| 00:06 | <Hixie> | only 206 e-mails left in the WF2 folder |
| 00:38 | <Hixie> | 168. |
| 00:47 | <blooberry> | hixie: 38 in a half hour? You'll be done by bedtime at that rate |
| 00:47 | <Hixie> | not all of them got direct replies, some got moved to other folders for bulk reply later |
| 00:48 | <Hixie> | but bulk replies are easy to do compared to individual replies and compared to finding the e-mails to bulk reply to |
| 00:54 | <Philip`> | "... if the returned value is an element node with either the local name datalist or the local name select, and (for XHTML) with the XHTML namespace" - why does the namespace bit only apply to XHTML? |
| 00:55 | <Philip`> | Oh, wait, that's the old WF2 spec |
| 00:55 | <Hixie> | wf2 is a mess |
| 00:55 | <Philip`> | Please ignore me :-) |
| 00:56 | <Hixie> | 152 |
| 00:57 | <Hixie> | i'll probably be a bit late to the whatwg meet |
| 00:57 | <Hixie> | i hope people don't leave early |
| 00:57 | <Hixie> | (i have to eat first then cycle home then go there) |
| 00:57 | <blooberry> | I think in this group, everyone will understand your excuse of "I was working on the spec" |
| 00:58 | <dbaron> | I couldn't quite tell if people were expecting to eat at the meetup or before it or after it... |
| 01:00 | <othermaciej> | oh yeah the meetup |
| 01:02 | <annevk2> | we can meat at the meetup I think |
| 01:02 | <annevk2> | I won't have eaten when I arrive |
| 01:03 | <Philip`> | Meat? |
| 01:03 | <annevk2> | eat |
| 01:03 | <Philip`> | Ah |
| 01:08 | <Hixie> | ok going to grab food. |
| 01:09 | <Hixie> | blooberry: you coming? |
| 01:09 | <dbaron> | I'm not sure about eating at a pub... |
| 01:09 | <Hixie> | actually that pub has ok food iirc |
| 01:09 | <Hixie> | but i wouldn't eat there either :-) |
| 01:09 | <dbaron> | of course, it might take me a good 20 minutes to find the people who I promised I'd give a ride... |
| 01:10 | <Hixie> | bbl |
| 01:11 | <annevk2> | i'm ok with eating somewhere else too |
| 01:11 | <annevk2> | if someone can name me a restaurant i'll show up |
| 01:19 | <annevk2> | looking at the time I'll probably eat at the pub |
| 11:21 | <Lachy> | I think that <q> discussion has effectively turned into a permathread. It's basically rehashing the same arguments that I've seen on www-html, www-style and possibly other lists for at least the past 5 years |
| 11:29 | <hendry> | maybe someone needs to blog on the pros and cons to act as a reference |
| 11:33 | <Lachy> | hendry, thanks for volunteering :-) |
| 11:59 | <zcorpan> | i don't read that thread; have i missed anything? |
| 12:13 | <Lachy> | zcorpan, here's a quick summary of the arguments so far: Different langagues have different quoting conventions, nested quotes and mixed langauges make things more complex. Various solutions proposed ranging from defining explicit quoting rules, to no quoting, to dropping <q> altogether |
| 12:13 | <Lachy> | now you haven't missed anything |
| 12:16 | <zcorpan> | Lachy: thanks |
| 12:17 | <zcorpan> | i'd prefer one of the latter proposals |
| 12:20 | <Lachy> | personally, I'd prefer to drop it at this stage. But in any case, the rules for rendering quotes need to be defined |
| 12:21 | <Lachy> | in which case, I'd like the spec to say render no quotes. But since IE8 will be rendering quotes now, we might be stuck with defining something else |
| 12:21 | <Lachy> | like always using the curly double and single quotes by default |
| 12:21 | <zcorpan> | Lachy: do you see adding it back in HTML6 if all browsers don't render quotes by then? |
| 12:22 | <Lachy> | if there are use cases to justify it, then it could stay in HTML5. But there haven't been any compelling use cases presented yet |
| 12:23 | <Lachy> | but I think if it renders quotes by default, then it will make it a less valuable solution, even if there are use cases for it |
