| 00:05 | <Dashiva> | "You can look at how standards evolve, like how quickly CSS2 became CSS 2.1" |
| 00:06 | <Dashiva> | And how CSS 2.1 was stuck in CR for how many years... |
| 00:26 | <Hixie> | "CSS2 became CSS 2.1"? |
| 00:29 | Philip` | wonders why "different aspects of HTML5" includes a link to an SVG-related post |
| 00:29 | <Philip`> | (in the ieblog post) |
| 00:48 | <jwalden> | Philip`: inline SVG in HTML maybe? |
| 00:50 | <othermaciej> | CSS2 went REC in 1998 and was rescinded in favor of 2.1 in 2008 |
| 00:54 | <karlcow> | Dashiva: about "And how CSS 2.1 was stuck in CR for how many years...", I would say, some people and companies were real world perfectionist |
| 00:57 | <Hixie> | the ".1" in CSS2.1 stands for "revision 1" |
| 00:57 | <Hixie> | CSS2.1 _is_ CSS2 |
| 02:39 | <JonathanNeal> | Hey all! |
| 02:58 | <MikeSmithX> | JonathanNeal: こんちは |
| 02:59 | <JonathanNeal> | Wow. |
| 03:03 | <JonathanNeal> | MikeSmithX, uncultivated land? |
| 03:11 | <JonathanNeal> | My life is so much happier since trying to translate that. I've been reading Japanese blogs poorly translated into English. |
| 03:27 | <boblet> | JonathanNeal: I’m betting a shortened form of こんにちは ;-) |
| 03:27 | <boblet> | your version sounds way more poetic though |
| 03:28 | <JonathanNeal> | "I will climb the palm tree; I will take hold of its fruit." |
| 03:28 | <boblet> | that’s what she said |
| 03:29 | <JonathanNeal> | Yea, you go and google that and see where it leads you, boblet. |
| 03:29 | <boblet> | woops, this is probably not the forum for obscure Halo:ODST refs huh |
| 04:33 | Hixie | is liking these conforming-with-warning feature less and less |
| 04:33 | <Hixie> | maybe i should raise an issue suggesting that we make them non-conforming but downplayed errors again |
| 05:24 | <MikeSmith> | Hixie: what would be your suggestion for how a "downplayed error" gets displayed by a conformance checker? |
| 05:28 | <MikeSmith> | JonathanNeal: Googling that quote leads me here: |
| 05:28 | <MikeSmith> | http://historymaker.wordpress.com/2010/01/13/a-husband-enjoying-his-wifes-body-by-pastor-phil/ |
| 05:28 | <MikeSmith> | "A Husband Enjoying His Wife’s Body, by Pastor Phil" |
| 05:28 | <MikeSmith> | "1. Start at the Bottom." |
| 05:51 | <MikeSmith> | I wonder who I should contact about updating http://www.chromium.org/developers/web-platform-status |
| 06:06 | <hsivonen> | hmmm. RDFa API mentioned in the context of TAG review of client-side storage APIs |
| 06:08 | <Hixie> | MikeSmith: collapse all the downplayed errors into one grayed out error at the bottom that says something like "...and 5 less important errors [show]" or something |
| 06:09 | <MikeSmith> | Hixie: OK, that'd be one way to do it in the Web UI |
| 06:09 | <MikeSmith> | but there's also the validator API |
| 06:10 | <Hixie> | there's lots of ways to do it in an api; the api can just have a boolean flag for each error that says whether it's "minor" or not, for instance |
| 06:11 | <MikeSmith> | well, then, the general problem is that I can't think of any similar apps that have a "downplayed error" level |
| 06:11 | <MikeSmith> | most have error, warning, informational |
| 06:12 | <MikeSmith> | with no additional level of granularity to distinguish different levels within "error" |
| 06:13 | <Hixie> | if you have to interoperate with such systems, put these into the "Warning" category" |
| 06:13 | <Hixie> | s/"$// |
| 06:13 | <MikeSmith> | I see |
| 06:13 | <MikeSmith> | anyway, I gotta get some food |
| 06:13 | <MikeSmith> | bbiab |
| 06:15 | <hsivonen> | "The SQL queries could be identified by URIs" |
| 06:17 | <hsivonen> | I'm not a fan of introducing downplayed errors alongside warnings esp. if the are close enough to be mapped to warnings for interop |
| 06:20 | <Hixie> | me either |
| 06:21 | <Hixie> | but i don't see how else they can be stuffed into an api that doesn't have the required level of detail |
| 06:21 | <Hixie> | i suppose you could also just put them into the "error" bucket with parentheses around them or something |
| 06:32 | <boblet> | MikeSmith: if only more of the religious right followed the lead of Pastor Phil, instead of being fixated on smiting |
| 06:44 | <JonathanNeal> | MikeSmith, yes that's an appropriate place to have Google lead you. |
| 07:15 | <annevk> | twitter tells me there's another flamewar on public-html |
| 07:15 | annevk | cannot wait to read it |
| 07:28 | <hsivonen> | annevk: flamewar about what? |
| 07:28 | hsivonen | is unable to maintain low unread counts on public-html |
| 07:28 | <annevk> | Hixie, in http://html5.org/tools/web-apps-tracker?from=4801&to=4802 you changed one into something else... |
| 07:29 | <annevk> | hsivonen, the tweet didn't say |
| 07:29 | <annevk> | and I haven't looked yet since I was being sarcastic |
| 07:30 | <othermaciej> | I am not sure what he was referring to |
| 07:31 | <othermaciej> | maybe a couple of messages from Roy? |
| 07:47 | <Hixie> | annevk: oops... which one? |
| 07:51 | <annevk> | Hixie, it now says "Run the following steps:" |
| 07:52 | <annevk> | rather than "Run these steps:" |
| 07:53 | <Hixie> | oh, i see |
| 07:53 | <Hixie> | thanks |
| 07:58 | <MikeSmith> | 869 messages this month on public-html |
| 07:58 | <MikeSmith> | so far |
| 07:59 | <MikeSmith> | 1459 messages last month |
| 07:59 | <MikeSmith> | 1459 x N people x N minutes per person |
| 08:00 | <rektide> | total inability to debug any SharedWorker code in Chromium is making me really really sad. :/ |
| 08:00 | <MikeSmith> | boblet: I think unfortunately Pastor Phil himself missed the irony of the term "bottom" in his "step 1. Start with the bottom" advice |
