| 01:47 | <Hixie> | Lachy: yt? |
| 05:16 | <Hixie> | i wonder what to do about the definition of "browsing context" |
| 05:16 | <Hixie> | (in the content of an author version of the spec) |
| 05:49 | <MikeSmith> | Hixie: define by example? |
| 05:50 | <Hixie> | well i was wondering whether i need a definition at all |
| 05:50 | <Hixie> | maybe i can just refer to frames and windows in the author text and leave all the complexities out |
| 05:51 | <MikeSmith> | Hixie: It does seem like there needs to be some definition or explanation of that term in the author view |
| 05:52 | <MikeSmith> | and seems necessary to leave out the complexities because otherwise you'd get into also needing to define a lot of other terms too |
| 05:52 | <MikeSmith> | like pulling a thread |
| 05:53 | <Hixie> | yeah |
| 05:55 | <MikeSmith> | but I would think you'd not be able to get away with just referring to frames and windows and not to "browsing context" at all |
| 05:55 | <Hixie> | oh well, i'll figure something out |
| 08:19 | <hsivonen> | yay! the whatwg archives are back! |
| 08:40 | <Lachy> | Hixie, I'm here now |
| 08:45 | <zcorpan> | http://spacergif.mobi/ |
| 08:46 | <zcorpan> | (for now, it just returns a spacer gif) |
| 09:01 | <Lachy> | Hixie, Requiem 1.8.7 is out now. The most recent changes provide support for iTunes 8.1, speed improvements and fixes for >2GB files |
| 11:34 | <annevk5> | http://twitter.com/jsdc/statuses/1335371346 we do? |
| 11:38 | <Lachy> | annevk5, no, we don't. I wonder where that misconception came from |
| 11:40 | <annevk5> | http://spacergif.mobi/ most excellent :) |
| 11:40 | <Lachy> | hmm, that no longer returns a spacer gif for me. All I get is an empty directory listing |
| 11:42 | <zcorpan> | fixed |
| 11:44 | <Lachy> | zcorpan, is that your domain? |
| 11:44 | <zcorpan> | yeah |
| 11:44 | <Lachy> | oh. Why did you get that? |
| 11:44 | <Lachy> | are you going to use it for anything but a spacer gif? |
| 11:44 | <zcorpan> | yeah i'll probably have a blog there |
| 11:44 | <Lachy> | ok |
| 11:45 | Lachy | looks forward to reading about all the different ways of creating and using spacer gifs on a mobile |
| 11:46 | <zcorpan> | now i wonder two things... how do i get genshi installed there and has someone written a blog using genshi already? |
| 11:46 | jgraham | hopes that zcorpan will blog in the style of a teenage girl |
| 11:47 | <zcorpan> | jgraham: sorry to disappoint you |
| 11:47 | <jgraham> | zcorpan: http://www.cmlenz.net/archives/2008/08/view-source#more |
| 11:47 | <jgraham> | Is the only one I can think of |
| 11:48 | <zcorpan> | jgraham: thanks |
| 11:49 | <jgraham> | zcorpan: Ah, shame. It could have been all like, well I got us this morning right? and there was all this email about XML Namespaces, y'know so I was just like you guys are so lame if you don't know what a script inserted xmlns decleration does and I laughed so hard I almost split my frosties on the cat |
| 11:50 | <zcorpan> | lol |
| 11:51 | <jgraham> | zcorpan: I was going to use it but didn't get around to installing couchdb. But I think I might have another go at some point. SO let me know if you use it and if you have any issues |
| 11:52 | <zcorpan> | ok |
| 11:52 | <annevk5> | jgraham, :D you should write a blog |
| 12:12 | jgraham | had forgotten that the aria taskforce plans to do HTML element -> accessibility api mappings |
| 12:20 | <takkaria> | wtf are people spending so much time discussing alternative calendars? |
| 12:20 | <takkaria> | am I missing something obvious? |
| 12:22 | <annevk5> | the fun in it |
| 12:22 | <annevk5> | though thank god I'm not a calendar geek and not wasting my time on this one :) |
| 12:24 | <zcorpan> | http://html5gallery.com/ |
| 12:26 | <rubys> | One of these days, I really need to complete this: http://rails.intertwingly.net/blog/ |
| 12:27 | <rubys> | (and convert planet intertwingly to use html5 features) |
| 12:34 | <Lachy> | that html5gallery site doesn't actually include any links or URLs for the sites it lists |
| 12:39 | <beowulf> | zcorpan: feel free to add http://havetheygoneawayyet.com/ :) |
| 14:20 | <Lachy> | Hixie, dfn.js is missing from http://www.whatwg.org/specs/web-apps/current-work/multipage/dfn.js but the script on the page still attempts to load it from there. |
| 14:58 | <annevk5> | http://twitter.com/diveintomark/status/1336512675 |
| 15:04 | <hsivonen> | the <time> email just keeps coming... |
| 15:05 | <Lachy> | yeah, it's becomming a permathread |
| 15:06 | karlcow | yawns |
| 15:07 | <Lachy> | I wonder who that tweet from diveintomark is responding to |