| 15:14 | <Lachy> | Hixie, yt? |
| 15:14 | <zcorpan> | Hixie: shouldn't a UA wait with load()ing until the </video> or </audio> end tag has been parsed? |
| 15:16 | <zcorpan> | Hixie: consider the UA getting "<video><source src=1>" and then the server waits a few seconds and then further gives "<source src=2>", the UA would already had started to load the first resource |
| 15:16 | annevk2 | doubts Hixie is awake; it's 8AM over here :) |
| 15:17 | smedero | rubs eyes |
| 15:17 | <smedero> | http://adactio.com/journal/1524/ |
| 15:17 | <zcorpan> | morning annevk2 |
| 15:20 | <zcorpan> | smedero: heh, without thinking about where the link came from i was about to paste it here after reading it |
| 15:20 | <smedero> | :-) |
| 15:20 | <zcorpan> | i guess browsers should "throttle the download" when there are many <audio>s |
| 15:22 | <smedero> | Is there an experimental build of Opera that support <audio>? I know there is a <video> one... |
| 15:22 | <zcorpan> | no, though we have an old impl of Audio() |
| 15:23 | <smedero> | ahh, ok. |
| 15:25 | <blooberry> | annevk2: what are you doing in silly valley right now again? |
| 15:29 | <annevk2> | http://www.opera.com/education/tours/ |
| 15:36 | <blooberry> | annevk2: ah, that. 8-} |
| 15:41 | <BenMillard> | here's something I just said in Mozilla's #accessibilty channel: "An 80% solution is many times more likely to fulfill the requirement of '1 major website accessible for each activity people do online' than a 1% solution, or a 4.5% solution" |
| 15:48 | <zcorpan> | firefox seems to require type on <source> but it shouldn't be required per spec afaict |
| 15:55 | <gsnedders> | Man. Too many emails. |
| 16:00 | <Philip`> | <q> is the new <acronym> |
| 16:16 | <aaronlev> | Hixie, BenMillard was just pinging me about tables |
| 16:16 | <doublec> | zcorpan, source isn't fully implemented - there's a bug to that effect |
| 16:17 | <aaronlev> | what do you guys want me to review |
| 16:17 | <doublec> | what's there was done mainly so another browser vendor could do a demo showing multiple video formats working across browsers |
| 16:20 | <zcorpan> | doublec: ok |
| 16:21 | <BenMillard> | aaronlev, here's the header association algorithm as currently specified: http://www.whatwg.org/specs/web-apps/current-work/multipage/tabular-data.html#header-and-data-cell-semantics |
| 16:21 | <aaronlev> | and was there some other algorithm it was going to merge with? |
| 16:21 | <doublec> | zcorpan, bug 449363 |
| 16:21 | <annevk2> | BenMillard, aaronlev and Hixie already communicated about that during TPAC |
| 16:22 | <annevk2> | BenMillard, Hixie was going to get back to aaronlev once he addressed outstanding feedback |
| 16:22 | <aaronlev> | annevk2: i wasn't at TPAC |
| 16:22 | <annevk2> | aaronlev, you still communicated though :) (over IRC) |
| 16:22 | <aaronlev> | i mean, hxie asked me to review something this week |
| 16:22 | <aaronlev> | so is it ready for my review or is there still outstanding feedback to fold in first |
| 16:23 | <annevk2> | the latter |
| 16:23 | <annevk2> | also, http://krijnhoetmer.nl/irc-logs/whatwg/20081023#l-263 suggests it would not necessarily be this week |
| 16:24 | <BenMillard> | aaronlev, there's a "semantics-tables" folder which lists some specific message Hixie will be going through: http://www.whatwg.org/issues/ |
| 16:25 | <BenMillard> | aaronlev, the "Smart Span" algorithm is a prototype James Graham documented here: http://lists.w3.org/Archives/Public/public-html/2008Mar/0075.html |
| 16:25 | <aaronlev> | i'll review it when it's got all of the current feedback folded in |
| 16:27 | <gsnedders> | Has Mr Last Week died? |