| 08:00 | <rektide> | had to go into the logs to see, hey, Chromium, you stopped re-requesting the page... even though its changing. |
| 08:00 | <MikeSmith> | rektide: nobody helped you out at all on that yet? |
| 08:01 | <rektide> | MikeSmith: zcorpan did show up w/r/t my base case, and he got that working |
| 08:01 | <boblet> | MikeSmith: nooo! don’t destroy my hope! |
| 08:01 | <MikeSmith> | boblet: this is case of "dramatic irony" |
| 08:01 | <rektide> | Drew Wilson also followed up /w a post on the Chromium-HTML5 group |
| 08:01 | <pesla> | MikeSmith: What has the N people to do with it? |
| 08:01 | <rektide> | http://groups.google.com/a/chromium.org/group/chromium-html5/browse_thread/thread/8a1a0042c8c32de3# |
| 08:02 | <MikeSmith> | pesla: read the message and try to figure out if there is any part of it -- even any small part -- that adds new information to the discussion, or that is actionable in any way |
| 08:03 | <rektide> | but now that the base case is out of hte way, i have MessageChannels flying everywhere and closures galore and i can only debug 1/3 the code... if i want better debugging i have to add new message channels to output the statu\s of the code I cant see |
| 08:03 | <MikeSmith> | rektide: you should get that feedback to JoePeck and/or #webkit-inspector at least |
| 08:03 | <MikeSmith> | not that it does you any good personally in the short term |
| 08:04 | <MikeSmith> | but could help others later, if they add some better debugging features for workers |
| 08:07 | <rektide> | drew wilson filed http://code.google.com/p/chromium/issues/detail?id=36577 |
| 08:08 | <rektide> | chromium is, so i'm told, lacking what capabilities do exist in webkit-inspector |
| 08:08 | <MikeSmith> | maybe we can suggest to people that they self-evaluate the content of their messages before sending them, and add [ignorable] if they are not providing any new information or if there's nothing actionable in the message |
| 08:08 | <rektide> | i dont believe i have any way to see whats in webkit-inspector, lacking a mac osx or windows machine. |
| 08:08 | <annevk> | naïve comes to mind |
| 08:09 | <MikeSmith> | rektide: huh? |
| 08:09 | <MikeSmith> | chrome runs on linux |
| 08:10 | <MikeSmith> | arora has web inspector too |
| 08:10 | <MikeSmith> | arora = browser based on Qt port of Webkit |
| 08:12 | <MikeSmith> | rektide: right-click and Inspect Element ? |
| 08:18 | <hsivonen> | the v8 test suite makes assumptions about how a race between a timeout and the HTML parser finishes |
| 08:18 | <hsivonen> | Web <3 |
| 08:19 | <Hixie> | i guess the spec doesn't put those two into the same queue, huh |
| 08:20 | <hsivonen> | Hixie: I'd expect that to be a spec bug then |
| 08:20 | <Hixie> | is this a 0-second timeout? |
| 08:20 | <hsivonen> | Hixie: so the v8 suite assumes that the event loop doesn't spin after an inline script |
| 08:20 | <hsivonen> | 25 ms |
| 08:20 | <Hixie> | oh, ok, that's different |
| 08:21 | <Hixie> | so if the inline script is the last thing, it wants the EOF handled in the same task? |
| 08:21 | <Hixie> | as the </script>? |
| 08:21 | <hsivonen> | no, it wants a div after the script to parse in the same task as the script |
| 08:21 | <Hixie> | oh well that's definitely a bug in the page then, we can't guarantee that kind of thing, it might be network-bound |
| 08:22 | <hsivonen> | Hixie: indeed, it's a bug in the test suite |
| 08:22 | <Hixie> | k |
| 08:22 | <hsivonen> | but it's still my problem |
| 08:23 | <annevk> | you have to be compatible with the page or something? |
| 08:23 | <annevk> | if so the spec should prolly say something about it |
| 08:23 | <annevk> | though it seems kind of odd |
| 08:23 | <hsivonen> | annevk: yeah, because it's part of Mozilla's test suites |
| 08:24 | <hsivonen> | anyway, I expect another perf fix to paper over this |
| 08:24 | <hsivonen> | just whining about test suites |
| 08:24 | <othermaciej_> | hsivonen: you should patch the copy in Mozilla's test suite IMO |
| 08:25 | <hsivonen> | othermaciej_: that, too |
| 08:26 | <beilabs> | Hey guys, I've been playing around a little with HTML5/cached pages for offline use....is it possible to run a script to cache many different pages, allowing for the user to have full use of the site in an offline context without them have to visit each one individually? |
| 08:27 | <annevk> | beilabs, you can list the pages in the manifest |
| 08:27 | <annevk> | beowulf, that ought to do it |
| 08:28 | <annevk> | that was not meant for beowulf but you get the idea... |
| 08:28 | hsivonen | decides to try to increse the sniffing limit to 1K as in WebKit in order to work around bogus Apache instances out there |
| 08:28 | <hsivonen> | including the one at lists.whatwg.org |
| 08:28 | <annevk> | o_O |
| 08:28 | <beilabs> | annevk, so just having 10 pages in the manifest will allow the user to browse all 10 pages even only after visiting one single page....the other 9 will be cached? |
| 08:29 | <Hixie> | hsivonen: please let adam know what you decide is the minimum required for compatibility |
| 08:29 | <beilabs> | just thinking about a concept app...sorry for my silly questions |
| 08:29 | <annevk> | beilabs, as long as the user visits the one that points to the manifest, yes |
| 08:30 | <beilabs> | annevk, ahh cool...thanks. |
| 08:30 | <hsivonen> | Hixie: I'm trying 1K, because I recall someone (ap?) saying that WebKit went to 1K and that worked |
| 08:30 | <hsivonen> | Hixie: but yeah, I'll let Adam know if 1K works |
| 08:30 | annevk | forgot how it worked when they all point to he same manifest |
| 08:31 | <asmodai> | g'morning. |
| 09:00 | <Hixie> | man, this autobuffer attribute (now preload) is the most weak-sauce attribute in the language |
