| 02:19 | <Philip`> | Will any browser ever implement renameNode? |
| 02:20 | <jruderman> | i hope not |
| 02:21 | <jruderman> | it sounds evil |
| 02:25 | <Philip`> | I can imagine it interacting kind of badly with the parser, like if you write <i id=a><script>document.renameNode(document.getElementById('a'), 'http://www.w3.org/1999/xhtml';, 'b')</script></b>whoo |
| 02:35 | <jwalden> | "renameNode"? |
| 02:35 | jwalden | hits the Google |
| 02:36 | <jwalden> | ah, nonstandard |
| 02:40 | <heycam> | renameNode is dom3 core |
| 02:55 | <roc> | the document "List of vents (SVG 1.2)" isn't as interesting as it sounds |
| 02:56 | <heycam> | haha |
| 02:57 | <roc> | wow |
| 02:57 | <roc> | that sounds hellish |
| 02:57 | <othermaciej> | roc: it's allowed to replace the node, I think |
| 02:58 | <roc> | that helps, but it's still hellish |
| 02:58 | <othermaciej> | http://www.w3.org/TR/DOM-Level-3-Core/core.html#Document3-renameNode |
| 02:58 | <roc> | I'm reading it |
| 02:58 | <othermaciej> | I don't think we have it in WebKit |
| 02:58 | <othermaciej> | it sounds pretty crazy indeed |
| 02:59 | <othermaciej> | I think Web DOM Core should drop it |
| 02:59 | <roc> | amen |
| 03:02 | <roc> | if they keep it, it really needs to specify how renameNode affects hidden DOM state like <iframe> subdocuments and form elements |
| 03:02 | <roc> | and whether references to the old node are redirected to the new node |
| 03:03 | <roc> | and it sounds like the decision about whether to rename or create could have observable effects, which means leaving that decision undefined is a big interop problem |
| 03:22 | <othermaciej> | yes, having it undefined whether to create a new node or somehow rename one in-place seems terrible |
| 03:25 | <othermaciej> | and having the method at all when the common case is to create a new node (which I think it must be, changing an element's class on the fly just sounds like crazy-land) then it is misleading to have the API at all |
| 03:31 | <Dashiva> | And if there's any case for renaming inplace, maybe those elements could've been better handled collapsed with a @type |
| 03:33 | <othermaciej> | I can imagine cases where it might be useful, it just seems really impractical to do for real |
| 08:02 | annevk | wonders what 5ce9134be8a642e1fbf3b93593da33e9 means |
| 08:03 | <hdh> | firefox accepts Content-Encoding: SHA-1? |
| 08:19 | <jwalden> | annevk: http://weblogs.mozillazine.org/roc/archives/2007/01/ec06b3461cf0eaf.html may help |
| 08:19 | <jwalden> | an extremely low-down, dastardly way to annoy everybody, I say |
| 08:20 | <hsivonen> | jwalden: has roc published that previous time capsule yet? |
| 08:20 | <jwalden> | I don't think he's published any of them |
| 08:20 | <hsivonen> | ok. are there more than two now? |
| 08:21 | <jwalden> | I think this is the third, lemme check |
| 08:23 | <jwalden> | http://weblogs.mozillazine.org/roc/archives/2007/12/e0f4ffc0ddc898c.html |
| 08:24 | <hsivonen> | thanks |
| 08:24 | <jwalden> | those are all |
| 08:24 | <hsivonen> | no announced ETA for the revelation of any of them :-( |
| 08:25 | <jwalden> | I seem to recall IRC discussion that suggested to me that at least one of them would never be published unless "something" went really, really, devastatingly bad |
| 08:25 | <hsivonen> | hmm. |
| 08:25 | <jwalden> | as I said, full-blown underhandedness |
| 08:27 | <doublec> | we tried to get it out of him in the office today but he's a tough nut to crack |
| 08:28 | <jwalden> | yes, he's definitely a nut all right, completely agree |
| 08:33 | <jwalden> | also, http://blogs.msdn.com/oldnewthing/archive/2006/05/23/604743.aspx |
| 08:34 | <jwalden> | which was revealed, eventually |
| 08:34 | <jwalden> | http://blogs.msdn.com/oldnewthing/archive/2006/12/29/1379914.aspx |
| 08:38 | <hsivonen> | hmm. can't do <kbd> on whatwg wiki |
| 08:52 | <MikeSmith> | hsivonen: I see that v.nu is now reporting duplicate IDs |
| 08:54 | <hsivonen> | MikeSmith: yeah, I fixed that |
| 08:54 | <hsivonen> | <a name> in HTML 4.01 / XHTML 1.0 checking is still broken, though |
| 08:55 | <MikeSmith> | hsivonen: OK |
| 08:55 | <MikeSmith> | the dupe ID thing is nice |
| 08:55 | <MikeSmith> | I caught a couple that I had in the HTML5 publication notes doc |
| 09:11 | <Hixie> | renameNode should be required to replace |
| 10:48 | <hsivonen> | Hixie: valid hashed reference is not defined as having to match an id value, but now name is required on map |
