| 01:45 | <Hixie> | ruby1: sam? |
| 01:46 | <ruby1> | yes, can't seem to get rubys at the moment |
| 01:46 | <Hixie> | hey dude |
| 01:47 | <Hixie> | i was wondering if you were aware that DanC was using your objections on <canvas> as a reason not to publish html5 as a working draft: |
| 01:47 | <Hixie> | http://lists.w3.org/Archives/Public/public-html/2007Nov/0416.html |
| 01:48 | <Hixie> | ...and whether, if that wasn't intentional, you could let the htmlwg know you didn't mean to imply that we shouldn't publish :-) |
| 01:48 | <othermaciej> | hey ruby1 |
| 01:48 | <ruby1> | I'm torn. I do believe in release early and often. And I'm also keenly aware that there are scope issues that need to be resolved sooner than later. |
| 01:49 | <ruby1> | maciej: why "other"? |
| 01:49 | <othermaciej> | ruby1: because someone else has "maciej" registered on this server |
| 01:50 | <rubys> | was able to "ghost" rubys. Much better. |
| 01:51 | <rubys> | I'll post a response to Dan on the subject tomorrow. |
| 01:51 | <othermaciej> | rubys: much appreciated |
| 01:52 | <Hixie> | rubys: cool, thanks |
| 01:52 | <Hixie> | rubys: i would be interested in hearing why you weren't convinced by the arguments that <canvas> is within the scope of html5 according to the charter, btw (e.g. those maciej wrote in: http://www.w3.org/2002/09/wbs/40318/tactics-gapi-canvas/results ) |
| 01:54 | <othermaciej> | oh yeah, I would too, if you are indeed not convinced |
| 01:58 | <rubys> | I'll provide a more complete answer in my post, but the short version is (1) by that reasoning SVG is in scope, (2) those that are working on the charter don't need to impact the people working on the spec (a point which I realize is partially invalidated by the fact that publishing this draft is on hold), and (3) I think namespaces are reasonable, so I guess that's in too? #3 is the most important issue. Charters aren't important when pe |
| 02:00 | <hober> | cut off at "when pe", btw |
| 02:01 | <rubys> | Charters aren't important when people agree, they are only important when people disagree. |
| 02:02 | <Dashiva> | Sounds like we might need the entire spec in the charter for this wg then :) |
| 02:03 | <hober> | I really doubt your #2 -- I suspect rechartering would be a giant time-sink political mess which would necessarily drag everyone into it |
| 02:04 | <hober> | Agreed on #1 -- I expect <canvas> and SVG to both be included |
| 02:04 | <othermaciej> | rubys: I do think SVG integration is in scope - I hope you do too |
| 02:05 | <othermaciej> | rubys: I also think namespaces are in scope (indeed, some mechanism to embed foreign languages is required, though the requirement is not specifically for namespaces) |
| 02:06 | <rubys> | if we can agree on those things, then they should be mentioned in the charter too. I'm not hopeful that we can agree on those things yet. |
| 02:07 | <othermaciej> | rubys: "The HTML WG is encouraged to provide a mechanism to permit independently developed vocabularies such as Internationalization Tag Set (ITS), Ruby, and RDFa to be mixed into HTML documents. Whether this occurs through the extensibility mechanism of XML, whether it is also allowed in the classic HTML serialization, and whether it uses the DTD and Schema modularization techniques, is for the HTML WG to determine." |
| 02:07 | <othermaciej> | rubys: that pretty clearly puts SVG integration and namespaces in scope |
| 02:07 | <othermaciej> | rubys: (though it does not specifically require either) |
| 02:08 | <rubys> | maciej: where are you quoting from? |
| 02:09 | <othermaciej> | rubys: the HTML Working Group charter |
| 02:09 | <othermaciej> | http://www.w3.org/2007/03/HTML-WG-charter.html |
| 02:09 | <rubys> | oops, I was looking at the wrong page for a moment |
| 02:09 | <othermaciej> | there are many old versions on the Google |
| 02:10 | <rubys> | "whether it is also allowed in the classic HTML serialization, ..., is for the HTML WG to determine." <= that's the part that I'm not hopeful on. |
| 02:10 | <othermaciej> | anyway, I think that charter clause as written is better than an exhaustive and exclusive list of languages to integrate, or a specific requirement to use the namespace syntax |
| 02:11 | <rubys> | that's a low bar. |
| 02:11 | <othermaciej> | since it gives us leeway to study the technical issues |
| 02:11 | <othermaciej> | I think nearly all WG members are in favor of integration of at least some specific languages with some syntax |
| 02:11 | <othermaciej> | I would guess most are in favor of SVG being one of these |
| 02:12 | <othermaciej> | and I think that many would favor or at least not object to an open-ended extensibility mechanism, likely compatible with Namespaces in XML |
| 02:12 | <othermaciej> | I am personally pro all three of those things |