| 09:01 | <Hixie> | it has like no UA criteria at all except the IDL attribute, when you get right down to it |
| 09:03 | <Hixie> | hsivonen: i don't suppose there's any way i could convince you to add some sort of support for the equivalent of #file and #line to the validator, is there? |
| 09:03 | <Hixie> | hsivonen: the "offset the line numbers by this much" thing stopped working when i started validating the post-preprocessor versions of the files |
| 09:09 | <doublec> | is autobuffer officially preload now? |
| 09:10 | <annevk> | it seems so |
| 09:10 | <annevk> | http://html5.org/tools/web-apps-tracker?from=4810&to=4811 |
| 09:12 | <Hixie> | just checked it in |
| 09:12 | <Hixie> | mere seconds ago |
| 09:12 | <zcorpan> | Hixie: thanks for addressing autobuffer |
| 09:14 | zcorpan | notes that "none" is the added feature |
| 09:14 | <zcorpan> | adding a "nobuffer" attribute would be more compatible with shipping firefox |
| 09:15 | <Hixie> | "metadata" is also an added feature |
| 09:15 | <doublec> | is the spec regenerating itself or something? All the multipages are 404 |
| 09:16 | <Hixie> | before, you had two options -- UA does what it likes, and UA does what it likes but probably will buffer a lot |
| 09:16 | <Hixie> | doublec: hm, seems to have broken, hold on |
| 09:17 | <zcorpan> | Hixie: and how we have three options -- UA does what it likes, UA does what it likes and UA does what it likes? :) |
| 09:17 | <Hixie> | four options |
| 09:17 | <zcorpan> | oh yeah sorry |
| 09:18 | <annevk> | at some point HTML5 needs to say something about https://bugzilla.mozilla.org/show_bug.cgi?id=524223 I think |
| 09:18 | <Hixie> | UA does what it likes (likely to be equivalent to one of the other three unless it invents a fourth state), UA does what it likes but likely downloads nothing, UA does what it likes but probably will buffer only metadata and first frame, and UA does what it likes but will either buffer until canplaythrough fires or until load fires. |
| 09:18 | <MikeSmith> | doublec: multipages at whatwg or at w3c? |
| 09:19 | <annevk> | i.e. under what circumstances something will be treated as a CSS style sheet; the rules changed |
| 09:19 | <doublec> | MikeSmith, whatwg. It's working now though. |
| 09:19 | <MikeSmith> | k |
| 09:19 | <hsivonen> | annevk: do you think it belongs in HTML5 because the CSS WG didn't address the issue? |
| 09:20 | <hsivonen> | Hixie: there's pretty much no way to convince me to add any validator features right now, sorry. I'm trying to get the HTML5 parser in Gecko into a state where we can flip the switch and turn it on by default |
| 09:20 | <gsnedders> | hsivonen: Is it not more an issue of whether you use the link element or not? |
| 09:20 | gsnedders | looked at that back in October when it was reported, and can't remember the detail entirely |
| 09:21 | <gsnedders> | hsivonen: Esp. in the case when what the link element points to isn't served as text/css |
| 09:21 | <Hixie> | doublec: yeah, dunno what happened there. seems fixed now. |
| 09:21 | <Hixie> | doublec: sometimes the rather brittle pipeline between all the services i use breaks :-) |
| 09:21 | <Hixie> | hsivonen: k |
| 09:21 | <doublec> | hehe |
| 09:22 | <Hixie> | hsivonen: maybe i should work on MikeSmith instead :-) |
| 09:22 | <Hixie> | hsivonen: is there some way to make the text/plain output to output the actual broken line at all? |
| 09:22 | <Hixie> | hsivonen: that would help a lot |
| 09:22 | Philip` | hopes the multipage problem wasn't his fault |
| 09:22 | <annevk> | hsivonen, because HTML5 defines the conditions now too I guess |
| 09:23 | <Hixie> | Philip`: no i think it was some side-effect of my bugzilla script dying when it tried to send ISO-8859-1 as UTF-8 |
| 09:23 | <Hixie> | i really should fix that at some point |
| 09:23 | <Hixie> | i think bugzilla is probably not sending back appropriate headers |
| 09:27 | <annevk> | I guess I could file a bug on the issue, but I wonder if that's the appropriate way to track it |
| 09:28 | <Hixie> | i commented on the bug |
| 09:29 | <MikeSmith> | Hixie: about the line-reporting thing, might be good to file a bug/enhancement request on bugzilla.validator.nu .. I can try to look myself at what'd take to add it. also, I did a presentation/demo here on validator.nu the other day and there were some people there who were interested in contributed but asked how many interesting/useful features there were left to add |
| 09:29 | <annevk> | btw, so bufferedAmount does not include framing overhead, but does include it being encoded as UTF-8? |
| 09:29 | <Hixie> | annevk: yeah |
| 09:29 | <Hixie> | MikeSmith: will do |
| 09:30 | <annevk> | somewhat weird imo |
| 09:30 | <Hixie> | MikeSmith: i imagine that there's a ton of useful work that can be done in making the validator more user-friendly |
| 09:30 | <MikeSmith> | yeah, definitely |
| 09:30 | <Hixie> | annevk: comment on the public-webapps list (smaug mailed the list iirc) |
| 09:30 | <MikeSmith> | but this request has an actual person behind it right now, waiting for it get added. which makes a big difference of course |
| 09:31 | <nessy> | I'm almost sorry I got so curious about the canvas accessibility discussion - that curiosity is bound to make me only enemies :-) |
| 09:31 | <Hixie> | MikeSmith: what component? |
| 09:31 | <Hixie> | nessy: i hear ya |
| 09:31 | <zcorpan> | "The above rules will block the attack because an HTML, XML etc. header will |
| 09:31 | <zcorpan> | always cause a broken first CSS descriptor." -- <!-- is valid css |
| 09:31 | <zcorpan> | from https://bugzilla.mozilla.org/show_bug.cgi?id=524223#c4 |