| 10:49 | <hsivonen> | roc: fwiw, I've used run-in in published non-test-case Web content (Re: IRC log) |
| 10:52 | <Philip`> | http://www.hobi.com/ uses display:run-in, though seemingly to avoid an Opera bug |
| 10:52 | <hsivonen> | I used it in http://hsivonen.iki.fi/thesis/html5-conformance-checker to conform to a LaTeX emulation requirement set by the professor |
| 10:52 | <Philip`> | ((That bug seems fixed in 9.5)) |
| 10:53 | <Philip`> | http://www.w3.org/TR/xhtml+voice/ uses it too |
| 10:54 | <Philip`> | though probably they didn't test that page in a browser that supported it, since their References section looks ugly in Opera |
| 10:54 | <annevk> | given the people on the editor's list, I'd be surprised |
| 11:06 | <Philip`> | annevk: I'd still be surprised if they could have tested it in Opera, since the run-in list is inconsistently and nastily spaced, and I can't believe that was intentional :-) |
| 11:06 | <Philip`> | Hmm, those pages are the only two out of 130K with display:run-in in inline stylesheets |
| 11:08 | <othermaciej> | does any browser support display:run-in? |
| 11:08 | <othermaciej> | I guess webkit might |
| 11:08 | <othermaciej> | but display: run-in is sort of a microcosm of what is wrong with CSS |
| 11:08 | <othermaciej> | it is so special-casey it is kind of ridiculous |
| 11:09 | <hsivonen> | othermaciej: Safari and Opera support it, and it seems it has regressed in WebKit nightlies |
| 11:19 | <zcorpan> | Hixie: i think now is a pretty good time to add the new elements to the html parser (<nav>, <source>, etc) |
| 11:20 | <hsivonen> | looks like the regression I'm seeing is like https://bugs.webkit.org/show_bug.cgi?id=12522 but without the printing aspect |
| 11:30 | annevk | agrees with zcorpan |
| 11:31 | <annevk> | zcorpan, is there an issue so we can put at the top of http://www.whatwg.org/issues/top ? :) |
| 11:55 | <zcorpan> | annevk: i don't remember if there's an email about it |
| 11:59 | <annevk> | nope |
| 11:59 | <zcorpan> | haha, there's a folder "zzz" |
| 14:10 | <annevk> | apparently there's also one called aaa-productity or something |
| 14:10 | <annevk> | euh, aaa-productivity |
| 14:27 | <gsnedders> | annevk: so, what? It makes you more productive? |
| 14:35 | <MikeSmith> | jgraham_: thanks for your support on the issue-tracker discussion |
| 14:35 | <MikeSmith> | I've added some text to the group home page: |
| 14:36 | <MikeSmith> | http://www.w3.org/html/wg/#issues |
| 15:16 | <zcorpan> | hah |
| 15:16 | <zcorpan> | <div/> div.innerHTML = 'x</div>y<div>z'; |
| 15:17 | <zcorpan> | doesn't throw in firefox |
| 15:35 | <annevk> | zcorpan, does it affect the DOM? |
| 15:35 | <annevk> | I mean, how does it? |
| 15:43 | <mayhemer> | hi, I would like to discuss one misunderstanding from my side of HTML5, 4.7 Offline Web applications spec |
| 15:43 | <mayhemer> | I am not sure when exactly opportunistic namespace matching entries sould be cached |
| 15:43 | <annevk> | feel free, though the right people might not be around |
| 15:44 | <mayhemer> | annevk: do you know namely who I could talk to? |
| 15:45 | <mayhemer> | annevk: it is about offline web applicatoins cache |
| 15:51 | <annevk> | well, you could e-mail the mailing list |
| 15:51 | <annevk> | i'd expect them to be cached during some kind of background process, btw |
| 15:51 | <mayhemer> | I will do, thanks |
| 15:51 | annevk | looks at the spec |
| 16:01 | <closure> | anyone know if there is an implementation of the connection interface anywhere yet? |
| 16:02 | <closure> | the spec is still a bit shady in that section. |
| 16:02 | <zcorpan> | annevk: same as div.innerHTML = 'x' |
| 16:04 | <annevk> | closure, people have been experimenting with it, but nothing native |
| 16:26 | <closure> | annevk: figures, but is source code available of prototypes? |
| 16:28 | <closure> | especially testsuites. |
| 16:28 | <annevk> | i don't think so, you're interested in implementing? |
| 16:30 | <closure> | heh, part of my job for google summer of code alas (for es operating system) |
| 16:31 | <closure> | but i might have to postpone that task a little because the spec is still too vague |
| 16:39 | <annevk> | closure, best course of action is probably to subscribe to the mailing list, make comments on where the spec is vague and say that it has some priority because you want to implement it this summer |