| 15:07 | <Lachy> | LOL http://twitter.com/diveintomark/status/1329398270 |
| 15:09 | karlcow | yawns bis |
| 15:09 | <jgraham> | Is any of the time email intersting? I keep dipping in and out and there seems to be a lot of discussion about year 0 and alternate calendar systems |
| 15:10 | jgraham | wonders how one can yawn bis. Or hat bis means. I thought they were some sort of pop band... |
| 15:10 | <jgraham> | s/hat/what/ |
| 15:10 | <annevk5> | bis can also mean second or some such |
| 15:10 | <annevk5> | e.g. house number 12 and 12bis |
| 15:11 | <annevk5> | numbers* meh |
| 15:13 | <Lachy> | jgraham, I don't think so. It's all repeating the same thing over and over again |
| 15:13 | <jgraham> | It also stands for Bandalag íslenskra skáta, the icelandic boy scouts. But I doubt karlcow yawned so wide he swallowed a whole nation's scouts by mistake |
| 15:15 | <Lachy> | it gets a little weird when you start to realise the contradictory arguments that seem to be getting put forth, like wanting to allow negative years, a year 0000, and have the year -0001 mean 1BCE instead of 2BCE |
| 15:15 | <jgraham> | It's only confusing if the contradictory positions are all taken by a single person at a single time |
| 15:16 | <Lachy> | there was one email in which that seemed to be the case |
| 15:16 | Lachy | looks it up |
| 15:16 | <Lachy> | http://lists.w3.org/Archives/Public/public-html/2009Mar/0341.html |
| 15:17 | <Philip`> | What if I want to represent dates in Flatland's calendar system, where there is a year 0? |
| 15:18 | <Lachy> | "I think [year 0000] should be allowed. Historians deny the existence of year 0, but astronomers use it... ISO 8601:2000 and above suggest that year 0000 be used and be considered year 1 BC, and then -0001 is 2 BC, etc. Most, I believe, will want year -0004 = 4BC (and this is what I'd suggest)." |
| 15:19 | <jgraham> | Funness |
| 15:20 | <Lachy> | Philip`, Flatland usually refers to an imaginary universe in which there are only 2 spatial dimensions. I'm not sure what that universe has to do with a year 0 |
| 15:20 | <karlcow> | yep bis aka a second time to which people could have replied "Bis repetita non placent" |
| 15:21 | <Lachy> | karlcow, google translate says that translates to "Repeat not place". What on earth does that mean? |
| 15:22 | <jgraham> | Lachy: But imagine I am writing a book about flatland and I want to say it was <time datetime=00-00-00>the first day</time>..." |
| 15:23 | <karlcow> | http://www.google.com/search?q="Bis+repetita+non+placent"; |
| 15:23 | <Philip`> | A Square says, "It was the last day of the 1999th year of our era. The pattering of the rain had long ago announced nightfall; and I was sitting in the company of my wife, musing on the events of the past and the prospects of the coming year, the coming century, the coming Millennium." |
| 15:24 | <Philip`> | and since the new millennium starts on the year 2000, their calendar system must have started with the year 0 |
| 15:25 | <jgraham> | karlcow: Limiting communications to latin should at least cut down on the noise in the channel |
| 15:25 | <karlcow> | http://yauba.com/?q=%22Bis+repetita+non+placent%22 or by the new one ;) |
| 15:25 | <Philip`> | "Acting, as was their wont, in strict accordance with precedent, the highest Circles of the realm were meeting in solemn conclave, as they had met on the first hour of the first day of the year 1000, and also on the first hour of the first day of the year 0." |
| 15:25 | <karlcow> | jgraham: indeed :) |
| 15:25 | <Philip`> | Ah, there's an explicit year 0 |
| 15:26 | <Lachy> | Philip`, what are you quoting from? |
| 15:26 | <Philip`> | Lachy: Flatland |
| 15:27 | <karlcow> | # Bis repetita non placent -- "Repetitions are not well-received." (Horace, Ars Poetica 365) |
| 15:27 | <Philip`> | (http://www.geom.uiuc.edu/~banchoff/Flatland/ etc, though that's not a nicely formatted version) |
| 15:27 | <Lachy> | I didn't know that was a book |
| 15:28 | <Philip`> | http://en.wikipedia.org/wiki/Flatland |
| 15:28 | <karlcow> | Lachy: be happy, being young means that there is a lot of classics to discover. |
| 15:28 | <Lachy> | I'd only ever heard of the place being described in quantum physics documentaries as a way of explaining the concept of 11 demensions to average people. |
| 15:29 | Philip` | wonders how Lachy could have gone through life without knowing of the Flatland book |
| 15:29 | <Lachy> | I don't read much |
| 15:31 | <Lachy> | anyway, I think marking up the calendar systems of fictional universes can be considered an edge case that we don't need to support in HTML5 |