| 09:31 | <nessy> | Hixie: yeah, I know :S |
| 09:32 | <MikeSmith> | Hixie: "general" component, I guess |
| 09:33 | <MikeSmith> | nessy: I will put together surveys on the two media proposals, and point you at the drafts of those so you can suggest refinements |
| 09:33 | <nessy> | MikeSmith: I'm not worried about those two proposals |
| 09:33 | <MikeSmith> | nessy: surveys similar to the recent ones on the canvas and summary-details proposals |
| 09:33 | <nessy> | MikeSmith: I am more worried about the format question |
| 09:34 | <MikeSmith> | which particular format question |
| 09:34 | <nessy> | the one which external formats get supported |
| 09:34 | <MikeSmith> | ah |
| 09:34 | <nessy> | it's actually a question relatively independent of those two specs |
| 09:35 | <nessy> | there is only a single line pointing to formats in the TextAssociations wiki page |
| 09:35 | <MikeSmith> | well, as far as the proposals, I'm not worried if you're not -- but the surveys are a sorta necessary formality |
| 09:35 | <nessy> | ah ok |
| 09:35 | <nessy> | well, then can we have three surveys? |
| 09:36 | <MikeSmith> | necessary at least as far as the conventions of this particular group (a11y TF) |
| 09:36 | <MikeSmith> | nessy: sure |
| 09:36 | <nessy> | I'm happy to help formulating them |
| 09:36 | <nessy> | thanks |
| 09:36 | <MikeSmith> | nessy: if you have time to draft up some text for the surveys and e-mail it to me, let's do that |
| 09:36 | <nessy> | ok, cool |
| 09:36 | <MikeSmith> | you can use the canvas survey and summary-details survey as models |
| 09:36 | <nessy> | does it have to be ready before tomorrow's meeting? |
| 09:36 | <Hixie> | hsivonen, MikeSmith: http://bugzilla.validator.nu/show_bug.cgi?id=720 |
| 09:37 | <MikeSmith> | nessy: models just for just the text.. I will take care of setting up the actual survey |
| 09:37 | <MikeSmith> | Hixie: thanks |
| 09:42 | <annevk> | oh crap |
| 09:42 | <annevk> | html5-diff needs an update now for preload |
| 09:42 | <Hixie> | heh |
| 09:43 | <Philip`> | Is it intentional that |
| 09:43 | <Philip`> | <time> |
| 09:43 | <Philip`> | 12:34</time> |
| 09:43 | <Philip`> | is seemingly non-conforming, because LF is not White_Space? |
| 09:44 | <gsnedders> | LF is White_Space |
| 09:44 | <gsnedders> | http://unicode.org/Public/UNIDATA/PropList.txt |
| 09:44 | <gsnedders> | LF is U+000A which is in the range 0009..000D |
| 09:45 | <Philip`> | Someone at http://lists.w3.org/Archives/Public/www-html/2010Jan/0010.html seemingly indicated it wasn't |
| 09:45 | <Hixie> | i saw someone else say White_Space wasn't LF recently |
| 09:45 | <Philip`> | but maybe that's a confusion between WS and White_Space |
| 09:45 | <Hixie> | didn't have time to investigate |
| 09:45 | <Philip`> | Oops, I didn't realise that was a month old |
| 09:45 | <hsivonen> | Hixie: speaking of white space, did you actually locate an email from me asking for to expand to LF? |
| 09:45 | <gsnedders> | Well, the UCD as of Unicode 5.2.0 (i.e., the above link) certainly has LF |
| 09:46 | <Hixie> | hsivonen: didn't look |
| 09:46 | <annevk> | hsivonen, he did |
| 09:46 | <Philip`> | http://unicode.org/Public/UNIDATA/UnicodeData.txt doesn't say WS |
| 09:47 | <annevk> | hsivonen, oops, nm |
| 09:47 | <hsivonen> | Hixie: may I assume the decision will be reversed regardless? |
| 09:47 | <Philip`> | Maybe some of the other letters are magic |
| 09:47 | <gsnedders> | White_Space: "Separator characters and control characters which should be treated by programming languages as "white space" for the purpose of parsing elements." |
| 09:47 | <gsnedders> | 'Note: There are other senses of "whitespace" that encompass a different set of characters.' |
| 09:47 | <Hixie> | hsivonen: why would we reverse it? |
| 09:48 | <gsnedders> | Philip`: Why would it? |
| 09:48 | <Philip`> | gsnedders: Because WS sounds a bit like White_Space |
| 09:48 | <gsnedders> | Ah. |
| 09:48 | <gsnedders> | There is White_Space, which is a binary property |
| 09:48 | <gsnedders> | There is also "WS" which is a possible value for Bidi_Class |
| 09:49 | <gsnedders> | Then there is the Zs (Space_Separator) in General_Category |
| 09:49 | <Philip`> | Ah |
| 09:49 | <hsivonen> | Hixie: because it's a gratuitous change from what all four engines do now |
| 09:49 | <Philip`> | Unicode is hard |
| 09:50 | gsnedders | mumbles it-really-isn't-that-hard |
| 09:51 | <Philip`> | It's harder than not supporting text |
| 09:51 | <gsnedders> | Well, yes, textual communication is overrated. |
| 09:53 | Philip` | hasn't even worked out which of the files in http://unicode.org/Public/UNIDATA/ contain original data and which are derived from others |
| 09:54 | <zcorpan> | hsivonen: is it a web compat problem? |
| 09:54 | <hsivonen> | zcorpan: I don't know yet. It's a mochitest compat problem. |
| 09:56 | <Hixie> | hsivonen: it was very intentional: http://html5.org/tools/web-apps-tracker?from=941&to=942 |
| 09:56 | <Hixie> | hsivonen: and it affects the parser quite dramatically |
| 09:56 | <hsivonen> | Hixie: dramatically? |
| 09:57 | <Hixie> | well, maybe "dramatically" is overstating the case |
| 09:57 | <Hixie> | it affects the tree construction phase in a dozen places |
| 09:57 | <hsivonen> | Hixie: only if you want to treat as whitespace |
| 09:57 | <hsivonen> | Hixie: but even so, not a huge deal |
| 09:59 | <Hixie> | http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-June/thread.html#11689 is the thread |
| 10:00 | <Hixie> | i guess it was anne requesting it, not you |
| 10:00 | <hsivonen> | I'm vindicated! |
| 10:00 | <Hixie> | indeed you said then what you're saying now :-) |