| 02:12 | <rubys> | An exclusive list or prematurely setting on a specific syntax are both non-starters. Agreeing to allow other vocabularies -- some that may be known today, and some that may not -- is something that I don't see agreement on. |
| 02:12 | <othermaciej> | although on the namespace syntax there are many devilish details to work out |
| 02:13 | <othermaciej> | I think there is some disagreement on whether extensibility needs to be open-ended or just support a fixed list of additional vocabularies |
| 02:13 | <othermaciej> | but I do not think the charter process is the best way to come to consensus on that point |
| 02:14 | <othermaciej> | I personally would like to take a stab at proposing an approach to namespaces that can have decent degradation and cross-XML/Classic properties when I have time |
| 02:15 | <rubys> | hopefully more people will comment on the proposal that you come up with than have commented on mine. |
| 02:15 | <Hixie> | it's still on my list of things to look at :-) |
| 02:16 | <othermaciej> | rubys: I did comment on one of yours at some point |
| 02:16 | <othermaciej> | I think we may need a couple of different proposals on the table to compare and contrast |
| 02:17 | <othermaciej> | cause there are tradeoffs here |
| 02:17 | <rubys> | discussion is good. |
| 02:17 | <Hixie> | rubys: (sorry, was getting dinner. i didn't really follow how your comments above (1,2,3) lead to you thinking <canvas> wasn't covered by the points in the charter that maciej lists as being ways in which <canvas> is covered, though, fwiw.) |
| 02:19 | <othermaciej> | anyway, addressing the foreign syntax issue is one of the few significant feature additions I really want to see |
| 02:19 | <othermaciej> | (for whatever my opinion's worth) |
| 02:50 | <MikeSmith> | Philip` - any chance you might be awake right now? |
| 02:52 | MikeSmith | wonders if there might be any canvas interoperability reports around other than Philip`'s |
| 02:52 | <MikeSmith> | http://canvex.lazyilluminati.com/tests/tests/results.html |
| 03:03 | <Hixie> | http://www.michaelsalamon.com/?p=20 is probably worth bearing in mind |
| 03:03 | <Hixie> | i'm sure i'm guilty of such problems in html5 :-/ |
| 03:03 | <Hixie> | (like maybe the inline/block stuff, i dunno) |
| 03:27 | <othermaciej> | Hixie: style attribute :-) |
| 03:32 | <Hixie> | yeah well that's an open issue |
| 03:32 | <Hixie> | i'm still considering having two levels of conformance, one for final production-level code and one for experimentation, one-offs, etc |
| 03:36 | othermaciej | points to the picture |
| 03:39 | <Hixie> | the difference here is that there are real benefits to not using media-specific inline styling |
| 03:57 | <othermaciej> | delete the word inline and I'll agree |
| 03:57 | <othermaciej> | add an explanation of how to distinguish media-specific styling from media-independent styling and I'll agree you have a potentially worthwhile conformance requirement |
| 03:59 | <othermaciej> | bonus points for explaining how this page would be improved by not using the style attribute: http://www.flickr.com/photos/tags/ |
| 04:02 | <othermaciej> | (alternatives I have seen include nesting em tags (ugh) and defining a class for each size and just using a less continuous size scale (also kinda ugh) |
| 04:03 | <othermaciej> | if you add a fine color grade as well, then the class approach becomes untenable |
| 04:25 | <Hixie> | othermaciej: why is nesting tags (not necessarily <em>) bad? |
| 04:25 | <othermaciej> | Hixie: makes the markup fugly, and more painful to generate programatically |
| 04:26 | <othermaciej> | Hixie: also takes more bandwidth |
| 04:26 | <Hixie> | but style="" is better in that respect? |
| 04:26 | <othermaciej> | especially if you go up to, like, 20 levels |
| 04:26 | <othermaciej> | sure, if I want to make a tagcloud on a log/log scale of frequency I can generate an int scaled to the right range and happily slap it into a style attribute |
| 04:27 | <othermaciej> | wrapping in 20 <em>s is more unpleasant both to read and write |
| 04:27 | <othermaciej> | also no good for a color grade |
| 04:27 | <Hixie> | hmm |
| 04:27 | <othermaciej> | unless you write totally insane style rules with one per nesting level |
| 04:27 | <Hixie> | wonder how to address this |
| 04:27 | <Hixie> | it's totally clear to me that style="" is a terrible way of doing this, but i agree that it may be the better way (for a colour grand, at least) |
| 04:29 | <othermaciej> | also if it is nonconforming at the DOM level too (not sure if the HTML5 spec defines conforming post-parse DOMs) then you can't do foo.style.left="4px" or whatever |
| 04:29 | <othermaciej> | since that creates a style attribute if you did not have one |
| 04:41 | <Hixie> | i'm not convinced .style.foo = '' is good style either |
| 04:41 | <Hixie> | but that's another story |
| 04:41 | <Hixie> | i don't expect to win any battles on that front any time soon |
| 05:01 | <markp> | http://www.crockford.com/html/ |
