use if we refuse to update bugzilla? |
| 16:53 | <MikeSmith> | at least I've not been able to find an option to enable such a feature |
| 16:54 | <Hixie> | zcorpan: (one for me, one for watching the various lists, one for the htmlwg lists who are cc'ed, one for contributor⊙.) |
| 16:54 | <zcorpan> | i recall something about default QA contact being different for each component and then watching that |
| 16:54 | <zcorpan> | as a workaround |
| 16:54 | <zcorpan> | MikeSmith: can that be pulled off? |
| 16:55 | <Hixie> | i love the article where Paul Cotton was interviewed and he says the difference between the HTMLWG and the WHATWG is that in the WHATWG one person decides whereas in the HTMLWG the group decides |
| 16:55 | <MikeSmith> | zcorpan: yes |
| 16:55 | <MikeSmith> | we can do that |
| 16:55 | <zcorpan> | MikeSmith: excellent |
| 16:55 | <Hixie> | hello, the HTMLWG process is that three people, of which he's one, take turns to decide! |
| 16:56 | <zewt> | Hixie: is he criticizing the htmlwg? we all know how bad design by committee is, after all~ |
| 16:57 | <Hixie> | no he was criticising me, for doing what he does :-) |
| 16:58 | <zewt> | sounds more like the other way around, though perhaps he's not aware of what he's doing :P |
| 16:58 | <MikeSmith> | zcorpan: I'll add some right now |
| 16:58 | <zcorpan> | MikeSmith: thanks |
| 17:02 | <Hixie> | annevk: still no reply from those ietf folk, i wonder if my last e-mail just made them give up |
| 17:03 | <annevk> | Hixie: yeah dunno, I'm making modest process in between talking with people here |
| 17:03 | <GPHemsley> | zcorpan, MikeSmith: Indeed, the old BMO way was to have fake QA Contacts that people could watch. I think it's a good workaround, if you don't actually have real QA Contacts. |
| 17:03 | <annevk> | Hixie: with the spec that is |
| 17:04 | <Hixie> | annevk: cool |
| 17:05 | GPHemsley | wonders why we don't just set up our own Bugzilla instance. |
| 17:05 | GPHemsley | then remembers that suggesting such things counts as volunteering and thus retracts his statement. |
| 17:06 | <zcorpan> | i think i heard GPHemsley volunteering to set up our own Bugzilla instance. |
| 17:06 | <GPHemsley> | zcorpan: You are imagining things. |
| 17:06 | <zcorpan> | oh. :-( |
| 17:06 | <GPHemsley> | zcorpan: It's the voice in your head. |
| 17:07 | <GPHemsley> | Which could mean that it's actually *you* volunteering! ;) |
| 17:07 | <zcorpan> | otoh, having our own instance would make it harder to clone and move bugs, no? |
| 17:07 | <zcorpan> | to other groups, i mean |
| 17:08 | <annevk> | yeah, sharing Bugzilla is very useful |
| 17:08 | <zcorpan> | just need to kick some w3c people to update it |
| 17:09 | <divya> | can bugzilla be skinned? |
| 17:10 | <zcorpan> | sure. just use a user stylesheet. |
| 17:10 | <zcorpan> | or glue a piece of skin to your screen. |
| 17:11 | divya | lols because csswg discussion is even more depressing |
| 17:11 | <divya> | zcorpan: no not like that, i mean is there a way to change the annoying dropdowns to autocomplete input boxes? |
| 17:12 | <zcorpan> | which field autocompletes? |
| 17:12 | <divya> | in w3c bugzilla you have to choose the specs from some 20 or so dropdown list |
| 17:13 | <divya> | or the scroll thing |
| 17:13 | <divya> | e.g. https://www.w3.org/Bugs/Public/show_bug.cgi?id=19735 |
| 17:13 | <divya> | Product |
| 17:13 | <divya> | Component |
| 17:13 | <divya> | Platform |
| 17:14 | <zcorpan> | oh. what's wrong with it? |
| 17:14 | <divya> | :|||||||||||||||||||||||||||||| |
| 17:14 | <divya> | never mind then. |
| 17:15 | <zcorpan> | i didn't mean to put you off, i'm more trying to understand what you think is bad |
| 17:16 | <zcorpan> | is the problem that there are 20 or so items in the list? or is the list itself hard to use? something else? |
| 17:16 | <divya> | yes zcorpan its painful |
| 17:16 | <MikeSmith> | zcorpan, Hixie and anybody else who's keen on bugmail: If I change the existing QA-contact address on the WHATWG HTML bugs, that's 1400+ bugs |
| 17:17 | <divya> | anyways BREAK TIME BAI |
| 17:17 | <MikeSmith> | and I think you might be a bugmail notifcation for each one of those |
| 17:17 | <zcorpan> | MikeSmith: so long as they're easily searchable so i can filter them out, i'm fine with that |