| 10:00 | <Hixie> | http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-June/011703.html |
| 10:01 | <hsivonen> | I have foresight, too. :-) |
| 10:01 | <gsnedders> | You have to live in exactly the moment, now. |
| 10:01 | <Hixie> | reopen the bug or send mail or whatever, so it gets on my queue -- if anyone disagrees, they should speak up |
| 10:01 | <hsivonen> | ok |
| 10:01 | <annevk> | i still think normalizing it makes sense |
| 10:02 | <annevk> | to death with CR |
| 10:02 | <hsivonen> | annevk: then Opera should ship it first to gauge the Web compat effect :-) |
| 10:06 | <smaug> | annevk: hey, you're about to send a message about bufferedAmount to webapps |
| 10:06 | <annevk> | not really |
| 10:06 | <smaug> | I'm not sure my email ever got through |
| 10:06 | <smaug> | since I wasn't using my usual email account |
| 10:06 | <annevk> | I'm not too interested in its semantics to be honest |
| 10:06 | <annevk> | just seemed kind of odd how it worked |
| 10:06 | <smaug> | ah |
| 10:07 | <smaug> | it is still odd |
| 10:07 | <smaug> | unless Hixie just changed it |
| 10:08 | <Hixie> | i just changed it as you asked |
| 10:08 | <Hixie> | UTF-8 bytes, no overhead bytes |
| 10:08 | <smaug> | ah, thanks |
| 10:10 | <rektide> | question; did the "expose EventTarget within JavaScript" thought-train ever get anywhere? http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-April/019387. |
| 10:10 | <smaug> | another thing is message event. What values should origin and eventid etc have . I assume null. |
| 10:11 | <rektide> | zcorpan: thanks for all the help with my SharedWorker example. you nailed all the issues i was having. |
| 10:11 | <Hixie> | i think that's defined, but yeah, null, empty string, etc, whatever the "empty" value is for the relevant types |
| 10:12 | <Hixie> | it's defined with the event interface iirc |
| 10:12 | <smaug> | Hixie: but null or empty string? |
| 10:12 | <smaug> | ah, I'll check there |
| 10:12 | <Hixie> | "Except where otherwise specified, when the user agent creates and dispatches a message event in the algorithms described in the following sections, the lastEventId attribute must be the empty string, the origin attribute must be the empty string, the source attribute must be null, and the ports attribute must be null." |
| 10:12 | <smaug> | right |
| 10:13 | <zcorpan> | rektide: np |
| 10:13 | <Hixie> | rektide: i don't think it ever got sent to the more appropriate list, but i don't follow that other list as carefully, so dunno |
| 10:14 | <rektide> | Hixie: thanks, i'll do some archival work & post if its still not mentioned. |
| 10:15 | <smaug> | there was some discussion about EventTarget JS objects |
| 10:15 | <smaug> | in w3c mailing lists |
| 10:15 | <smaug> | webapps or dom |
| 10:15 | <smaug> | but IIRC it was still missing good use cases |
| 10:16 | <rektide> | well, if you ever want to make an object that can publish events, it'd be handy to have |
| 10:17 | <smaug> | "publish events" ? |
| 10:17 | <rektide> | i just cant really imagine a spec that would help with the nitty gritty; |
| 10:17 | <smaug> | but that is not a use case |
| 10:17 | <smaug> | want to make an object to do this or that isn't a use case ;) |
| 10:18 | <rektide> | i have a use case, i dont really care to explain it or reduce it to mortal terms. |
| 10:21 | <MikeSmith> | rektide: OK if I borrow that sentence from you for later use of my own? |
| 10:22 | <MikeSmith> | I like that sentence |
| 10:22 | <nessy> | Hixie: nice work on the preload attribute |
| 10:22 | <jgraham> | It puits me in the mind of Fermat |
| 10:22 | <MikeSmith> | heh |
| 10:22 | <rektide> | be my guest MikeSmith |
| 10:22 | <nessy> | Hixie: I think you missed something though, but let me run it past you |
| 10:22 | <gsnedders> | jgraham: You're being very philisophical today… |
| 10:22 | <jgraham> | You think? |
| 10:23 | <gsnedders> | Certainly far moreso than normal |
| 10:23 | <nessy> | Hixie: when autoplay is used, the video should regard preload as set to "auto" - no? |
| 10:23 | <gsnedders> | That isn't hard, though |
| 10:24 | <Hixie> | nessy: so long as it plays, i don't really mind what buffering happens if autoplay="" is set |
| 10:25 | <nessy> | Hixie: yeah, so it is up to the browser, i.e. "auto" ? |
| 10:25 | <Hixie> | the preload attribute is just ignored if autoplay is set |
| 10:25 | <remysharp> | I'm trying to clarify a point - does HTML5 (and various implementations of specs surrounding it) extend: a) the DOM, b) the BOM (browser object model) or c) both (I thought it was this, but not completely sure now I've gone over and over in my head)? |
| 10:26 | <nessy> | Hixie: I guess it's self-evident that neither metadata nor none would be appropriate, but it may make sense to state the obvious |
| 10:27 | <hsivonen> | Filed the CR thing as http://www.w3.org/Bugs/Public/show_bug.cgi?id=9144 |
| 10:27 | <Hixie> | nessy: the spec says the attribute is ignored altogether, so it doesn't matter what value it's treated as |
| 10:27 | <nessy> | ah I see: so it's up to the browser anyway - ok, I get it |
| 10:28 | <Philip`> | remysharp: Depends what you mean by DOM and BOM, I guess |
| 10:28 | <Hixie> | nessy: well how preload="" is handled is entirely up to the browser anyway, so it's kinda hard to tell the difference :-) |
| 10:28 | <Hixie> | nessy: it's just a hint |
| 10:28 | <Philip`> | remysharp: I don't think HTML5 redefines the basic DOM things like getAttribute and createElement etc - it just defers to DOM 3 Core for that |
| 10:29 | <nessy> | fair enough :) |
| 10:29 | <remysharp> | Philip`: so things like the applicationCache, would you say that's BOM then? |