| 15:32 | <hsivonen> | Lachy: next you are going to say we don't need Klingon in Unicode! |
| 15:33 | <Philip`> | It's kind of a weird book, since it's half a discussion of dimensional geometry and half a social commentary on Victorian life |
| 15:33 | <Philip`> | hsivonen: But Unicode is all about edge cases |
| 15:35 | <Lachy> | hsivonen, no, that's a different issue, because writing about fictional universes with different calendar systems doesn't gain anything by having explicit <time> element support, nor lose anything by not having it. |
| 15:37 | <Lachy> | however, it is questionable whether Klingon should be added to Unicode or not. There are several fictional writing systems that I can think of which are not supported in Unicode, such as those from the Lord of the Rings, or the Ancients' alphabet from Stargate |
| 15:38 | <jgraham> | Kligon was rejected from unicode |
| 15:38 | <jgraham> | *Klingon |
| 15:40 | <Philip`> | There are practical reasons to include artificial languages in Unicode - people want to use normal software (text editors, copy-and-paste, TTF fonts, HTML, etc) to manipulate text in those languages, and that software is all designed around being able to represent any text using Unicode |
| 15:41 | hsivonen | proceeds to remove table tainting from the V.nu parser |
| 15:41 | <Philip`> | (though Private Use Areas are probably a sufficient level of inclusion) |
| 15:44 | <Lachy> | for such characters to be useful for such purposes, the code points used within the Private Use Areas would need to have some level of agreement between the applications involved |
| 15:45 | <Lachy> | at the very least, font vendors making different fonts for those characters would need to agree on the code points |
| 15:48 | <Philip`> | Indeed, so you have some simple low-cost external registry to assign character ranges to anyone who wants them |
| 15:50 | <hsivonen> | is it OK if I update html5lib tests to the new AAA tomorrow? what about getting rid of taint and doing foster-parenting the WebKit way? |
| 15:53 | <hsivonen> | I should also update the tests to deal with the new frameset-ok stuff |
| 15:54 | <hsivonen> | good thing I didn't change the tests for spec rev 2730 |
| 16:03 | <gsnedders> | annevk5: ping |
| 16:03 | <annevk5> | pang |
| 16:03 | <gsnedders> | annevk5: What's the state of setting Authorization via XHR? |
| 16:04 | <annevk5> | forbidden I think |
| 16:04 | <gsnedders> | annevk5: I found an old email from 10 months ago saying you'd disallowed it, but it isn't in the latest TR |
| 16:04 | <gsnedders> | Actually, TR version is too old |
| 16:04 | <annevk5> | maybe because TR is from more than 10 months ago? |
| 16:06 | gsnedders | is speaking to people about the possibility of using it for OAuth, which needs Authorization |
| 16:09 | <annevk5> | per XHR Authorization is UA controlled |
| 16:10 | <annevk5> | even though we might be able to change that for same-origin requests it would not be possible for cross-origin so we might as well keep them in sync |
| 16:31 | <annevk5> | does anyone know if SWSX or whatever it's called has live transcripts of sessions? |
| 16:31 | <annevk5> | e.g. the Browser War Panel might be fun to follow |
| 16:38 | beowulf | wonders if browsers will be put in harms way during this panel |
| 17:40 | <Lachy> | "... Rob [Burns] really has a lot to say and I'm looking forward to read more of his ideas and opinions" -- http://lists.w3.org/Archives/Public/public-html/2009Mar/0346.html |
| 17:45 | Philip` | agrees with half of that sentence |
| 18:46 | <sayrer> | jgraham, Philip, mozilla implementors have brought up the fact that HTML5 doesn't define the SQL syntax |
| 18:46 | <sayrer> | rubys, my use of brendan's quote was meant to be considered in relation to your desire to bring in more "content creators" |
| 18:47 | <annevk5> | yeah, SQL syntax is an issue |
| 18:47 | <sayrer> | I am not sure what you meant by that. People from big websites? They always have lots of feature requests |
| 18:48 | <sayrer> | annevk5, yes it fails a number of the design principles as currently specified |
| 18:50 | <annevk5> | and basic LC exit criteria |
| 18:50 | <annevk5> | I don't think anyone is contesting that :) |
| 18:51 | <Hixie> | indeed not |
| 18:51 | <sayrer> | I am actually kinda bummed about that |
| 18:52 | <sayrer> | what a pain, and everyone is using sqlite anyway |
| 18:52 | <annevk5> | afaik it was announced when we added this feature that speccing the dialect would be done once we had some impl experience |