| 17:18 | <MikeSmith> | I think most sane people have their bugmail prefs set so that they don't get nofications about changes to the QA contact |
| 17:18 | <zcorpan> | i just unchecked the qa contact checkbox so i guess i won't get those emails now |
| 17:18 | <MikeSmith> | ok |
| 17:18 | <annevk> | let me change my settings before you hit submit |
| 17:18 | <MikeSmith> | hai |
| 17:19 | <MikeSmith> | anybody else please change their bugmail settings now please |
| 17:19 | <MikeSmith> | and anybody who doesn't and then gets the mailbomb, please don't kill me |
| 17:19 | <zcorpan> | https://www.w3.org/Bugs/Public/describecomponents.cgi?product=WHATWG seems to have default qa already |
| 17:20 | <Hixie> | MikeSmith: why would we change the qa contact? |
| 17:20 | <Hixie> | isn't it already contributor@? |
| 17:20 | <MikeSmith> | Hixie: because Simon wants a specific address to watch |
| 17:21 | <MikeSmith> | Hixie: is that the only one where we using contributor@ as QA contact? |
| 17:21 | <Hixie> | i hope so |
| 17:21 | MikeSmith | vaguely recalls he talked with Hixie about this before |
| 17:21 | <MikeSmith> | OK |
| 17:21 | <MikeSmith> | I won't change that one then |
| 17:21 | <Hixie> | k |
| 17:21 | <MikeSmith> | zcorpan: follow contributor@ for WHATWG HTML spec bugs |
| 17:22 | <zcorpan> | ok. thanks! |
| 17:22 | <MikeSmith> | for the rest of WHATWG components I changed em already |
| 17:25 | <MikeSmith> | divya, we wrapping up here in webapps |
| 17:25 | <MikeSmith> | you prepared for eating food now? |
| 17:27 | <zcorpan> | mmmm... fooood... |
| 17:33 | <GPHemsley> | MikeSmith: Why'd you make them your e-mail address? |
| 17:34 | <MikeSmith> | GPHemsley: because I don't own no other seven e-mail addresses to use isntead |
| 17:35 | <Hixie> | i can set up arbitrary addresses @whatwg.org if you need 'em |
| 17:35 | <Hixie> | in fact, they don't even have to receive mail |
| 17:35 | <Hixie> | so you can just use arbitrary addresses without me doing anything |
| 17:51 | <Hixie> | i got SEVEN copies of the bugmail for bug 19773 when it was created |
| 17:51 | <Hixie> | wtf |
| 17:53 | <WeirdAl> | stupid chatzilla - it thought you meant 19773 @ Mozilla's bugzilla |
| 17:53 | <Ms2ger> | netwerk/base/public/nsIIOService.idl interface status? |
| 17:53 | <Ms2ger> | Could've been! |
| 17:53 | <WeirdAl> | :D |
| 17:59 | <GPHemsley> | MikeSmith, Hixie: Yeah, I think arbitrary whatwg.org e-mail addresses would be better. I was thinking something like mimesniff⊙swo |
| 18:55 | <GPHemsley> | Hixie: I would like to change the wiki e-mail address(es) so that the messages seem to come from the wiki instead of you, but I seem to recall you objecting to that in the past. |
| 18:56 | <Hixie> | it should come from a person, since a person is responsible for dealing with problems, imho |
| 18:57 | <Hixie> | but as admin of the wiki, you can do with it as you wish |
| 18:57 | <Hixie> | just keep us posted :-) |
| 19:21 | <GPHemsley> | Hixie: Out of curiosity, does anybody have @whatwg.org e-mail addresses (besides contributor@)? |
| 19:36 | <Hixie> | GPHemsley: no |
| 19:46 | <smaug____> | Ms2ger: come back :) |
| 19:47 | <Ms2ger> | :) |
| 19:48 | Ms2ger | should be used to bz knowing everything by now |
| 19:50 | <smaug____> | I somehow think he didn't get this one right |
| 20:04 | <GPHemsley> | Registered users 1,428 |
| 20:04 | <GPHemsley> | Active users (list of members) |
| 20:04 | <GPHemsley> | (Users who have performed an action in the last 30 days) 14 |
| 20:04 | <GPHemsley> | ;_; |
| 20:04 | <karlcow> | GPHemsley: the stats for? |
| 20:05 | <GPHemsley> | the wiki |
| 20:05 | <karlcow> | which one? |
| 20:05 | <GPHemsley> | WHATWG |
| 20:05 | <karlcow> | ah expected. |
| 20:06 | <GPHemsley> | Hixie: Can you create an e-mail alias for me? |
| 20:06 | <GPHemsley> | Hixie: (for wiki.whatwg.org) |
| 20:17 | <Hixie> | GPHemsley: yeah, e-mail me details, running to lunch now |
| 20:17 | <GPHemsley> | alright |
| 20:32 | <GPHemsley> | Hixie: Mail sent. |
| 20:33 | <GPHemsley> | Does somebody want to try editing my talk page to test the e-mail system? |
| 21:46 | <Hixie> | GPHemsley: admin@wiki now goes to you and me |
| 21:46 | <GPHemsley> | k, thanks |