| 10:29 | <Philip`> | remysharp: but it redefines all the interfaces that HTML pages expose to scripts via objects in the DOM |
| 10:29 | <Philip`> | remysharp: I would never use the term "BOM" :-) |
| 10:30 | <Philip`> | remysharp: (except for byte order marks) |
| 10:30 | <annevk> | remysharp, BOM means byte-order-mark |
| 10:30 | <annevk> | remysharp, we refer to everything as DOM usually |
| 10:30 | <remysharp> | Philip`: okay cool. That makes sense. I've been staring at the question so long none of the letters made much sense anymore! |
| 10:30 | <annevk> | remysharp, including e.g. XMLHttpRequest, applicationCache, etc. |
| 10:30 | <remysharp> | cheers for the help chaps. |
| 10:58 | <annevk> | I'd like to add some kind of overview table to http://dev.w3.org/html5/html4-differences/ but I can't really work out what it should look like |
| 10:59 | <zcorpan> | overview of what? |
| 10:59 | <annevk> | what changed and what stayed the same between HTML4/HTML5 |
| 10:59 | <Hixie> | get a png of the whole of html4, and the whole of html5, with width 10px or something, and draw confident looking lines from one picture to the other showing how things have moved to different parts of the spec |
| 11:00 | <Hixie> | and then wait and see if anyone actually asks about it |
| 11:00 | <annevk> | merging DOM2HTML and HTML4 for such a table would be nice though maybe APIs and markup should stay separate |
| 11:00 | <annevk> | Hixie, I'll get right on that! |
| 11:00 | <Hixie> | oh i wasn't suggesting the lines should be real, btw |
| 11:00 | <Hixie> | hence confident-_looking_ |
| 11:06 | <Philip`> | You could draw a graph of HTML version number vs awesomeness |
| 11:09 | <Dashiva> | How would that handle the unversioned future? |
| 11:19 | <annevk> | Dashiva, NaN exception |
| 11:42 | <Philip`> | Dashiva: Just extrapolate the line into extreme awesomeness |
| 11:42 | <Dashiva> | But where on the x axis? |
| 11:42 | <Philip`> | On the right |
| 11:43 | <Philip`> | because it's more awesome than HTML5, and the line slopes upwards |
| 11:43 | <Philip`> | so it must go there |
| 11:45 | <Hixie> | ok i'm going skiing, should be back in about 40 hours. |
| 11:45 | <virtuelv> | How does it handle HTML6 |
| 11:46 | <virtuelv> | 6.8.3, even, these days |
| 11:48 | <annevk> | Hixie, you can go skiing in California? |
| 11:48 | <Hixie> | we have a mountain |
| 11:48 | <Hixie> | it's not very good skiing |
| 11:48 | <annevk> | downhill? |
| 11:48 | <virtuelv> | annevk: www.skicalifornia.com |
| 11:48 | <Hixie> | i mean it's no bruson or verbier or bergen or anything |
| 11:49 | <annevk> | didn't know you did skiing |
| 11:49 | annevk | wants to go skiing again too |
| 11:49 | <Hixie> | i'm swiss! |
| 11:49 | <Hixie> | didn't know i did skiing |
| 11:49 | <Hixie> | next you'll say you didn't know i liked cheese! |
| 11:49 | <annevk> | fair enough |
| 11:50 | <annevk> | or fondue? |
| 11:50 | annevk | doesn't like fondue |
| 11:50 | <virtuelv> | or guns |
| 11:50 | <Dashiva> | Hey, there are plenty of Norwegians who swear off skiing for life once they reach 18 |
| 11:51 | <Hixie> | fondue is cheese |
| 11:51 | <annevk> | when you say skiing in Norway they always assume cross-country |
| 11:51 | <Hixie> | yeah, crazy norwegians |
| 11:51 | <Hixie> | in french we call than "ski de fond" |
| 11:51 | <Hixie> | i call it "walking" |
| 11:52 | <Hixie> | specs are generated for publication today, btw, in case the team decides to give the ok |
| 11:52 | <virtuelv> | annevk: you should try downhill skiing on cross-country skis sometime |
| 11:52 | <hsivonen> | I've done cross-country skiing multiple times this winter. for the first time since military service. |
| 11:52 | <Hixie> | if changes are needed, i guess i'll do them this weekend |
| 11:53 | virtuelv | is considering buying cross-country skis again, for the first time since around my military service |
| 11:53 | <virtuelv> | (I spent my naval days in a now-closed mountain in 93-94) |
| 11:55 | <Hixie> | ok i'm really gone |
| 11:55 | karlcow | loves cross-country skiing but didn't have many opportunities these last few years |
| 11:56 | <annevk> | virtuelv, I'm having enough trouble as it is on my normal skis on black |
| 11:57 | <annevk> | I'm not falling anymore, but I'm going way faster than other people and not always intentionally so... |
| 11:57 | <virtuelv> | annevk: is tryvann the only black course you've tried? |
| 11:57 | <annevk> | yeah, I hear it's not a very difficult black course, at least according to Lachy |
| 11:58 | <virtuelv> | I've actually never skied there, but given the terrain around there, I can't imagine it being more than blue/red |
| 11:59 | <annevk> | it's quite a bit steeper than their red/blue tracks |
| 11:59 | <annevk> | but maybe those are actually green? and their green is nothing? |
| 11:59 | <annevk> | I've only been to one other place in Oslo so I don't have much to compare with |
| 12:00 | <Dashiva> | I can't imagine Tryvann having a black course |
| 12:29 | <Lachy> | annevk, the problem with the black run at Tryvann is that it's a completely straight, fully groomed run. It's just fun because it's long and designed for high speed. |
| 12:38 | <asmodai> | Did HTML ever allow the valign attribute to have "center" as value? |
| 12:45 | <gsnedders> | asmodai: Off the top of my head, no |
| 12:45 | <asmodai> | Guess they meant align then |
| 12:46 | <virtuelv> | Lachy: *no* slope in Norway is designed for high speed if you have to carry a lift card |
| 12:46 | <virtuelv> | (in other word on open days) |
| 13:04 | <Philip`> | asmodai: I think there was some issue like you couldn't have the same keyword in two different enumerated attributes |