| 05:02 | markp | wanders off mumbling "to the fairest" under his breath |
| 05:28 | <MikeSmith> | interesting that Doug writes "The only character encoding permitted in [in my proposed version of] HTML 5 is UTF-8." but he has <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> in the source for the doc.. |
| 08:03 | <Philip`> | MikeSmith: There wasn't much chance, and I went to bed early anyway |
| 08:04 | <Philip`> | I'm not aware of any other canvas interoperability reports |
| 08:04 | <MikeSmith> | OK |
| 08:05 | <MikeSmith> | I had mistakenly thought your report didn't cover Webkit |
| 08:05 | <MikeSmith> | but Maciej pointed out to me that it did |
| 08:05 | <MikeSmith> | after he had manually run through all the test cases |
| 08:06 | <MikeSmith> | with results that turn out to align pretty much with the results in your report |
| 08:06 | <MikeSmith> | see earlier discussion over on #webkit |
| 08:10 | <Philip`> | Okay - those results are just hidden in the column with heading "...AppleWebKit..." :-) |
| 08:15 | <MikeSmith> | Philip` - yeah, I drank too much breakfast beer this morning before screwing my head back on |
| 08:15 | <MikeSmith> | feeling better now |
| 08:16 | <Philip`> | The test report thing at http://canvex.lazyilluminati.com/tests/reportgenentry.html ought to make it easy to collect results, except it's not really designed to be usable by anyone other than me, so I should probably fix that at some point |
| 08:18 | <MikeSmith> | Philip` - Maciej said "fwiw it's not unlikey we will implement those two features in the not-too-distant future and also go over the tests and fix our implementation or try to get the spec fixed as appropriate" |
| 08:18 | <MikeSmith> | so maybe would be worthwhile for them |
| 08:18 | <MikeSmith> | I dunno |
| 08:18 | <MikeSmith> | anyway, thanks for making that report |
| 08:19 | <othermaciej> | Philip`: how do I use it? |
| 08:19 | <MikeSmith> | it's very useful to have something referenceable to cite |
| 08:19 | <othermaciej> | oh, I see |
| 08:19 | <othermaciej> | Philip`: does more than 150 items per page make Safari sad? |
| 08:19 | <Philip`> | There's a couple of tests that I know are wrong, and haven't got around to re-uploading yet, but otherwise it ought to be about correct |
| 08:20 | <Philip`> | othermaciej: Yes - when I last looked, WebKit had a hardcoded limit of 200 frames per document |
| 08:21 | <othermaciej> | oh |
| 08:21 | <othermaciej> | we could probably relax that now |
| 08:22 | <Philip`> | (There were several other bugs with Safari (on Windows) when I first did this, hence the comment on the reportgen page about WebKit problems, but I think things worked much better when I last tried it) |
| 08:22 | <Philip`> | (It's still incredibly painful to run the tests with Opera Mini, though) |
| 08:27 | Philip` | should try testing all the browsers on all the platforms to see if they're doing anything funny |
| 08:27 | <othermaciej> | I'd be curious if there are any Safari discrepancies between Mac and Win |
| 08:28 | <othermaciej> | I am not sure any of the Linux WebKit ports are ready for prime time though |
| 08:28 | <othermaciej> | (yet) |
| 08:29 | <Philip`> | There were significant differences in e.g. radial gradients between Safari 3 on Windows and Safari 3 on OSX 10.4, presumably because that's handled by system libraries instead of by WebKit, so maybe they're the same now on 10.5 |
| 08:30 | <othermaciej> | might be so |
| 08:31 | <Philip`> | (If I remember correctly, http://canvex.lazyilluminati.com/tests/tests/2d.gradient.radial.inside3.html was an odd reddish-green shade on Windows, but worked correctly with the same WebKit on 10.4) |
| 08:36 | <MikeSmith> | Philip` - btw, http://canvex.lazyilluminati.com/83/play.xhtml doesn't seem to display as expected in current Webkit nightly |
| 08:44 | <Philip`> | MikeSmith: Is the game view pushed down below the ugly brown border, or is it something else? |
| 08:45 | <othermaciej> | it's an odd reddish green in Safari 3 on Mac |
| 08:45 | <othermaciej> | the view is pushed down |
| 08:45 | <othermaciej> | I think that bug might be filed in bugs.webkit.org but not sure |
| 08:46 | <Philip`> | http://bugs.webkit.org/show_bug.cgi?id=13538 |
| 08:46 | <Philip`> | othermaciej: Okay, sounds like that's the same bug I see on Windows, so at least it's consistent when using the latest versions :-) |
| 08:49 | <Philip`> | Looks quite similar to http://bugs.webkit.org/show_bug.cgi?id=16045 |
| 08:51 | <othermaciej> | sounds like |
| 08:56 | <MikeSmith> | Philip` - just the gamed pushed below the border |
| 08:56 | <MikeSmith> | that's all |
| 08:56 | <MikeSmith> | otherwise works fine for me |
| 08:56 | <MikeSmith> | as far as the behavior |
| 08:59 | <othermaciej> | the game itself plays quite smoothly in Safari |
| 09:04 | <Philip`> | I think I don't even have any WebKit-specific hacks in there, whereas I do for Firefox and Opera, but maybe that's just because I couldn't test it in WebKit until after the WebKit developers had already fixed the bugs |