| 21:47 | <Hixie> | we could add anne too if he wants |
| 21:47 | <Hixie> | then if we need to have people ask us for accounts we can have them mail that dadress instead |
| 21:48 | <GPHemsley> | well, I've set it up so that all trusted users can create accounts |
| 21:50 | <Hixie> | k |
| 23:10 | <zcorpan> | http://filamentgroup.com/lab/rwd_img_compression/ is interesting |
| 23:11 | <zcorpan> | if that is good enough for retina, yet has smaller file size than what people do today for non-retina, it seems like a solved problem with just <img src>. |
| 23:13 | <zcorpan> | well, <img src width height>. we might want a way to say "2x" without knowing the exact dimensions |
| 23:13 | <zewt> | fg |
| 23:14 | <WeirdAl> | Hey folks - looking at svn.whatwg.org/webapps, and I'm wondering how the multipage version is generated... I'm wondering if you have a script for that which could be checked in. :) |
| 23:14 | <zewt> | yeah, that. |
| 23:14 | <zcorpan> | (and design a solution just for the remaining problems with the art direction use case) |
| 23:16 | <zewt> | dear whoever this person is: files are not measured in pounds |
| 23:17 | <zcorpan> | i wonder how webp compares with this technique |
| 23:17 | <zewt> | zcorpan: it's roundabout, but it sounds like a wordy way of saying "high pixel density images can tolerate more artifacts" and "just quadruple your image resolution and target the same file size" |
| 23:18 | <zewt> | it's roundabout, so i'm not sure if i'm getting from it what was intended (and I'm a bit distracted) |
| 23:19 | <zcorpan> | zewt: right. though the bigger image has worse artifacts even when scaled down for the sky, as the first comment points out. which is why i wonder how webp compares. |
| 23:19 | <zewt> | a proper test would be three images: low-dpi high-bitrate ("regular"), high-dpi high-bitrate ("regular retina") and high-dpi low-bitrate-matching-#1s-file-size |
| 23:20 | <zcorpan> | and various different images |
| 23:20 | <zewt> | (i have an ipad 3 in front of me, but without a decent test I don't feel like painstakingly typing that URL into it; chrome to ipad wanted) |
| 23:22 | <zcorpan> | maybe i can hack together a test. i'll see if i have interesting images. also, what is a good webp encoder and what is a good jpeg encoder these days? |
| 23:22 | <zewt> | (don't know; i stopped paying attention to webp since it didn't look like it was a serious attempt at a general-purpose jpeg replacement) |
| 23:23 | <zewt> | seemed (a couple years ago?) like it was hyperfocused on "a lossy image format for web pages" and not caring about replacing jpeg in general, and I don't think I want a world where we have two production lossy image formats and people need to make complex judgements to decide which to use |
| 23:24 | <zcorpan> | i thought webp was being improved. though i haven't followed it closely. |
| 23:26 | <zewt> | it's not that it isn't being improved (no idea on that, really), but that it seemed they were happy to not support things that aren't directly relevant to web images |
| 23:26 | <zcorpan> | i wish more time was spent on improving jpeg encoders |
| 23:26 | <zewt> | for example, jpeg's max file size I believe is 64x64k, webp 16x16k |
| 23:26 | <zewt> | users shouldn't need criteria like "use webp! but if your file is very large, use jpeg instead!" |
| 23:27 | <zewt> | (that's just the random thing I happened to recall off the top of my head, I think there were others but it's been too long) |
| 23:27 | <zcorpan> | right |
| 23:28 | <zewt> | cmyk might have been another |
| 23:28 | <zewt> | (i guess those are both print things, where you probably don't want jpeg either) |
| 23:29 | <zcorpan> | though 16x16k seems enough even for big images that have twice the resolution you want |
| 23:29 | <zcorpan> | for the web, i mean |
| 23:29 | <zewt> | for the web, sure (more than that will tend to choke browsers anyway) |
| 23:51 | <Hixie> | gah |
| 23:51 | <Hixie> | anyone remember where i put those tests recently that checked <embed> behaviour when display:none / height:0 / etc? |
| 23:59 | <WeirdAl> | Hixie: no idea, but on another note: how do you generate the multipage editions of the HTML5 spec from the source file? |
| 23:59 | <WeirdAl> | (see my question posted 45 minutes ago) |