| 16:27 | <BenMillard> | aaronlev, ok...just thought you might want to get a headstart, like trying out the various ideas prototyped here: http://james.html5.org/tables/table_inspector.html |
| 16:27 | <aaronlev> | BenMillard: if possible i would like others more familiar with the issues to arrive at a consensus |
| 16:27 | <aaronlev> | i'll check the algorithm to see if there are potential perf issues or whatever |
| 16:28 | <aaronlev> | and we can code it so people can play with the results |
| 16:40 | <aaronlev> | well my first feedback is that you have a list with 4 levels and all levels use arabic numbers |
| 16:40 | <aaronlev> | in the whatwg spec |
| 16:40 | <aaronlev> | kind of easy to lose your place |
| 16:40 | <Philip`> | Argh, I can't even draw a table with borders without getting totally different results in Opera and Firefox |
| 16:40 | Philip` | wonders why HTML/CSS is so hard |
| 16:42 | <Philip`> | http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0D%0A%3Cstyle%3E%0D%0Atable%20%7B%20border-collapse%3A%20collapse%3B%20%7D%0D%0Atable%20td.left%20%7B%20border%3A%201px%20solid%3B%20%7D%0D%0Atable%20td.right%20%7B%20border%3A%201px%20solid%3B%20%7D%0D%0A%3C%2Fstyle%3E%0D%0A%3Ctable%20class%3D%22matrix%22%3E%0D%0A%3Ctr%3E%0D%0A%3Ctd%20class%3D%22left%22%20rowspan%3D%222%22%3Ea%0D%0A%3Ctd%3Ex%0D%0A%3Ctd%20class%3D%22right%22%2 |
| 16:43 | <Philip`> | Oops |
| 16:43 | <Philip`> | http://tinyurl.com/57bksm |
| 16:43 | <Philip`> | Is Firefox (3.0) wrong in drawing a border underneath the 'y'? |
| 16:46 | <Philip`> | All I want to do is use table hacks to draw a matrix :-( |
| 17:10 | <svl> | Philip`: so it seems to set the border-bottom of td.left to be equal to the border-bottom of the second tr - and then apply the border styles of both td and tr to it. I don't know if the behaviour in the case of rowspan is actually defined anywhere... (And in the case of "undefined", I don't think gecko can be said to be wrong, though it's certainly counter-intuitive.) |
| 17:12 | <svl> | Looks to be https://bugzilla.mozilla.org/show_bug.cgi?id=332740 |
| 17:12 | <svl> | yeah, the testcase matches |
| 17:47 | <Philip`> | svl: Okay, thanks |
| 17:47 | <Philip`> | I think I might have actually run into the same problem in the past in a totally different context, since the bug looks familiar |
| 17:49 | <Philip`> | But I'm perfectly happy to blame Gecko for violating my expectations, regardless of what specs say, particularly if other browsers do what I want :-) |
| 17:51 | <svl> | Heh. :) |
| 18:06 | <Philip`> | (In this case I don't actually need border-collapse at all, so I can just remove it and it's all fine) |
| 18:16 | <Hixie> | zcorpan: no, it should look with every source, that way it starts getting one as soon as there is a good one |
| 18:20 | <hober> | http://www.jroller.com/tedgoddard/entry/websocket_is_neither_web_nor |
| 18:27 | <Lachy> | Hixie, I have some issues with your disposition-of-comments2html.py script, which I borrowed from XBL2. It doesn't seem to work the way I expected it would. |
| 18:27 | <Lachy> | see http://dev.w3.org/2006/webapi/selectors-api/disposition-of-comments.txt and http://dev.w3.org/2006/webapi/selectors-api/disposition-of-comments.html |
| 18:27 | <Lachy> | it seems it totally ignored the summary field and used the subject line from the email as the heading instead. |
| 18:27 | <john_fallows> | hober: there seem to be some technical inaccuracies in that post |
| 18:29 | <john_fallows> | hober: but it does raise two interesting questions about IANA ports and WebSocket connection limits |
| 18:30 | <john_fallows> | Hixie: when would it be appropriate for IANA to be consulted regarding ws:// port 81 and wss:// port 815 ? |
| 18:32 | <Lachy> | john_fallows, when the web sockets section is relatively stable |
| 18:34 | <john_fallows> | Lachy: are the default port numbers for ws:// and wss:// schemes not yet considered stable? |