| 13:04 | <Philip`> | so it was align="center" and valign="middle" |
| 13:04 | <Philip`> | (due to weirdness of SGML DTDs) |
| 13:04 | <Philip`> | though I could be entirely making that up |
| 13:05 | <zcorpan> | i think that's accurate; sgml had a feature where you could omit the attribute name and just have the value |
| 13:05 | <zcorpan> | like <td center middle> |
| 13:05 | <Philip`> | Did that feature have to be explicitly enabled for each attribute in the DTD? |
| 13:05 | <zcorpan> | no, it's enabled for all enumerated attributes |
| 13:07 | <zcorpan> | (there's no concept of boolean attributes in sgml; they're just enumerated with one value) |
| 13:07 | <Philip`> | Hmm, indeed |
| 13:07 | <TabAtkins> | I really wonder what the SGML people were smoking sometimes. |
| 13:07 | <Philip`> | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"><title></title><table><tr><td center middle></table> |
| 13:07 | <Philip`> | validator.w3.org says that's valid |
| 13:07 | <asmodai> | Philip`: They actually did mean align="center". Just replaced it with some CSS to text-align it. |
| 13:08 | <Philip`> | I like how <table valign=middle> causes the error message "Attribute "VALIGN" is not a valid attribute. Did you mean "valign"?" |
| 13:09 | <zcorpan> | has someone reported a bug about that? |
| 13:10 | <asmodai> | Philip`: heuh? |
| 13:10 | <TabAtkins> | Does anyone recall any improvements in IE8 that can be classified as "implementing HTML4 better"? |
| 13:10 | <TabAtkins> | I don't particular know of any lacks in their HTML4 implementation that they fixed in IE8. |
| 13:11 | <zcorpan> | <abbr>? <object>? |
| 13:11 | <Philip`> | zcorpan: http://lists.w3.org/Archives/Public/www-validator/2010Feb/0014.html mentions the bug |
| 13:11 | <zcorpan> | <img alt> |
| 13:11 | <TabAtkins> | What was wrong with their <img alt>? |
| 13:11 | <Philip`> | <q>? |
| 13:12 | <zcorpan> | they showed a tooltip |
| 13:12 | <TabAtkins> | Well, I suppose they showed it in a tooltip. |
| 13:12 | <TabAtkins> | Yeah. |
| 13:12 | <TabAtkins> | That one's not really implementing HTML4 better (closer to the spec), just changing a UI issue to better match expectations, right? |
| 13:12 | <zcorpan> | they changed the parser for end tags |
| 13:13 | <TabAtkins> | Ah, I really had no idea about that. |
| 13:13 | <zcorpan> | at least </li> |
| 13:14 | <zcorpan> | which became incompatible with everyone else |
| 13:15 | <virtuelv> | TabAtkins: you're implying the SGML people only smoked some times? |
| 13:16 | <TabAtkins> | virtuelv: Presumably they were sane when they were around their families. |
| 13:16 | <virtuelv> | you assume they had families? |
| 13:17 | <Philip`> | Helps them blend in to society, while they pursue their madcap markup schemes |
| 13:19 | <zcorpan> | in the future people will imply we were smoking for making the platform overly insane for compat with a mere few billion documents |
| 13:22 | <jgraham> | What do you mean "in the future" people ask that all the time in the present |
| 13:22 | <jgraham> | And in the past it was the theory that gave us XHTML 2 |
| 13:22 | <nessy> | MikeSmith: just sent you the three questions to vote on |
| 13:23 | <MikeSmith> | nessy: thanks very much |
| 13:23 | <nessy> | hope it helps :) |
| 13:24 | <nessy> | gotta grab some sleep now - got an early meeting tomorrow |
| 13:24 | <MikeSmith> | OK, nn |
| 13:24 | <nessy> | cya nn |
| 16:38 | <JonathanNeal> | Goodmorning! |
| 16:39 | <AryehGregor> | Good morning! |
| 16:40 | <JonathanNeal> | AryehGregor, thanks for your help the other day with <menu> and <command> stuff, I was able to finish a new HTML5 outline I'm proposing to put in our product (you can see it @ http://sandbox.thewikies.com/html5-layout/ ) |
| 16:42 | <JonathanNeal> | The bottom links point to alternate css for the page. |
| 16:47 | <Dashiva> | MikeSmith: What's バリデーター? |
| 16:50 | <JonathanNeal> | Why, dominoes. |
| 17:09 | <gsnedders> | Dashiva: Japanese |
| 17:10 | <Philip`> | http://www.google.com/search?q=%E3%83%90%E3%83%AA%E3%83%87%E3%83%BC%E3%82%BF%E3%83%BC highlights the word "validator" on a lot of pages |
| 17:11 | <Philip`> | so maybe it's that |
| 17:23 | <Dashiva> | Oh, duh |
| 17:23 | <Dashiva> | I read the last part as 'data' |
| 17:44 | <asmodai> | Dashiva: it says baride-ta- -> validator |
| 17:49 | <MikeSmith> | Dashiva: validator |
| 17:50 | <MikeSmith> | ah, what asmodai said |
| 17:50 | AryehGregor | figured that out a couple of hours ago thanks to Google Translate |
| 17:50 | <Dashiva> | I know enough Japanese to know that google translate is not all that reliable :) |
| 17:53 | <AryehGregor> | Well, it seemed to fit awfully well. |
| 18:08 | <MikeSmith> | hmm, I guess バリデータ is the more common way to write it, rather than バリデーター |
| 18:09 | <MikeSmith> | I suppose I should have checked that before tweeting it |
| 18:09 | <MikeSmith> | oh well |
| 18:09 | <MikeSmith> | that common usage sounds wrong to me, though |
| 18:09 | <MikeSmith> | バリデーター makes it a long "a" sound at the end |
| 18:10 | <MikeSmith> | with is the usual way of transliterating a final "r", like in "er" |
| 18:10 | <MikeSmith> | which there is no real equivalent for in Japanese |
| 18:10 | <MikeSmith> | バリデータ makes the last part sound more like "data", like Dashiva points out |
| 18:11 | <MikeSmith> | anyway, a lot of common ways of transliterating stuff in Japanese a plain wrong |
| 18:11 | <MikeSmith> | like マイケル for Michael |
| 18:12 | <MikeSmith> | it should be マイクル or even マイコ |
| 18:14 | gsnedders | almost fails to notice the difference between the first two |
| 18:17 | <Dashiva> | MikeSmith: It makes more sense if you assume the people transliterating it only knew the written form and not the pronounciation |