| 10:16 | <hsivonen> | hendry: I no longer see anything "transparent" on http://validator.nu/?doc=http%3A%2F%2Fvideo.natalian.org%2F |
| 10:33 | <zcorpan> | hsivonen: "Contexts in which element hr may be used: |
| 10:33 | <zcorpan> | Empty." |
| 10:33 | <zcorpan> | doesn't seem to be correct |
| 10:34 | <hsivonen> | zcorpan: indeed not. it's the content model |
| 10:34 | <zcorpan> | yep |
| 10:48 | <Hixie> | nn |
| 11:16 | <hendry> | hsivonen: http://validator.nu/?doc=http%3A%2F%2Fvideo.natalian.org%2F06 |
| 11:16 | <hendry> | hsivonen: i re-arranged some stuff |
| 11:17 | <hendry> | nice email on the webkit list about support for HTML5. Sweeet. |
| 11:18 | <hsivonen> | hendry: thanks. |
| 11:18 | <hsivonen> | it appears that there's a parser bug |
| 11:18 | <zcorpan> | hendry: pointer? |
| 11:19 | <hsivonen> | too bad that it will be tedious to catch when it happens midway a document of the size of the HTML5 spec |
| 11:19 | <hsivonen> | the string "Contexts in which this element may be used:" gets some garbage inserted into it |
| 11:21 | <hendry> | zcorpan: http://lists.webkit.org/pipermail/webkit-dev/2007-November/002921.html |
| 11:27 | <hendry> | http://bugs.webkit.org/show_bug.cgi?id=16145 |
| 11:27 | <hendry> | wtf is Mozilla doing with <video>. they seem to be dragging their heels with HTML5 stuff |
| 11:29 | <hsivonen> | hendry: any particular wtf in mind? |
| 11:33 | <hendry> | hsivonen: i am not tracking their bts, so i am not sure what's going on |
| 11:34 | <hendry> | though I recall asking if video/element would be in firefox 3 and that didn't seem to be happening |
| 11:34 | <hendry> | sorry, i should really check myself how far they are on WF2 & video/audio before I bitch :) |
| 11:37 | <hendry> | i hate bugzilla. i guess i should read some manual, how to bookmark the bugs I want to see, saved searches isn't cutting it. |
| 11:40 | <hsivonen> | hendry: from reading b.m.o, it seems to me that <video> was too late to make it to Firefox 3 |
| 11:40 | <hsivonen> | But I don't work on Firefox, so don't trust me on this point. |
| 11:44 | <hendry> | any ideas about WF2? I think WF2 is low hanging fruit, isn't it? There is JS for IE support, no? http://olav.dk/wf2/demo/ |
| 11:44 | <hendry> | https://bugzilla.mozilla.org/show_bug.cgi?id=345822 a bug I found about WF2 in Firefox |
| 11:47 | <hendry> | does that bug make any sense to you guys? |
| 11:47 | <hendry> | oh and there is https://bugzilla.mozilla.org/show_bug.cgi?id=344614 |
| 11:49 | <hsivonen> | hendry: I'm not up to speed with the WF2 situation |
| 11:50 | <hendry> | that bug depresses me |
| 11:50 | <hendry> | or dissapoints rather :) |
| 11:56 | <zcorpan> | hendry: thanks |
| 12:09 | <hsivonen> | hendry: fixed: http://validator.nu/?doc=http%3A%2F%2Fvideo.natalian.org%2F06 |
| 13:23 | <hendry> | hsivonen: great |
| 13:24 | <hendry> | so now all i have is block/inline issues |
| 13:24 | <hendry> | i hate the distinction between block/inline. Is there _really_ i need for this distinction or am I asking a really dumb question this lunch time? :) |
| 13:27 | <hsivonen> | hendry: you need to discuss the block/inline thing with Hixie |
| 13:31 | zcorpan | suspects that block/inline is going to make some authors ignore conf checking altogether |
| 13:32 | <hsivonen> | I'd like to see glazou state his block/inline opinions/requirements on public-html |
| 13:49 | hendry | agrees with zcorpan |
| 13:52 | <Philip`> | I almost never want to check for conformance, I just want to check for errors, where I define "errors" as anything I did without noticing or without knowing the consequences |
| 13:53 | <Philip`> | but the only tools available are validators / conformance checkers, so I have to put up with them not doing exactly what I want them to do |
| 13:54 | <hsivonen> | Philip`: do you have a precise spec for what you want checked? :-) |
| 13:55 | <Philip`> | hsivonen: No, since it would involve a tool that could read my mind :-) |
| 13:55 | <hendry> | hsivonen: can you validator check for payload? gzip type ?? |
| 13:56 | <hendry> | sorry my keyb went mad on me. payload size I meant. |
| 13:56 | <hendry> | be good if you could link into CSS validators and things like jslint.com too |
| 13:57 | <hsivonen> | hendry: request recorded but I didn't understand the question |
| 13:57 | <hendry> | hsivonen: i think it is useful to know how big your Web page is. And to ensure that something sensible like gzip compression on the server is being utilised |
| 13:58 | <hsivonen> | hendry: ah. ok |
| 13:58 | <hsivonen> | hendry: the payload size is checked only as a countermeasure agaings DoS |
| 13:58 | <hendry> | other components of a web page, like CSS and JavaScript should also be validated and "linted" too somehow. |
| 13:58 | <hsivonen> | hendry: not in the UI ATM |
| 13:58 | <hsivonen> | hendry: request recorded |