| 18:37 | <Lachy> | I don't know |
| 18:38 | <Lachy> | are there browser vendors looking at implementing really soon? |
| 18:39 | <john_fallows> | you'd have to ask them :-) |
| 18:40 | <john_fallows> | presumably this IANA request would be one of the last steps to stabilize the WebSocket section before browser vendors would provide an implementation |
| 18:41 | <Lachy> | it would need to be done before any release version was shipped with support for web sockets. But browsers could proceed with experimental versions before it was finalised and provide feedback |
| 18:41 | <Hixie> | Lachy: yes, that is as designed (by summaries sucked) |
| 18:42 | <Hixie> | s/by/my/ |
| 18:42 | <Lachy> | oh. Is there another field that I can use that will be output by the script? |
| 18:42 | <Hixie> | no idea |
| 18:42 | <Lachy> | or will I have to modify the script to output the summaries myself? |
| 18:42 | <Lachy> | ok |
| 18:42 | <john_fallows> | what is the proposed relationship between WebSocket connection limits and regular HTTP connection limits in the browser? |
| 18:42 | <Hixie> | i didn't even remember the script existed until 2 minutes ago |
| 18:43 | <Hixie> | my impression was that there was no IANA consultation process, just a registration process, and I do not want to register anything until we know we're sure we're doing it |
| 18:43 | <Hixie> | same with the headers, mime types, etc |
| 18:43 | <Hixie> | about the only thing html5 doesn't need registering is a new top-level mime hierarchy |
| 18:43 | <john_fallows> | Hixie: sounds like exactly the right strategy |
| 18:44 | <Hixie> | (we're even talking about registering new registries!) |
| 18:44 | <john_fallows> | lol |
| 18:44 | <john_fallows> | what is the proposed relationship between WebSocket connection limits and regular HTTP connection limits in the browser? |
| 18:44 | <jcranmer> | Hixie: I'm sure someone will come up with a top-level MIME hierarchy for HTML 5 soon |
| 18:45 | <gsnedders> | Hixie: We were discussing unknown as top-level, remember? |
| 18:45 | <Hixie> | john_fallows: i don't think anything is proposed |
| 18:45 | <Hixie> | gsnedders: oh jeez, you're right |
| 18:45 | <Hixie> | we're EVEN considering registering a top-level mime hierarchy |
| 18:45 | <Hixie> | what a jip |
| 18:45 | <Hixie> | this is ridiculous |
| 18:46 | <Hixie> | http headers, ports, schemes, mime types, mime hierarchies, registries |
| 18:46 | <Hixie> | anything else? |
| 18:47 | <jcranmer> | want more stuff to consider adding for the sake of registering? |
| 18:47 | <Hixie> | lord no |
| 18:47 | <Hixie> | just making sure i know the whole list |
| 18:48 | <john_fallows> | given that a WebSocket connection cannot be reused, it seems appropriate to specify connection limits independent from HTTP, to avoid WebSocket taking up all the HTTP connections |
| 18:48 | <jcranmer> | well, this is the lot: http://iana.org/protocols/ |
| 18:49 | <Hixie> | john_fallows: yes i would think it is reasonable not to apply http limits to websocket |
| 18:49 | <Hixie> | that's probably the default, too, since http != websocket :-) |
| 18:50 | <john_fallows> | yes, agreed, although it probably ought to be specified explicitly to make sure browser vendors keep them separate |
| 18:50 | <Hixie> | yeah maybe |
| 18:51 | <john_fallows> | Server-sent events would also benefit from being kept separate from regular HTTP connections |
| 18:53 | <john_fallows> | in fact, it is probably reasonable to place a combined limit on SSE and WebSocket connections, separate from the limit on regular HTTP connections |
| 18:53 | <jcranmer> | I'm guessing webservice is TCP, right? |
| 18:53 | <Hixie> | wow, my filters totally caught the entire <q> thread and considered it low-priority |