| 18:53 | <sayrer> | that is sort of like specifying a barn door after the horse has escaped |
| 18:53 | <sayrer> | but oh well |
| 18:54 | <sayrer> | I was also wondering what this big effort Hixie keeps giving percentage updates on is |
| 18:55 | <sayrer> | some sort of automated spec split? |
| 18:55 | <Hixie> | going through marking up all the non-author stuff so it can be hidden by a style sheet |
| 18:55 | <Hixie> | if you load the spec in firefox switch to one of the other stylesheets to see the effect |
| 18:55 | <Hixie> | re sql, i think it would have been a mistake to spec the language before having impl experience |
| 18:55 | <sayrer> | so you get something akin to Mike Smith's doc by CSS switching? |
| 18:56 | <Hixie> | looks more like a cut-down version of the html5 spec, but sort of, yes |
| 18:56 | <sayrer> | are there any compatible HTML5 SQL impls? |
| 18:56 | <sayrer> | I thought Gears didn't match WebKit |
| 18:56 | <sayrer> | does Opera match WebKit? |
| 18:57 | <annevk5> | we haven't implemented the database API yet |
| 18:57 | <sayrer> | I guess the Gears one was never claimed to be compatible |
| 18:57 | <annevk5> | though when we will I expect us to use SQLite as well |
| 18:57 | <Hixie> | the gears one and the webkit one have differences, but are more similar than different |
| 18:57 | <Hixie> | much like implementations of everything else :-) |
| 19:25 | <Lachy> | hey Hixie, you pinged me last night? |
| 19:25 | <Hixie> | was going to ask you about requiem |
| 19:26 | <Lachy> | ok |
| 19:26 | <Lachy> | I sent you the latest version |
| 19:26 | <Hixie> | sweet |
| 19:26 | <Hixie> | thanks dude |
| 19:34 | Hixie | discovers that the split-window concept breaks the multiple views concept |
| 19:34 | <Hixie> | what should UIEvent.view return? |
| 19:34 | <Hixie> | shepazu: this may affect DOM3 Events |
| 19:50 | <annevk5> | we could of course do away with multiple views and have a concept of primary view which is the one exposed to scripts... |
| 20:07 | virtuelv | innocently wonders what split-window concept is |
| 20:09 | <virtuelv> | (In other words, where do I look for a description?) |
| 20:13 | <gsnedders> | virtuelv: /dev/null |
| 20:13 | gsnedders | ducks |
| 20:13 | <virtuelv> | :P |
| 20:14 | <gsnedders> | /dev/random should have it somewhere, actually |
| 20:14 | <virtuelv> | gsnedders: generally, ducking is not adviced |
| 20:14 | <jcranmer> | /dev/urandom is faster |
| 20:14 | <jcranmer> | now, can someone get this purr-bucket off my lap? |
| 20:15 | <virtuelv> | gsnedders: if it's a real-world fight, that is, because it'll end with your opponent kicking your face instead of your chest or nuts) |
| 20:15 | <gsnedders> | See, this is why I don't have a face. |
| 20:15 | gsnedders | takes the purr-bucket and gives it to Hixie |
| 20:16 | <annevk5> | virtuelv, HTML5 defines it |
| 20:16 | <annevk5> | virtuelv, look for e.g. WindowProxy |
| 20:16 | <gsnedders> | Silly people giving useful answers… |
| 20:17 | <virtuelv> | "As mentioned earlier, each browsing context has a WindowProxy object. This object is unusual in that it must proxy all operations to the Window object of the browsing context's active document. It is thus indistinguishable from that Window object in every way, except that it is not equal to it." |
| 20:22 | <gsnedders> | awww… there's no such thing as \subsubsubsection |
| 20:23 | <atw> | It looks like the webworker spec implies that MessagePorts implement EventTarget (so you can call addEventListener() on them) but I don't actually see that explicitly stated anywhere. I'm curious, because setting the onmessage attribute effectively calls start() on the port, which perhaps implies that calling addEventListener() does too? |
| 20:23 | <virtuelv> | reading on from there. Why is Window exposing browser chrome constructs, such as the location bar |
| 20:23 | <virtuelv> | knowing that there is no location bar only lets me create more convincing phishing UI |
| 20:24 | <virtuelv> | (Besides, there are contexts wherein application UI constructs are completely irrelevant |
| 20:24 | <virtuelv> | mobile, widgets and embedded cases |
| 20:25 | <sayrer> | http://intertwingly.net/blog/2009/03/16/Yahoo-and-HTML-Rip-Mix-Burn |
| 20:25 | <virtuelv> | chiefly usable in the XUL case, no? |
| 20:26 | <virtuelv> | in which case I wonder: Why is it in HTML5? |
| 20:27 | <annevk5> | atw, "Objects implementing the MessagePort interface must also implement the EventTarget interface." |
| 20:27 | <annevk5> | atw, -- http://www.whatwg.org/specs/web-apps/current-work/#messageport |