| 13:59 | <Philip`> | hsivonen: There is e.g. http://html5.validator.nu/?doc=http://www.cl.cam.ac.uk/~pjt47/ where the conformance checker complains about something that I did intentionally, which makes the tool useless since that stops it looking for anything that I might consider a real error that I would like to fix |
| 14:00 | <hsivonen> | Philip`: if we as a group decide that what you are doing is a useful backwards compat design patter, I'd expect Hixie to make it conforming |
| 14:01 | <Philip`> | The group's view of what is an error will not always be exactly the same as my view, and I want a tool that exactly reflects my view |
| 14:03 | <Philip`> | (at least when I'm being an author - I have different desires when being a user of other people's pages) |
| 14:04 | <Lachy> | Philip`, why not use divs instead of span, which would be conforming? |
| 14:05 | <Philip`> | Lachy: Because it won't get parsed correctly in Firefox |
| 14:05 | <Lachy> | oh, so you're using the span has a compatibility hack |
| 14:05 | <Philip`> | Yes |
| 14:06 | <Philip`> | so it's not something I'm going to fix in my code, because I put it there intentionally, and it's not helpful for a conformance checker to keep complaining about it |
| 14:08 | <Philip`> | I haven't got a clue how/whether it's possible to implement a better approximation to my desires than the current single 'document conformance' concept |
| 14:10 | <Philip`> | (so I can't give any constructive suggestions) |
| 14:12 | <Philip`> | (and the current approximation isn't necessarily bad, and couldn't necessarily be made better in practice, but it's just not perfect :-) ) |
| 14:23 | <hendry> | hsivonen: could you use the WHATWG favicon on validator? I have so many tabs open nowadays. If things don't have a favicon I'm lost :) |
| 14:23 | <hsivonen> | hendry: using the WHATWG favicon would be a bad idea, but I think having *a* favicon would be good |
| 14:24 | <hsivonen> | hendry: recording request |
| 14:24 | <nickshanks> | Hixie: are you around? |
| 14:24 | <hendry> | hsivonen: http://www.favicon.cc/ |
| 14:24 | <hendry> | hsivonen: why is a bad idea to suggest some sort of affiliation between whatwg and validator.nu? |
| 14:25 | <nickshanks> | hsivonen: you might know! did ian (or anyone else) ever write a good piece about the theme of "great URLs never die" |
| 14:25 | <nickshanks> | i want to educate some people |
| 14:25 | <hsivonen> | nickshanks: TimBL wrote Cool URIs Don't Change |
| 14:25 | <nickshanks> | yeah, that'll do |
| 14:25 | <hsivonen> | hendry: because the WHATWG doesn't endorse a single validator |
| 14:36 | <Philip`> | The WHATWG endorses validators in general, and there's only one in practice, so that sort of counts as an implicit endorsement :-) |
| 14:53 | <dglazkov> | hello kind HTML5 folk |
| 14:54 | <zcorpan> | hi |
| 14:54 | <dglazkov> | a while back, I put together the HTML5 SQL player and posted about it on WHATWG list |
| 14:54 | <nickshanks> | what, HTML5 folk?! where? |
| 14:55 | <dglazkov> | I haven't heard any response, so I am wondering if it would be a good thing for me to continue improving upon it |
| 14:56 | <dglazkov> | my POV is that it's always better to develop/enhance the spec based on an implementation, rather than thin air |
| 14:57 | <dglazkov> | but I could be wrong |
| 14:58 | zcorpan | shares that POV |
| 14:59 | <zcorpan> | though, the sql part of html5 is not my expertise |
| 15:00 | <nickshanks> | HTML5 is getting bloated IMO |
| 15:01 | <zcorpan> | yeah |
| 15:01 | <nickshanks> | SQL should be in a seperate, optional module |
| 15:01 | <zcorpan> | moving stuff doesn't reduce bloat, though :) |
| 15:02 | <zcorpan> | if the web starts using sql, implementors will have to support sql |
| 15:02 | <zcorpan> | no matter where or even whether it is defined |
| 15:02 | <nickshanks> | hmm, maybe I should write a web browser that only renders HTML 1.0 |
| 15:02 | <zcorpan> | how about xhtml2? :) |
| 15:03 | <nickshanks> | http://web.nickshanks.com/books/gettinggold.xhtml2 |
| 15:03 | <nickshanks> | i'm waiting for that URL to begin working one day |
| 15:05 | <Philip`> | "Error 404"? |
| 15:06 | <hsivonen> | zcorpan: what's the current best list of what ARIA states apply to which roles |
| 15:06 | <nickshanks> | oops, it's hyphenated |
| 15:06 | <hsivonen> | ? |
| 15:06 | <zcorpan> | hsivonen: not sure |
| 15:09 | <zcorpan> | perhaps http://developer.mozilla.org/en/docs/ARIA_to_API_mapping#Role_mappings.2C_with_properties_that_depend_on_the_role |
| 15:13 | <hsivonen> | zcorpan: hmm. that's not normative and it has restrictions that http://www.w3.org/TR/2007/WD-aria-role-20071019/ and http://www.w3.org/TR/aria-state/ do not |
| 15:13 | <hsivonen> | have |
| 15:14 | <nickshanks> | Philip`: are you a student? |
| 15:15 | <hsivonen> | zcorpan: does "Inherits into Roles" mean that the state or property can also occur directly on those roles? |