| 18:53 | <jcranmer> | s/service/socket/ |
| 18:53 | <Hixie> | my filters rock |
| 18:54 | <Hixie> | jcranmer: yes |
| 18:54 | <jcranmer> | so you don't need to register a protocol number |
| 18:54 | <BenMillard> | cool, aaronlev sent some early feedback while I was eating a kebab: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-October/016915.html |
| 18:55 | <jcranmer> | that appears to be the whole list then |
| 18:55 | <aaronlev> | BenMillard: but my feedback isn't nearly as tasty |
| 18:55 | <BenMillard> | :P |
| 18:55 | <BenMillard> | Hixie, since IE8 will be shipping soon with an implementation of <q> which matches HTML4 but differs from HTML5, maybe it should get bumped up the queue? |
| 18:55 | <BenMillard> | (for some definition of "soon") |
| 18:56 | <Hixie> | i don't see what we can do but drop the entire element at this point |
| 18:56 | <Philip`> | I suggest renaming the element from <q> to " |
| 18:56 | <Hixie> | right |
| 18:57 | <BenMillard> | in my research, authors commonly use punctuation instead of <q> even when they do " |
| 18:57 | <BenMillard> | but they also use <em> with punctuation to avoid using <q> with CSS |
| 18:59 | <BenMillard> | so authors are happy to type the punctuation, but sometimes want styling as well...so if <q> was consistent with other phrase-level elements and didn't generate punctuation, that would be neat way of achieving this |
| 19:03 | <campd> | Hixie: so another offline nit... should a resource that matched a fallback namespace, but didn't actually fallback, be considered "loaded from the application cache" for the purposes of the cache selection algorithm? |
| 19:03 | <campd> | it seems like it should, but it's not explicit. |
| 19:05 | <john_fallows> | Hixie: since there is no explicit limit set on WebSocket connections in the spec, does that imply there is no limit, so any number of WebSocket connections could be made from the same browser to the same WebSocket server? |
| 19:07 | <Hixie> | BenMillard: well apparently <q> generates punctuation |
| 19:08 | <aaronlev> | BenMillard: i did not understand kristof's reply to me :/ |
| 19:08 | <Hixie> | campd: top-level? |
| 19:09 | <Hixie> | john_fallows: yes |
| 19:09 | <Hixie> | aaronlev: nobody understands kristof really |
| 19:09 | <aaronlev> | snort |
| 19:09 | <john_fallows> | Hixie: ok, do you have any objection to making that implication explicit in the spec? |
| 19:11 | <Hixie> | john_fallows: not particularily, send feedback. :[) |
| 19:11 | <Hixie> | er |
| 19:11 | <Hixie> | :-) |
| 19:11 | <john_fallows> | ok, will do. |
| 19:15 | <Hixie> | thanks |
| 19:16 | <john_fallows> | done |
| 19:16 | <takkaria> | Hixie: I'd like to see your filters, they must have some serious intention-detecting voodoo going on |
| 19:17 | <BenMillard> | I feel sorry for Chris Wilson...his opening message is an implementor giving feedback during their final approach to shipping a product and there's very little he can take away from the thread which resulted |
| 19:19 | <BenMillard> | once all 4 main browsers are generating " on <q> and </q>, I guess we're stuck with it...even though hardly anyone uses <q> and it is impossible to internationise if it generates punctutation (as David Baron explained during the HTMLWG meeting) |
| 19:19 | <BenMillard> | s/punctutation/punctuation/ |
| 19:20 | <BenMillard> | s/internationise/internationalise/ |
| 19:21 | <Philip`> | takkaria: They shouldn't need much voodoo - they can be based simply on the existing distributed socially-networked human-powered filtering system, i.e. if certain people choose to respond to a topic then you probably don't want to read it |
| 19:23 | <Philip`> | Those people do all the hard work of reading the emails and classifying them into "I must give my opinion here!" and "this is just technical details, I don't care about this", so it's classified along the same boundaries that you want |