| 20:40 | <atw> | anne - sigh, there it is right under the interface description. I must be blind. |
| 20:41 | <atw> | Is the implication that adding an event listener for (say) the "message" event is equivalent to setting an onmessage() handler (meaning it triggers side effects like starting the message port)? |
| 20:41 | <annevk5> | atw, long term it should probably be part of the interface, but Web IDL isn't quite there yet |
| 20:41 | <Hixie> | atw: no |
| 20:41 | <Hixie> | atw: there's no implication |
| 20:41 | <Hixie> | atw: the spec says exactly what it means and no less :-) |
| 20:42 | annevk5 | wonders why there's magic for onmessage |
| 20:43 | <Hixie> | so that you don't have to call .start() in the simple case |
| 20:43 | <roc> | does anyone know what Crockford wants to do in his "HTML 4.2", in concrete terms? |
| 20:43 | <atw> | Yeah, it does seem kind of magical for onmessage() to have side effects but nothing else. I'm playing with the WebKit implementation now to see if they've implemented that side effect, as I'm not getting it to work. |
| 20:44 | <Hixie> | roc: http://www.crockford.com/html/ |
| 20:44 | <sayrer> | roc, I think he mainly wants something similar to seamless iframes, not many other new features, and possibly some breaking changes |
| 20:44 | <Hixie> | roc: q.v. also http://diveintomark.org/archives/2008/02/21/the-bolero-of-troll |
| 20:45 | <atw> | Hixie: got it. We can argue why addEventListener() isn't part of the simple case, but the motivation for the side-effect seems clear. |
| 20:45 | <Hixie> | atw: well you're only going to use aEL() if you're hooking multiple listeners up |
| 20:45 | <Hixie> | atw: in which case you might be doing asynchronously |
| 20:45 | <Hixie> | atw: so we don't want to start the message queue in that case |
| 20:45 | <Hixie> | atw: or you might miss messages |
| 20:46 | <roc> | ah OK, so he wants a completely different language with a version switch |
| 20:46 | <sayrer> | I dunno, I think his experience with ES3.1 will make it less completely new |
| 20:47 | <roc> | are we sure that's still his plan? that page is quite old |
| 20:47 | <sayrer> | yeah, my point |
| 20:47 | <roc> | ok |
| 20:47 | <Hixie> | roc: no idea |
| 20:47 | <sayrer> | oh, he'll want postMessage too |
| 20:50 | <atw2> | hixie: thx, makes sense |
| 20:53 | <Hixie> | roc: if he has other concrete ideas, he hasn't said them |
| 20:54 | <Hixie> | I wish he would |
| 20:59 | <sayrer> | Hixie, I've seen videos where he touches on it |
| 21:00 | <sayrer> | and I expect he'll raise it at the ECMA/W3C joint meeting later this year |
| 21:00 | <Hixie> | would be nice if he actually told us directly |
| 21:00 | <sayrer> | that is the entire purpose of the co-location, afaik |
| 21:01 | <Hixie> | are the numerous mailing lists and irc channels not open enough for him? |
| 21:01 | <annevk5> | he's the entire reason for co-location? wow |
| 21:01 | <sayrer> | he asked for it, iirc |
| 21:01 | <sayrer> | the ECMA committee is only 12 people in a room usually, and 4 of them seem to never come back |
| 21:02 | <annevk5> | true |
| 21:03 | <Hixie> | oh maybe the html5 work is in fact _too_ open for him |
| 21:07 | <sayrer> | Hixie, well, in my experience he isn't too interested in open goal setting, just open bug fixes |
| 21:09 | <Hixie> | sayrer: well he's not reported any bugs in our numerous open bug reporting mechanisms either |
| 21:10 | <sayrer> | Hixie, I didn't mean in that direction |
| 21:10 | <sayrer> | you fix his bugs, see :) |
| 21:11 | <Hixie> | while it does look like one day i might end up fixing the mess he made with JSON, i'm assuming that's not what you meant :-) |
| 21:11 | <gsnedders> | But JSON is pefect! |
| 21:11 | jgraham | doesn't quite understand why his opinion has such disproportionate weight |
| 21:11 | <jgraham> | Or at least why people act as if it did |
| 21:12 | <Hixie> | 64%! |
| 21:12 | <Hixie> | jgraham: dunno, but if you find out please let me know, it'd be a nice trick to reproduce! |
| 21:13 | annevk5 | finds an email in the <time> thread where the three people still debating have declared agreement |
| 21:14 | <Hixie> | speaking of the <time> thread |
| 21:14 | <sayrer> | Hixie, what do you want to change? |
| 21:14 | <sayrer> | about JSON, I mean |
| 21:14 | <Hixie> | is there an e-mail somewhere in that beast that describes use cases? |
| 21:14 | <Hixie> | sayrer: error handling, adding comments, adding NaN and Infinity |
| 21:15 | <sayrer> | error handling, hmm |
| 21:15 | <sayrer> | do you want to add error handling to every format? |