| 16:40 | <annevk> | closure, not guaranteed it works btw, but it's likely your best shot (and if it doesn't work out you'll at least be ensured that your feedback will be addressed at some point in the future) |
| 16:41 | <closure> | i'll probably lurk on the mailing list then, although i do understand it's taking a backseat for now in the spec since it's a new feature |
| 16:42 | <annevk> | the spec is quite implementation driven actually |
| 16:44 | <annevk> | so if browser vendor Y says it needs <canvas> feedback integrated within a few weeks that will actually happen most of the time |
| 16:50 | <Philip`> | Uh, please s/ctx/canvas/ in my recent email |
| 16:54 | <annevk> | The two parameter version seems less elegant since it does not apply to all types |
| 16:54 | <annevk> | oh well |
| 16:57 | <Philip`> | annevk: What types other than PNG and JPEG is anyone ever going to use? |
| 16:58 | <Philip`> | (It doesn't seem a good idea to make JPEG harder to use just because of some hypothetical future that will probably never occur) |
| 17:02 | <annevk> | Philip`, maybe true, yes |
| 17:04 | <annevk> | the only other problem with the argument version is that implementing it could not be done using a prefixed attribute |
| 17:04 | Philip` | wonders if it matters if the JPEG quality value is implementation-dependent, so 0.80 could look rubbish in one browser and great in another |
| 17:04 | <Philip`> | annevk: Not quite sure what you mean |
| 17:05 | <annevk> | sorry, if we implemented this before it was standardized we'd basically deadlock the second argument of toDataURL(), but maybe it's not so bad |
| 17:06 | <Philip`> | Nowadays it only takes days to get new features standardised :-) |
| 17:07 | <annevk> | that'd be nice |
| 17:07 | <Philip`> | Named arguments like canvas.toDataURL('image/jpeg', { quality: 0.8 }) would be nice except they seem to not exist in the DOM and it'd be nicer to avoid unconventionality |
| 17:20 | <annevk> | Philip`, what exactly is the concern with html5-diff? |
| 17:20 | <annevk> | as it exists with http://www.w3.org/html/wg/html5/diff/ for instance |
| 17:20 | <annevk> | s/with/at/ |
| 17:21 | <Philip`> | annevk: Ah, I didn't realise that that and http://www.w3.org/TR/html5-diff/ etc all said "html5", and only http://dev.w3.org/html5/html4-differences/ says "html4" |
| 17:22 | <annevk> | I was trying to tell you that :) |
| 17:22 | <Philip`> | Ah, I failed to understand :-) |
| 17:22 | <annevk> | so no problem then, apart from html5-diff versus html5-differences which I'll declare a non-issue :) |
| 17:24 | <Philip`> | I'm happy to accept "html4-differences" as being a deprecated name that only exists in CVS, and CVS makes renaming far too much of a pain to bother with :-) |
| 18:59 | <annevk> | jgraham_++ |
| 20:13 | <Philip`> | http://meyerweb.com/eric/html-xhtml/html5-linking.html |
| 20:15 | <Hixie> | send mail |
| 20:15 | <Hixie> | but thanks |
| 20:15 | <Hixie> | that's a really useful page |
| 20:15 | <Hixie> | i wonder why eric isn't just posting to the list |
| 20:15 | <gsnedders> | for the code example, it would require @href to have per element specific semantics |
| 20:15 | <Philip`> | The accompanying blog post says "Now, as to why I’m blogging this instead of taking it to one or another of the relevant mailing lists: I’m looking for community input in order to strengthen the document." |
| 20:16 | <Hixie> | the community is on the mailing list... |
| 20:16 | <Philip`> | so I assume he will email it eventually |
| 20:16 | <Philip`> | There is more than one community in the world :-) |
| 20:16 | <gsnedders> | Philip`: We are… teh community! |
| 20:16 | <Hixie> | fair enough |
| 20:16 | <Hixie> | i just hate it when people do things and get it to where they think it's perfect before bringing it to the table |
| 20:17 | <Hixie> | because then if there's a fundamental flaw (like "browsers have already said they'll never do it") then they get all upset that they spent all this time on nothing |
| 20:18 | <Philip`> | Browser developers can change their minds if given sufficiently compelling arguments |
| 20:18 | <Dashiva> | "It would cause browsers to explode and destroy most of our users" |
| 20:20 | <Philip`> | Dashiva: If that were the case, all browser developers should be requesting strongly that the feature should be added to the spec, in the hope that their competitors will foolishly decide to actually implement it and kill their users and lose their market share |