| 19:25 | <takkaria> | I imagine that filtering any threads which generate so much noise from so few people would work quite well too |
| 19:51 | <campd> | Hixie: yeah |
| 19:53 | <campd> | Hixie: (or any context, if we end up making each iframe an island) |
| 20:10 | <Hixie> | campd: i mean, as opposed to an <img> image |
| 20:10 | <Hixie> | campd: as far as i can tell, it's all already defined |
| 20:11 | <Hixie> | campd: if it's top-level, matched a namespace but came from the network, it's only in the cache if it has a manifest |
| 20:12 | <campd> | ok |
| 21:11 | <gsnedders> | should any event fire on auto-completion? do we want one to? |
| 21:36 | <BenMillard> | in the end, I couldn't resist the <q> thread: http://lists.w3.org/Archives/Public/public-html/2008Oct/0221.html |
| 21:37 | <BenMillard> | I looked on the ESW wiki for things about <q> but there were only position papers, with few references to sources other than those the main contributor had written |
| 21:47 | <Lachy> | BenMillard, at least your post was well thought out and contained a lot of useul information, rather than being totally useless like most others in the thread |
| 21:57 | <BenMillard> | Lachy, thanks. :) |
| 21:57 | <hober> | indeed, it's been almost the only email worth reading in the thread |
| 22:05 | <Dashiva> | So what's the next <acronym> after we're done with <q>? |
| 22:06 | <smedero> | we'll probably just rehash <i> and <b>. |
| 22:06 | <Hixie> | man, hsivonen_ would love what's going on in geolocation |
| 22:07 | <Hixie> | people are trying to get an API changed to push their political agenda |
| 22:07 | <smedero> | :( |
| 22:07 | <Dashiva> | Hixie: Isn't that just business as usual in w3c? |
| 22:07 | <Hixie> | no comment |
| 22:07 | <Hixie> | hey is mr last week ok? it's been nearly 5 days since his last post |
| 22:08 | annevk2 | wants a copy of that picture of Hixie and John Foliot |
| 22:08 | <Dashiva> | Did you actually say that Ali G line? |
| 22:08 | <Hixie> | hm? |
| 22:08 | <smedero> | maybe he went out to the russian cabaret |
| 22:09 | <Dashiva> | This one: http://lastweekinhtml5.blogspot.com/2008/10/tpacker.html |
| 22:09 | <annevk2> | smedero, you should've been there |
| 22:09 | <smedero> | yeah, I felt a bit of regret after reading the twitter recap. |
| 22:10 | <smedero> | though I had fun discussion of US immigration and driving laws |
| 22:10 | <BenMillard> | smedero, I was there for that! It wasn't my idea of fun! |
| 22:10 | <smedero> | oh |
| 22:15 | <BenMillard> | the ice cream was a dream, though |
| 22:24 | <Hixie> | Dashiva: sure, I often say "yo wassup" when someone asks me if I'm there. :-) |
| 22:25 | <Hixie> | i didn't say that riviera quote though |
| 23:05 | <met_> | Someone is experimenting with audio element in firefox http://adactio.com/journal/1524 |
| 23:08 | <doublec> | "Support for the audio element is on its way in Firefox 3.1, albeit in a crippled Ogg-only way." |
| 23:08 | <doublec> | nice |
| 23:11 | <olliej> | met_: it would be nice if they filed a bug on the <audio> preloading behaviour so we were aware of a use case in which it's a problem |
| 23:13 | <met_> | olliej, ask them, there is email address in page footer |
| 23:13 | <olliej> | i've already filed a bug |
| 23:14 | <olliej> | it just stuns me that people still insist on complaining about bugs in blogs, rather than just, you know, reporting the bug |
| 23:15 | met_ | agrees |
| 23:44 | <Hixie> | olliej: yeah no kidding |
| 23:45 | <Hixie> | olliej: bugs me when people complain about spec bugs in bug databases instead of telling me on the mailing lists too though :-) |
| 23:45 | <gsnedders> | olliej: Bug trackers suck though |