| 21:15 | <Hixie> | me personally? no |
| 21:16 | <sayrer> | do you want error handling added? |
| 21:16 | <Hixie> | formats that expect to be interoperably implemented amongst a wide number of UAs and used amongst a very wide array of producers and consumers all have error handling rules, whether they are de-facto or de-jure |
| 21:17 | <Hixie> | by having de-jure rules (i.e. by speccing them) is a huge help in getting interoperability sooner. it doesn't matter whether those rules are draconian like XML, or ignore-unknown like CSS, or crazy-complicated like HTML, or whatever. |
| 21:17 | <Hixie> | s/by// |
| 21:18 | <sayrer> | JSON in ES3.1 has draconian rules, then |
| 21:18 | <Hixie> | excellent |
| 21:18 | <sayrer> | anything that doesn't match the grammar is supposed to raise a syntax error |
| 21:18 | <sayrer> | we'll see how long that lasts |
| 21:18 | <gsnedders> | But that doesn't happen now. |
| 21:19 | <Hixie> | does JSON in ES3.1 replace the json.org JSON? |
| 21:19 | <sayrer> | the json.org JSON keeps changing to match ES3.1 |
| 21:19 | <sayrer> | ES3.1 allows more than the RFC does |
| 21:19 | <roc> | there's the problem that the market rewards violations of draconian-ness, but we can try to hold the line |
| 21:20 | <Dashiva> | Is there still that silly duality of / being allowed both as / and as \/? |
| 21:21 | <sayrer> | I asked him why Infinity and NaN weren't allowed |
| 21:21 | <sayrer> | he said he didn't want to tie it to IEEE floating point |
| 21:23 | <roc> | is there a rational reason not to, given JS and most other languages depend them? |
| 21:24 | <sayrer> | not that I know of |
| 21:24 | <sayrer> | the JSON specified by the RFC allows you to support whatever extensions you want, btw |
| 21:25 | <Hixie> | lovely |
| 21:25 | <sayrer> | the main change in ES3.1 is to allow primitive root values |
| 21:25 | Hixie | strikes "error handling" from his JSON fixit list and adds "utter mess caused by allowing arbitrary extensions" |
| 21:26 | <Dashiva> | Just require namespaces for the extensions |
| 21:26 | <sayrer> | see, JSON really is replacing XML! |
| 21:59 | <Hixie> | 67% |
| 22:11 | <Philip`> | I like how example 1 on http://developer.search.yahoo.com/help/objects/games is giving bogus metadata |
| 22:11 | <Philip`> | (It says it's talking about resource="http://example.com/video_object.swf"; when actually it's a game_object.swf instead) |
| 22:11 | <Hixie> | 68% |
| 22:17 | <Hixie> | i don't really understand what the resource="" URL is for on that example |
| 22:17 | <Hixie> | couldn't it just be anything? |
| 22:20 | <annevk5> | sigh, the markup they recommend will not exactly do as shown in text/html |
| 22:20 | <annevk5> | (apart from the xmlns clutter, that is) |
| 22:21 | <Philip`> | Hixie: The resource is to tell the RDFa processor what object the <a rel href>s are properties of |
| 22:22 | <Hixie> | so how does yahoo know which URLs to examine the properties of? |
| 22:22 | <Hixie> | is there any way to test this in anything approaching real time? |
| 22:23 | <Philip`> | It looks like the idea is that when a page comes up in the search results, they look at its RDF triples and if there's one saying "the media:thumbnail of foo is bar" then they'll show the little bar picture and make it link to foo |
| 22:23 | <Hixie> | so then it doesn't matter what the value is, right? |
| 22:24 | <Hixie> | oh wait they make it link to foo? |
| 22:24 | <Hixie> | not to the page? |
| 22:24 | <Philip`> | Oh, but that's probably not all since it's got <object rel=media:game> too, which I guess must do something |
| 22:24 | <Hixie> | sigh, authors are never going to make heads or tails of this |
| 22:24 | <Philip`> | They can copy and paste the examples |
| 22:24 | <Hixie> | the examples are wrong, as you just pointed out! |
| 22:25 | <Philip`> | and we'll have a thousand pages all making assertions about http://example.com/video_object.swf |
| 22:26 | <Philip`> | Hixie: I assume the "Test Markup" button on that page will let you test things in real time |
| 22:28 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F368 |
| 22:29 | <Philip`> | Works fine with tag soup too |
| 22:30 | <Hixie> | if you change <object> to <div> it says that one of the required attributes is missing |
| 22:30 | <Hixie> | ... |
| 22:30 | <Hixie> | so it's not rdfa |
| 22:31 | <Philip`> | Really? Works fine for me with <div> |
| 22:31 | <Hixie> | hm, yes, indeed |
| 22:31 | <Hixie> | weird |
| 22:31 | <Hixie> | compare |
| 22:31 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F369 |
