| 00:10 | Philip` | wonders if he can do a regexp conversion of the tree construction algorithm into executable code |
| 02:32 | <Hixie> | was http://blog.whatwg.org/faq/ ever a valid URI? |
| 02:32 | <Hixie> | if so the redirect is wrong |
| 02:39 | <Lachy> | ok, I'll fix it |
| 02:40 | <Lachy> | fixed |
| 02:41 | <Lachy> | good night |
| 07:55 | <hsivonen> | http://www.imc.org/atom-syntax/mail-archive/msg20402.html |
| 07:58 | <othermaciej> | tasty! |
| 08:02 | <othermaciej> | it's good to know that Namespaces in XML are still busy providing non-solutions to non-problems |
| 08:13 | <MikeSmith> | that seems worth quoting |
| 08:13 | <MikeSmith> | "I think the lesson here is to write better specs, not rail against programmers when they are presented with bad ones." |
| 08:14 | <hsivonen> | zcorpan: stray script happens after </body> |
| 08:14 | <hsivonen> | http://html5.validator.nu/?doc=http%3A%2F%2Fwww.imageshack.us%2F&showsource=yes#l563c73 |
| 08:14 | <hsivonen> | http://html5.validator.nu/?doc=http%3A%2F%2Fwww.google.com.hk%2F&showsource=yes#l3c4938 |
| 09:56 | <Philip`> | https://bugzilla.mozilla.org/show_bug.cgi?id=415168 |
| 11:21 | <annevk> | http://googleblog.blogspot.com/2008/02/yahoo-and-future-of-internet.html |
| 12:04 | <hsivonen> | zcorpan: Validator.nu now permutes attributes so that the RELAX NG layer always sees the type attribute of <input> and <command> before the other element-specific attributes, the content attribute of <meta> after the other attributes and the template attribute before everything else |
| 12:04 | <hsivonen> | oh, and the wrap attribute of textarea is hoisted, too |
| 12:35 | <hsivonen> | zcorpan_: replies in logs |
| 12:36 | zcorpan_ | looks |
| 12:37 | <zcorpan_> | hsivonen: nice |
| 14:13 | <Philip`> | Where HTML5 says "insert an HTML element with the same tag name as the token's.", that sounds like you don't give it the same attributes as the current token's, but that doesn't seem to make sense |
| 14:16 | <annevk> | seems that http://www.whatwg.org/specs/web-apps/current-work/multipage/section-tree-construction.html#creating addresses that |
| 14:21 | <Philip`> | "create an element for a token" happens during "insert an HTML element for a token", but inserting an element with another token's tag name is not the same as inserting an element for that token |
| 14:21 | <annevk> | indeed, the difference is the tag name |
| 14:23 | <Philip`> | The difference is that inserting an element with another token's tag name does not define an actual token to insert the HTML element for |
| 14:23 | <Philip`> | I don't see why it doesn't just say "Insert an HTML element for the token" |
| 14:25 | <annevk> | you're right, that would make more sense |
| 14:46 | <zcorpan_> | http://forums.whatwg.org/viewtopic.php?t=142 |
| 14:51 | <annevk> | replied |
| 14:59 | <annevk> | zcorpan_, in http://simon.html5.org/specs/xml-stylesheet5 "content type" should probably say "content type, excluding parameters," |
| 15:09 | <zcorpan_> | annevk: thanks, changed to "MIME type" and said to ignore parameters where appropriate |
| 15:10 | <annevk> | fwiw, I believe the preferred term is "media type" |
| 15:11 | <zcorpan_> | what's the difference? a MIME type can have parameters and media type is a subset of MIME type? |
| 15:12 | <annevk> | no, they both mean the same thing, it's just that one is preferred per my recollection |
| 15:13 | <zcorpan_> | oh, ok. media type conflicts with MQ in my mind :) |
| 15:46 | <aroben> | Hixie: the tracker seems to be non-functional (e.g., <http://html5.org/tools/web-apps-tracker?from=1196&to=1197>;) |
| 15:47 | <aroben> | Hixie: for me it says "No result." |
| 15:49 | <annevk> | he's not responsible |
| 15:49 | <annevk> | seems that svn.whatwg.org is down |
| 15:49 | <annevk> | for which he is responsible :p |
| 16:51 | <Philip`> | Hixie: On a page like http://www.whatwg.org/specs/web-apps/current-work/multipage/section-tree-construction.html the annotation system login box hides some text behind it, which is quite undesirable |
| 18:21 | <gsnedders> | what's the POSIX command that gets the path of a command (i.e., curl would return /usr/bin/curl if $PATH said to look there first and it was there)? |
| 18:23 | <gsnedders> | which |
| 18:27 | <Philip`> | That doesn't seem to be POSIX |
| 18:28 | <Philip`> | (Hooray, now I can parse token streams that contain nothing but comments) |