| 15:15 | <Philip`> | nickshanks: Yes |
| 15:15 | <nickshanks> | studying what? |
| 15:16 | <Philip`> | Currently doing CS PhD |
| 15:17 | <zcorpan> | hsivonen: don't know, actually |
| 15:17 | <nickshanks> | does that entitle you to avoid council tax? |
| 15:18 | <zcorpan> | i need to get an understanding of how aria really works, and spec it down |
| 15:19 | <Philip`> | nickshanks: I assume so, since I'm not paying any tax at all |
| 15:19 | <nickshanks> | i have a CS PhD student who's just moved into my house and i'm wondering if that means i lose my single person discount |
| 15:38 | <Philip`> | nickshanks: Hmm, I've got no idea - I'm just living in a college-owned house with other students, and all the details are taken care of by other people :-) |
| 15:38 | <nickshanks> | fair enough |
| 17:26 | <Philip`> | MikeSmith: About your tactics-gapi-canvas comment: I wouldn't think that OpenGL experts would have much expertise relevant to the current canvas API, since it's quite a different area |
| 17:27 | <MikeSmith> | Philip` - that was an indirect comment I was attempting to pass on from other team members |
| 17:27 | <Philip`> | (though they would know relevant things for a future 3D canvas API) |
| 17:27 | <Philip`> | MikeSmith: Okay |
| 17:28 | <MikeSmith> | I hope to get more people on the team to read/review the actual canvas API spec |
| 17:39 | <Philip`> | It's harder to review now that it's about maintenance/evolution of the specification of an old technology, and not about the design of something new, so there's all the 'don't break the web' problems that make it hard to change much |
| 17:41 | <zcorpan> | 'text-indent' doesn't inherit to tables in firefox and ie, apparently |
| 17:41 | <Philip`> | but it'd be good to have people look at it if they're aware of what modifications are acceptable |
| 17:50 | <MikeSmith> | Philip` - yeah. Sorta perhaps indicates the spec and implementations around canvas are of a level of maturity that's a bit beyond what is normally required for publishing as part of a FPWD |
| 17:50 | <MikeSmith> | some might say |
| 18:03 | Philip` | finds another case where Firefox 2 writes random memory contents to a <canvas>, but, given the consequences of the most recent attempt at patching the same kind of bug in FF2, expects it would be safer to not tell anyone about it |
| 18:06 | <gavin> | bah! |
| 18:06 | <gavin> | you should certainly file a bug about it |
| 18:06 | <gavin> | the most recent troubles with the patch for that bug were caused by process issues, not code issues |
| 18:12 | <Hixie> | Lachy_: anything useful come out of the telecon? |
| 18:13 | <gavin> | Philip`: when you do file it, CC me? |
| 18:17 | <Philip`> | gavin: Okay, will do - just trying to implement an exploit for the bug, to make it more fun |
| 18:18 | <gavin> | perhaps you should mark the bug security-sensitive |
| 18:18 | <gavin> | I forget - did you do that for the last one? |
| 18:22 | <Philip`> | "_closedTabs:[{state:{entries:[{url:"http://www.mozilla.com/en-US/firefox/";, children:[], title:"Firefox web browser | Faster, more secure, & customizable", cacheKey:0, ID:1, scroll:"0,0"}], index:1, zoom:1, disallow:"", xultab:"", extData:null, text:""}, ..." |
| 18:22 | <Philip`> | I'm guessing web pages shouldn't have access to strings like that |
| 18:24 | <Philip`> | Bah, I thought I found another problem but actually it was just dust on my monitor that looked like random pixels |
| 18:26 | <gavin> | heh |
| 18:26 | <gavin> | Philip`: that's right, web pages don't |
| 18:26 | <gavin> | we just use a JSON-like format for serializing to disk |
| 18:27 | <Philip`> | gavin: s/don't/shouldn't/ :-p |
| 18:27 | <gavin> | I'm telling you that they don't |
| 18:28 | <gavin> | if you have evidence otherwise, please to be filing a bug? :) |
| 18:35 | <Philip`> | gavin: https://bugzilla.mozilla.org/show_bug.cgi?id=406036 |
| 18:35 | <Philip`> | I would CC you if I knew what your email address was :-) |
| 18:35 | <gavin> | that's OK, I can CC myself |
| 18:36 | <gavin> | (gavin.sharp matches me) |
| 18:37 | <Philip`> | gavin: I didn't mark the last bug security-sensitive, because I didn't really think about the security issue at first, and then I did realise but couldn't sensitise it myself so someone else did |
| 18:37 | <Lfe> | ok |
| 18:37 | <gavin> | Philip`: ok |
| 18:37 | <Lfe> | oops :o |
| 18:38 | gavin | is put off by a branch build's non-native looking Mac widgets |
| 18:38 | <zcorpan> | hsivonen: i've implemented collapsing of messages for validator.nu with javascript |
| 18:39 | <zcorpan> | hsivonen: http://simon.html5.org/temp/validator-nu-collapse.html |
| 18:40 | <zcorpan> | hmm, but the button's label is wrong :) |
| 18:41 | <Lachy> | zcorpan, clicking collapse multiple times has a weird bug |
| 18:42 | <hsivonen> | zcorpan: cool. Thanks. |
| 18:42 | <zcorpan> | Lachy: now why would you do that ;) |