| 18:17 | <MikeSmith> | yeah, I know |
| 18:17 | <MikeSmith> | but it's actually changing |
| 18:18 | <MikeSmith> | I see more and more people on twitter and such who are dispensing with the bookish canonical transliterations of names and words and instead just writing them closer to the way they really sound |
| 18:19 | <Dashiva> | I'm guessing tech-oriented Japanese have to be able to transliterate independently just to keep up with new terms |
| 18:20 | <Dashiva> | Or are there a semi-official "authorities" for that? |
| 18:21 | <asmodai> | Not really. |
| 18:21 | <asmodai> | There's the G5 collection and some katakana dictionaries, but most of it is quite informal. |
| 18:21 | <asmodai> | Nothing like Kojien, Meikyou or such dictionaries tend to really touch it, unless the term is so common. |
| 19:03 | <AryehGregor> | Hmm, why does this say Chrome support for Theora is "Not reliable"? http://openvideoalliance.org/wiki/index.php?title=Playback |
| 19:06 | <annevk> | weren't they the guys working on a new codec? |
| 19:06 | <annevk> | oh, maybe that's yet another thing |
| 19:06 | <kinetik> | annevk: that's OMS |
| 19:08 | <AryehGregor> | MikeSmith, you know, if your name is trademarked, I'm pretty sure you've lost rights to the trademark by now via dilution. I had to sift through a whole lot of Google results to find your Twitter feed. |
| 19:11 | <zcorpan> | AryehGregor: i've heard the sound is not synced correctly with the video in chrome for ogg/theora/vorbis |
| 19:11 | <AryehGregor> | I noticed that once, I think. |
| 19:12 | <annevk> | OMS being http://openmediacommons.org/ |
| 19:13 | <annevk> | last update June 2009 |
| 19:13 | <zcorpan> | what does second life use? |
| 19:14 | <zcorpan> | does it have a browser? |
| 19:15 | <kinetik> | annevk: i've heard that OMS is dead, but there doesn't seem to be anything saying so publically |
| 19:31 | <zcorpan> | > So, perhaps what would be easier for all of you is to simply have an |
| 19:31 | <zcorpan> | > attribute that says include the subtree in the navigable document structure |
| 19:31 | <zcorpan> | > when canvas is rendered? This is not the case now. |
| 19:31 | <zcorpan> | i thought it actually was the case now (in the spec) |
| 19:38 | <annevk> | it is, it's the reverse that's not possible |
| 19:38 | <annevk> | (though it's unclear whether that's wanted) |
| 19:39 | <zcorpan> | i guess it's possible by adding tabindex=-1 to otherwise focusable elements in the subtree |
| 19:39 | <zcorpan> | (with script) |
| 19:39 | <zcorpan> | but it seems simpler to just do canvas.innerHTML = "" |
| 19:51 | <Philip`> | zcorpan: http://wiki.secondlife.com/wiki/Web_Browser |
| 19:51 | <Philip`> | says Mozilla |
| 19:52 | <Philip`> | They've done work on WebKit porting too but I don't know how far that went |
| 19:52 | <Philip`> | http://wiki.secondlife.com/wiki/WebKit_worklist |
| 20:17 | <annevk> | Voor iedereen die Nederlands spreekt: http://twitter.com/#/list/DonaldDuckNL/duckstad (via #fronteers) |
| 20:28 | zcorpan__ | heeft geen twitter account |
| 20:29 | <annevk> | zijn er geen feeds? |
| 20:29 | <svl> | http://twitter.com/DonaldDuckNL/duckstad is the url not requiring an account |
| 20:29 | <annevk> | oh my bad |
| 20:29 | <annevk> | thanks svl |
| 20:48 | <zcorpan__> | ah, is there green light for publication now? |
| 21:00 | <AryehGregor> | Haha: http://img.moronail.net/img/1/0/2010.jpg |
| 21:00 | AryehGregor | has never watched CSI, and doesn't want to :P |
| 21:00 | <miketaylr> | you pretty much just watched every episode ever, AryehGregor |
| 21:01 | <AryehGregor> | I hope that isn't an actual sequence from the show? |
| 21:02 | <annevk> | CSI is for those who do not understand The Wire |
| 21:02 | <annevk> | or something |
| 21:02 | <miketaylr> | no, but it's pretty close |
| 22:45 | <zcorpan__> | Hixie: the message in dom-intro boxes looks like issue boxes. could you make them look less like issue boxes? |
| 22:47 | <zcorpan__> | Hixie: e.g. be styled the same way as other "this section is non-normative" paragraphs; possibly have .impl style |
| 22:47 | zcorpan__ | files a bug |
| 22:50 | <daedb> | "Microdata brings torment and dismay." <- Awesome, I've been wanting to add more torment to my documents :) |
| 22:50 | <othermaciej> | daedb: all you have to do is add the @torment attribute - the rest is automatic |
| 22:51 | <TabAtkins> | @itemtorment, actually. |
| 22:51 | <TabAtkins> | Only on a top-level microdata item. |
| 22:51 | <Dashiva> | daedb: Where's this? |
| 22:51 | <daedb> | Dashiva: http://lists.w3.org/Archives/Public/public-html/2010Feb/0865.html |
| 22:51 | <TabAtkins> | (You can specify the exact level of dismay in @itemdismay on @itemprop elements.) |
| 23:24 | <AryehGregor> | I like the commit message here: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62903 |
| 23:32 | <zcorpan__> | btw i'm going to write an article about html5 video, intending to cover the whole dom api, all events, and everything |
| 23:33 | <zcorpan__> | advice on how to structure it, or specific questions i should cover (like how do i do this, how do i do that) are appreciated |
| 23:34 | <zcorpan__> | write it here on #whatwg (i'll read the logs) or send to simonp⊙oc |
| 23:34 | <TabAtkins> | zcorpan__: How do I use <video> and have it play in all the browsers? |
| 23:35 | <zcorpan__> | thanks, keep em coming :) i'll go to bed now; nn |
| 23:50 | <AryehGregor> | * May harm the reputation of people who designed longdesc="" in the |
| 23:50 | <AryehGregor> | first place. |
| 23:50 | <AryehGregor> | Hixie, that kind of thing really isn't productive. |