| 22:31 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F370 |
| 22:32 | <Hixie> | they parse the URLs incorrectly, too |
| 22:32 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F372 |
| 22:34 | <Philip`> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Frdfa%2Fmisc01.html |
| 22:35 | <Hixie> | wow, you killed it |
| 22:35 | <Philip`> | (http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Frdfa%2Fmisc00.html does work) |
| 22:35 | <Philip`> | Gosh, I never meant to kill it |
| 22:35 | <Hixie> | dude they totally don't actually support xmlns |
| 22:35 | <Hixie> | they use magic prefixes! |
| 22:35 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F375 |
| 22:35 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fjunkyard.damowmow.com%2F376 |
| 22:36 | <Hixie> | they do check the URL in the xmlns:media="" attribute though |
| 22:38 | <Hixie> | they don't support CURIEs either |
| 22:38 | <Hixie> | they just have magic prefixes and they check that they are declared to the expected value |
| 22:39 | <rubys> | The HTML 4.2 page is dated March 7, 2009. Doesn't seem that old to me. |
| 22:40 | <Philip`> | http://www.w3.org/RDF/Validator/ARPServlet?URI=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Frdfa%2Fmisc02.html&PARSE=Parse+URI%3A+&TRIPLES_AND_GRAPH=PRINT_TRIPLES&FORMAT=PNG_EMBED |
| 22:41 | <Hixie> | Philip`: nice |
| 22:41 | <Hixie> | Philip`: guess you should report it |
| 22:43 | <Philip`> | Hmm, they have a nice problem-reporting form but it just uses mailto: and so it doesn't work |
| 22:44 | <annevk5> | rubys, http://www.crockford.com/html/ is relatively old, the 4.2 page doesn't amount to anything last I checked... |
| 22:47 | <roc> | the "HTML 4.2" blog post just says that he will simplify, streamline and generalize |
| 22:47 | Philip` | sends a manual email instead |
| 22:47 | <Hixie> | Philip`: hahaha, they support these attributes on any namespace, including the null namespace: http://developer.search.yahoo.com/help/objectfinder?url=hixie.ch%2Fwww%2Ftests%2Fadhoc%2Frdfa%2F001.xml |
| 22:47 | <roc> | which is obviously something no-one can disagree with |
| 22:48 | <roc> | I'd like to see how he actually plans to do that without breaking the Web |
| 22:48 | <Hixie> | you can simplify too much, and you can generalise too much |
| 22:48 | <Hixie> | i have no idea what "streamlining" means in this context |
| 22:48 | <gsnedders> | where is the ES 3.1 draft? |
| 22:48 | Philip` | notes that http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2009Mar/0096.html is where this was mentioned |
| 22:48 | <roc> | and without leaving behaviour unspecified |
| 22:49 | <rubys> | I'm sure that his intentions are good, not evil. <http://www.json.org/license.html>; |
| 22:50 | <Hixie> | Philip`: oh hey, i totally misunderstood what that rdfa was doing |
| 22:50 | <Hixie> | Philip`: the <a>s nested in the <object> are misleading |
| 22:50 | <Hixie> | Philip`: they're all using the implied subject of the whole page |
| 22:51 | <Philip`> | Hixie: Do you just mean they support the RDFa attributes on elements in any namespace? (which doesn't seem inherently unreasonable) |
| 22:51 | <Hixie> | yes |
| 22:51 | <Hixie> | that's what i meant above up to the point where i posted the 001 link |
| 22:51 | <Hixie> | haha this "rdfa" support is completely bogus |
| 22:52 | <Philip`> | Hixie: The <a>s are using the subject defined by the nearest @resource, I think |
| 22:52 | <Hixie> | nope, i can get the same result without any nesting |
| 22:52 | <Philip`> | and they themselves define the predicate and object that goes with that subject |
| 22:54 | <Hixie> | good lord almighty rdfa is really feaking confusing |
| 22:54 | <rubys> | blasphemer! <ducks> |
| 22:54 | <annevk5> | your English is too :p |
| 22:56 | <Philip`> | http://developer.search.yahoo.com/help/objectfinder?url=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Frdfa%2Fmisc03.html |
| 22:56 | <Philip`> | The last one looks wrong in that view, but correct in the raw RDF that it passes to the validator |
| 22:58 | <Hixie> | can you confirm that http://hixie.ch/tests/adhoc/rdfa/003.xml and http://hixie.ch/tests/adhoc/rdfa/004.xml are equivalent in RDFa? |
| 22:58 | <Hixie> | i'm having issues understanding it |
| 22:58 | <Hixie> | it = the spec |
| 22:59 | <gsnedders> | RDFa is pointless |
| 22:59 | gsnedders | hides |
| 23:01 | <Philip`> | Hixie: It looks to me like they ought to be the same :-) |
| 23:01 | <Hixie> | they're not in the yahoo thing |
| 23:02 | gsnedders | writes to Edinburgh to defer his application for 2010 entry |