| 18:29 | <Philip`> | (except it gets it backwards if you have >= 1 comment) |
| 20:57 | <blooberry> | grr. renee saarsoo's dmoz research has disappeared from his site. *consults google cache* |
| 21:03 | <Philip`> | blooberry: http://web.archive.org/*/triin.net/2006/06/12/HTML seems to work |
| 21:04 | <blooberry> | philip`: *nod* I'll email him too. I just received a response from him a week or so about his research, so I doubt that the site is stale |
| 21:04 | <blooberry> | something probably just went all broke-y. 8-} |
| 21:12 | <jwalden> | Hixie: https://bugzilla.mozilla.org/show_bug.cgi?id=387706#c52 |
| 21:12 | <jwalden> | ^ also for anyone else who cares about HTML5 postMessage |
| 21:12 | <jwalden> | :-\ |
| 21:15 | <SadEagle> | jwalden: http://lxr.kde.org/source/KDE/kdelibs/kdeui/util/fixx11h.h?v=4.0-branch#250 |
| 21:18 | <othermaciej> | jwalden: seems like a Mozilla issue that to name something postMessage you need to conflict with the name PostMessage |
| 21:18 | <jwalden> | I agree, more or less |
| 21:19 | <jwalden> | I think it's more a Windows issue than anything else, tho |
| 21:19 | <jwalden> | SadEagle: does that work in headers? claim at the top of that file is it doesn't |
| 21:19 | <othermaciej> | well in WebKit for instance, there doesn't have to be any identifier named "PostMessage" to name something "postMessage", but I agree also that Windows system headers are a cesspool of despair |
| 21:20 | <Philip`> | Just #undef PostMessage and then use PostMessageW whenever you want to talk to Windows |
| 21:20 | <SadEagle> | jwalden: I don't see why it won't. The problem is, the trick would rename the Win32 postMessage to something else. And your windows port might actually use that |
| 21:21 | Philip` | assumes it works sort of like that, but could be totally wrong |
| 21:21 | <jwalden> | not our code that's breaking, is the problem |
| 21:21 | <SadEagle> | Philip`: except when PostMessageA is wanted :-) |
| 21:21 | <jwalden> | I'd totally use that fix if we were broken |
| 21:21 | <Philip`> | SadEagle: But nobody wants ANSI any more :-) |
| 21:21 | <SadEagle> | othermaciej: they got nothing on X headers :-) |
| 21:39 | <roc> | aaargh, the memories |
| 21:39 | <roc> | "#define Status int" |
| 21:41 | <othermaciej> | given that Microsoft's strategy is to focus on developers, you have to wonder what it would be like if they focused on end uers |
| 22:01 | <jwalden> | hrm |
| 22:02 | <jwalden> | MS extensions for the sorta-win, at a glance: http://msdn2.microsoft.com/en-us/library/hsttss76.aspx |
| 22:03 | <Philip`> | Hmm, they totally broke MSDN in Opera |
| 22:03 | <jwalden> | it's pretty horked in Firefox too |
| 22:03 | <webben_> | MSDN is amazingly broken most of the time even in newer versions of their own browser |
| 22:03 | <jwalden> | parts of the page blanked out corresponding to their dropdown menus? |
| 22:03 | webben_ | remembers struggling to read it in IE7 Beta |
| 22:04 | <jwalden> | if that's what it is, hovering to open them and then moving off to close usually fixes it for me |
| 22:04 | <SadEagle> | It works in konq when identifying as firefox... but nor normally |
| 22:06 | <Philip`> | It probably doesn't help that one of their CSS files is an HTML page saying "Sorry, we were unable to service your request." as text/css |
| 22:06 | <jwalden> | har har |
| 22:08 | webben_ | just wishes they'd fix it; it's a massively useful source they've made incredibly difficult to use |
| 22:10 | <Philip`> | In Opera it just has no stylesheets |
| 22:10 | <Philip`> | I've got no idea where Firefox is getting its stylesheet content from, though |
| 22:12 | <SadEagle> | Philip`: does whatever opera you do support getters/setters? |
| 22:13 | <roc> | i hear the silverlight version is going to be awesome |
| 22:15 | webben_ | would just like a plain HTML search and a plain HTML output format. |
| 22:15 | <jwalden> | yeah, MSDN's going to be really flashy -- sorry, Silverlight-y -- when they get that out |
| 22:17 | <Philip`> | SadEagle: I think it doesn't |
| 22:17 | <Philip`> | (since it's Opera 9.2) |
| 22:19 | <Dashiva> | msdn looks (relatively) good in 9.5 |
| 22:19 | <Dashiva> | Not bad in 9.2 either, really |
| 22:19 | <Dashiva> | Sure, the top text is above the top of the screen, but nobody uses that text anyhow :) |
| 23:11 | Philip` | 's parser now can handle comments and <span> |
| 23:11 | <Philip`> | (but not </span> yet) |