| 18:43 | <hsivonen> | zcorpan: I will look into integrating this |
| 18:43 | <Lachy> | I thought it would toggle it |
| 18:43 | <hsivonen> | zcorpan: is this under the MIT license? |
| 18:43 | <zcorpan> | hsivonen: if it helps :) |
| 18:43 | <hsivonen> | zcorpan: it would |
| 18:43 | <zcorpan> | then it is |
| 18:44 | <hsivonen> | zcorpan: excellent. Thank you |
| 18:44 | <Lachy> | hsivonen, I spoke to DanC earlier about the licence for the authoring guide |
| 18:44 | <zcorpan> | do i need to include a boilerplate? |
| 18:44 | <hendry> | is this old news? http://www.crockford.com/html/ |
| 18:44 | <Lachy> | he assigned anne to look into the issue |
| 18:44 | <hsivonen> | zcorpan: preferably, yes, so I don't misrepresent anything by adding it myself |
| 18:45 | gavin | apologizes for spamming Philip` with bugmail |
| 18:45 | <zcorpan> | pointer to such a boilerplate, please? |
| 18:45 | <hsivonen> | zcorpan: http://www.opensource.org/licenses/mit-license.php |
| 18:45 | <Lachy> | hsivonen, http://lists.w3.org/Archives/Public/spec-prod/2007OctDec/0007 |
| 18:46 | <hsivonen> | Lachy: excellent |
| 18:46 | <Philip`> | gavin: No problem, I tend to ignore that email address anyway :-) |
| 18:47 | <hsivonen> | zcorpan: with JS-compatible comments: http://about.validator.nu/htmlparser/apidocs/src-html/nu/validator/htmlparser/common/DoctypeExpectation.html |
| 18:48 | <zcorpan> | hsivonen: ok, added |
| 18:49 | <hsivonen> | zcorpan: thank you |
| 18:50 | <nickshanks> | sometimes i hate google |
| 18:50 | <nickshanks> | it always ranks my website too high |
| 18:50 | <Hixie> | i know the problem |
| 18:50 | <nickshanks> | http://images.google.com/images?q=keyboard+layout |
| 18:51 | <zcorpan> | Lachy: bug fixed |
| 18:51 | <hsivonen> | nickshanks: yeah, that has interesting effects. like people asking me about Japanese fonts |
| 18:51 | <gsnedders> | peh. take a look at http://google.com/search?q=homophobic+insults |
| 18:51 | <gsnedders> | I'm the second result (or I was a few days ago)! |
| 18:51 | <nickshanks> | heh |
| 18:52 | <Philip`> | http://krijnhoetmer.nl/irc-logs/ was high on the image search for "periodic table" a while ago |
| 18:52 | <nickshanks> | thing is this costs me money because i've gone way over my bandwidth limits this month |
| 18:52 | <zcorpan> | nickshanks: Disallow / |
| 18:52 | <zcorpan> | or is it Disallow: / |
| 18:53 | <nickshanks> | there's apachectl stop |
| 18:53 | <nickshanks> | which saves even more |
| 18:53 | <nickshanks> | most hits i get are Windows people looking for free fonts |
| 18:54 | <nickshanks> | all my free fonts are AAT-based .dfont files |
| 18:55 | <gsnedders> | most common search referrer to my site: sports day speech |
| 18:55 | <zcorpan> | ah, i still have a bug |
| 18:55 | <gsnedders> | second is gsnedders |
| 18:55 | <gsnedders> | third is: causes of love |
| 18:55 | <gsnedders> | how lovely |
| 18:55 | <gsnedders> | :\ |
| 18:55 | <nickshanks> | where is 'causes of divorce' ? |
| 18:56 | <gsnedders> | no, I only tag posts with lust or love, not divorce. |
| 18:56 | <gsnedders> | I haven't done the marrying part yet (and legally can't) |
| 18:56 | <nickshanks> | causes of lust? |
| 18:56 | <zcorpan> | there |
| 18:56 | <nickshanks> | how old are you? |
| 18:56 | <gsnedders> | nickshanks: yeah, I have that (causes of lust) |
| 18:56 | <gsnedders> | nickshanks: 15 |
| 18:56 | <nickshanks> | you can get married |
| 18:56 | <gsnedders> | "causes of emo" too |
| 18:56 | <nickshanks> | france 15 |
| 18:57 | <nickshanks> | scotland 14 |
| 18:57 | <gsnedders> | nickshanks: Scotland is 16 |
| 18:57 | <nickshanks> | japan: any age |
| 18:57 | <nickshanks> | 14 with consent |
| 18:57 | <nickshanks> | of parents |
| 18:57 | <gsnedders> | nickshanks: no, there is no difference between consent of parents or not |
| 18:57 | <gsnedders> | nickshanks: such a law only exists in England and Wales |
| 18:58 | <nickshanks> | oh. hmm. |
| 18:58 | <hober> | (and NI) |
| 18:58 | <zcorpan> | hmm, doesn't hasAttribute() work in ie? |
| 18:59 | <nickshanks> | gsnedders: http://web.nickshanks.com/analog/ |
| 18:59 | <gsnedders> | more odd searches: how long does it take to bleed to death from slitting wrists, get me an essay about anything, emo poems about falling in love but you can't have them |
| 18:59 | <nickshanks> | scroll down to search terms |
| 18:59 | zcorpan | changes to getAttribute |
| 19:00 | <nickshanks> | (i just changed webservers so it's a couple of weeks out of date and the images don't work - need to set analog back up again on the new server) |
| 19:00 | <gsnedders> | nickshanks: heh. I only have stuff since I changed host. Too much rubbish in older ones (from spam bots and the like) |
| 19:01 | <Dashiva> | zcorpan: You probably want to check .attributes[name].defined or what it is |
| 19:01 | <nickshanks> | if you look at index-full.html that goes back to may 2005 |