| 23:02 | <Philip`> | http://torrez.us/rdfa/ says they're the same |
| 23:02 | <Hixie> | seriously though, rdfa is really confusing |
| 23:02 | <annevk5> | Is Y! using a proper parser though? I thought the expectation was that people wouldn't use such a thing... |
| 23:02 | <Hixie> | when does one use rel and when does one use property? |
| 23:03 | <Hixie> | content or resource or href? |
| 23:03 | <Philip`> | ("same" in the sense that they parse into an equivalent RDF graph) |
| 23:03 | <Philip`> | annevk5: What do you mean by "proper"? |
| 23:03 | <Philip`> | annevk5: It's totally tag soup and not at all XML, and is case-insensitive etc |
| 23:04 | <annevk5> | Philip`, I meant in the RDF sense, but that's interesting too :D |
| 23:04 | <Hixie> | i was triggering their xml parser with my most recent tests |
| 23:04 | <Hixie> | illformed content caused it to fail |
| 23:04 | <Philip`> | annevk5: but I assumed the RDFa people's idea is that you extract data from the DOM in the same way (modulo real vs fake namespace declarations etc) |
| 23:04 | <Hixie> | the yahoo thing is definitely not following rdfa rules correctly |
| 23:05 | <Philip`> | annevk5: and I thought that was what existing tools tended to do, e.g. using Tidy to convert stuff into XML before parsing it |
| 23:05 | <annevk5> | Philip`, which is why I meant in the RDF sense |
| 23:05 | <Hixie> | actually i lie |
| 23:05 | <Hixie> | they are getting the same triples out |
| 23:05 | <Hixie> | they're just reading their rdf tree wrongly |
| 23:07 | <Philip`> | s/tree/graph/ |
| 23:11 | <Hixie> | http://developer.search.yahoo.com/help/objectfinder?url=hixie.ch%2Fwww%2Ftests%2Fadhoc%2Frdfa%2F002.xml |
| 23:11 | <Hixie> | i made the button disappear! |
| 23:12 | <Philip`> | Can you use this to sabotage Yahoo search result pages? |
| 23:13 | <Hixie> | i don't understand http://torrez.us/services/rdfa/?url=http%3A%2F%2Fhixie.ch%2Fwww%2Ftests%2Fadhoc%2Frdfa%2F002.xml |
| 23:13 | <Hixie> | am i misunderstanding curies? |
| 23:14 | <Philip`> | What bit don't you understand? |
| 23:14 | <Hixie> | where did my namespaces go? |
| 23:14 | <Hixie> | http://hixie.ch/www/tests/adhoc/rdfa/002.xml |
| 23:14 | <Philip`> | It doesn't matter where they went |
| 23:14 | <Hixie> | in particular test3 |
| 23:15 | <Hixie> | i mean the namespace values, not the prefixes |
| 23:15 | <Philip`> | Oh |
| 23:15 | <Philip`> | Hmm... |
| 23:17 | <Philip`> | You have to use Safe CURIEs |
| 23:17 | <Philip`> | rel="[med:ia/game]" |
| 23:18 | <Hixie> | ...why? |
| 23:19 | <Hixie> | 5.4.4. Use of CURIEs in Specific Attributes says rel="" takes regular curies |
| 23:19 | <Hixie> | "@rel and @rev do not differentiate their two types of data by using [safe CURIE]s. Instead, any value that matches an entry in the list of link types in the section The rel attribute, MUST be treated as if it was a URI within the XHTML vocabulary, and all other values must be CURIEs." |
| 23:19 | <Philip`> | Oops, I was reading the wrong bit |
| 23:20 | <Hixie> | this is a seriously confusing technology |
| 23:20 | <Philip`> | but if you use rel="[med:ia/game]" in http://torrez.us/rdfa/ then it works like you want :-) |
| 23:20 | <Hixie> | so http://torrez.us/rdfa/ is buggy? |
| 23:21 | <Philip`> | Seems so |
| 23:22 | <Philip`> | Since the rel token isn't a known keyword, it should be processed a CURIE with prefix "[med" and name "ia/game]", I think |
| 23:23 | <Philip`> | though I don't see anything that defines what should happen when the prefix isn't in scope |
| 23:23 | Philip` | guesses Hixie is not being convinced of the value and quality of RDFa |
| 23:23 | <Hixie> | oh i'm being convinced... |
| 23:25 | <Lachy> | Hixie, what? Are you seriously almost convinved to add RDFa to HTML5? |
| 23:26 | <Philip`> | HTML is way too easy to understand already |
| 23:26 | <Lachy> | LOL |
| 23:26 | <Philip`> | We need to do something about that |
| 23:26 | <Hixie> | i think it's pretty clear that there are use cases that need to be addressed |
| 23:26 | <Lachy> | possibly. But does RDFa really address those cases? |
| 23:27 | <Hixie> | hard to tell, given how mind-bendingly confusing it is |
| 23:29 | <Lachy> | I'm not sure what the use cases are exactly, since I haven't followed most of the RDF threads. But I don't like the idea of introducing RDFa mostly due to its overall complexity |
| 23:33 | <Philip`> | http://www.w3.org/2007/08/pyRdfa/extract?uri=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Frdfa%2Fmisc01.html&format=pretty-xml&warnings=false&parser=lax&host=xhtml&space-preserve=true&submit=Go%21 |