| 19:02 | <Hixie> | aha, this is why i'm getting flaky mail service right now http://www.dreamhoststatus.com/2007/11/17/gmail-forwarding-slowness |
| 19:02 | <zcorpan> | Dashiva: oh yep. though getAttribute worked as well |
| 19:02 | <zcorpan> | hsivonen: it doesn't work in ie6 because it doesn't support attribute selectors |
| 19:03 | zcorpan | has to go now |
| 19:03 | <hsivonen> | zcorpan: ok |
| 19:22 | <Lachy__> | Apparently, I'm "not so democratic or balanced" http://lists.w3.org/Archives/Public/www-archive/2007Nov/0104.html |
| 19:35 | <Dashiva> | Lachy__: You aren't me, so you're obviously not balanced |
| 19:36 | <hsivonen> | Lachy__: you can become an unbalanced dictator |
| 19:37 | <Lachy__> | I don't mind being called unbalanced, but I fail to see how any of my responses in regards to the web dev guide have been undemocratice |
| 19:38 | <Lachy__> | s/undemocratice/undemocratic/ |
| 19:38 | <Dashiva> | "The HTMLWG is becoming less and less democratic everyday. It has become a dictatorship driven by three companies: Google, Apple and Opera." |
| 19:38 | <Dashiva> | If that was true, we'd be publishing HTML 5 drafts by now >:O |
| 19:39 | <Hixie> | i don't understand what Roy and Julian want |
| 19:40 | <Hixie> | what would it mean to describe the language without UA requirements? |
| 19:40 | <hsivonen> | Hixie: see the ODF spec :-) |
| 19:40 | <Hixie> | but that's a _bad_ thing |
| 19:42 | <Hixie> | is it just the scripting parts he wants taken out? |
| 19:42 | <Lachy> | I don't see how it is possible to address Roy's concern, since the group already resolved to call the spec HTML5 and the group knew exactly what the proposal contained when the decision was made |
| 19:42 | <Lachy> | he wan't anything that isn't related to the markup taken out, despite the fact that it's not actually possible to do with some things |
| 19:43 | <Hixie> | what isn't related to the markup? |
| 19:43 | <Hixie> | as far as i can see, everything in the spec is related to the markup |
| 19:44 | <Lachy> | no, I mean, just the syntax and element semantics |
| 19:44 | <tndH> | i read it as "pretend script doesn't exist and all documents are valid" |
| 19:45 | <Hixie> | "just the syntax and element semantics" would be mostly everything but scripting parts, no? |
| 19:45 | <Hixie> | does he want the parser out too? |
| 19:45 | <Lachy> | I presume he would be aware that we can't define how to parse HTML without at least considering document.write and innerHTML |
| 19:45 | <Hixie> | i can't tell |
| 19:45 | <Hixie> | i guess i'll wait for my mail to catch up and then ask him |
| 19:45 | <Lachy> | There are certainly some people who don't think the parser should be in there |
| 19:46 | <Hixie> | it's going to be hard to address everyone's desires in that case |
| 19:46 | Hixie | wonders how to address DanC's latest claim that we don't have critical mass |
| 19:50 | <Dashiva> | It didn't take long before the talk about "w3c members" as opposed to "html wg members" came up... |
| 19:51 | <Hixie> | 8 months, that's not that short |
| 19:52 | <Dashiva> | Well, I was counting since the vote :) |
| 19:52 | <Hixie> | i'm amused because the companies he mentioned haven't, by and large, voted on any of the votes |
| 19:52 | <Hixie> | it's not clear how to address his concern |
| 19:55 | <hsivonen> | "W3C Reopens Emotion Incubator Group" |
| 20:25 | <othermaciej> | our charter specifically requires us to specify the DOM |
| 20:25 | <othermaciej> | Roy's request that we violate the charter is not really in order |
| 20:25 | <Hixie> | hah |
| 20:25 | <Hixie> | feel free to call him on it |
| 20:26 | <othermaciej> | I think you addressed his comments sufficiently |
| 20:26 | <Hixie> | well i still haven't received that e-mail |
| 20:26 | <Philip`> | The charter says it's "in scope" - does that actually mean it's required? |
| 20:28 | <othermaciej> | well, it's hard to argue that we "shouldn't" do a category of things that the charter explicitly declares in scope |
| 21:41 | <Hixie> | hey, the forms tf missed their deadline |
| 23:37 | <Hixie> | wow, what insanity |
| 23:37 | <Hixie> | search for inurl:"nodetag.js" on Yahoo! and follow the trail |
| 23:38 | <Hixie> | i honestly thought it was some sort of botnet artefact for a while |
| 23:38 | <Hixie> | turns out to be what appears to be part of a legitimate survey network (questionmarket.com) |
| 23:40 | <Hixie> | their scripts have such wacky things as: |
| 23:40 | <Hixie> | function DL_rw_0_8721_0() |
| 23:40 | <Hixie> | { |
| 23:40 | <Hixie> | if (0 != 4 && !0) { |
| 23:40 | <Hixie> | setTimeout("DL_jsc_0_8721_0()" , Math.floor(Math.random() * 5) * 200 + 500); |
| 23:40 | <Hixie> | } else { |
| 23:41 | <gavin> | o_O |
| 23:42 | <Hixie> | wow, a whole bunch of sites have identical stat.js files |
| 23:43 | <Hixie> | aha, part of ShopFactory |
| 23:59 | <Philip`> | "The field of view shall be greater than zero and smaller than π" - that's quite an unhelpful character-misencoding :-( |