| 20:20 | <Hixie> | that would be unethical |
| 20:20 | <Hixie> | and arguing against browser developers tends to alienate them |
| 20:20 | <Hixie> | i'd like to reserve that for the critical things |
| 20:21 | <Dashiva> | Philip`: That wouldn't work, the web developers would be victimized before they could make their sites public |
| 20:23 | <Philip`> | Dashiva: Not the developers who randomly copy-and-paste bits of tutorials and code from other sites into their own pages, then only test in IE5 and see that it looks kind of alright so they publish it |
| 20:23 | <gsnedders> | But IE5 is awesome! |
| 20:23 | <Dashiva> | But IE is the only browser that could implement it, the other browsers could never get it out as it would fail QA :) |
| 20:24 | <Philip`> | Dashiva: Nobody who experienced the bug would be in a fit state to file a bug report, so the developers would never know there was a problem |
| 20:25 | <Dashiva> | It's not a bug, it's a feature. |
| 20:25 | <Dashiva> | There would be feature testing :) |
| 20:25 | <gsnedders> | I mean, my presence here is a feature, not a bug. |
| 20:25 | <gsnedders> | Right? |
| 20:26 | <Dashiva> | It's a bug in your use-every-waking-moment-to-study program |
| 20:26 | <gsnedders> | Dashiva: I know I have physics stuff on my lap, but that's because I'm helping a friend in need! |
| 20:29 | <Dashiva> | Philip`: It seems to be using "some new element like link" is a bad example for a new element :) |
| 20:30 | <Dashiva> | *to me |
| 20:31 | <Philip`> | Dashiva: Who suggested a new element like link? |
| 20:31 | <Dashiva> | In Eric's blog post, last case |
| 20:31 | <Philip`> | Oh, I only saw the one about the attribute like link |
| 20:33 | Philip` | thinks linkable <tr>s seems like the most useful case, since lots of forum systems use that kind of UI (where there's lots of structured information so it's reasonable to use a table, but it ought to be as easy as possible to click on a row to see the corresponding full post) |
| 20:34 | <Dashiva> | I'm not very enthusiastic about linking block elements. I can understand making a <tr> a link, for the same reason I want to make a <tr> be a form. But blocks is just asking for misclicks leading to leaving the page |
| 20:34 | <Dashiva> | And for inline content, <a></a> is only seven characters more. You'd still be using the attributes |
| 20:38 | Philip` | wonders what elements/attributes Googlebot follows |
| 20:57 | gsnedders | has decided how to be profound |
| 20:57 | <gsnedders> | Write poetry based on what people have searched for and found my site! |
| 20:57 | <gsnedders> | I mean, people search for such profound things. |
| 20:59 | <Lachy> | gsnedders, like what? |
| 21:00 | <gsnedders> | Lachy: I dunno. But it's profound. |
| 21:00 | <Lachy> | ok, whatever. |
| 21:01 | <gsnedders> | Lachy: "why must we love today and to be hurt tomorrow" |
| 21:02 | <gsnedders> | There a lot of things about sexuality, too |
| 21:02 | <Lachy> | wow. what exactly do you have on your site that would make it show up for searches like that? |
| 21:02 | <gsnedders> | Lachy: I don't know. |
| 21:03 | <gsnedders> | Lachy: There's quite a lot of extreme stuff from when I was suicidally depressed last year, though, mostly revolving around love |
| 21:04 | <gsnedders> | "oh therapy can you please fill the void" |
| 21:04 | <gsnedders> | oh, wait. that's actually a Green Day lyric that's quoted on my site. |
| 21:04 | <Lachy> | woah, I didn't know you were suicidal. |
| 21:05 | <gsnedders> | Lachy: I'm not anymore (mostly) |
| 21:05 | <Lachy> | good |
| 21:05 | <gsnedders> | Well, like as much as I have been for most of my life up until last May |
| 21:05 | <gsnedders> | "my teacher said yes when i asked her out" :\ |
| 21:05 | <Lachy> | what happened last may? |
| 21:06 | <gsnedders> | Lachy: Stuff |
| 21:06 | <gsnedders> | :) |
| 21:06 | <gsnedders> | (Useless answer, I know) |
| 21:06 | <Lachy> | that's fair enough. |
| 21:06 | <gsnedders> | Only around three people as well as me know the full story. One is a psychologist. |
| 21:07 | <Philip`> | Did the psychologist say yes when you asked him out? |
| 21:07 | <gsnedders> | Philip`: Well, he's married and I haven't asked him. |
| 21:08 | <Philip`> | Oh, okay then |
| 21:08 | <gsnedders> | (Nor do I intend on doing so, might I add) |
| 21:08 | <gsnedders> | Lachy: http://gsnedders.com/about-the-author |
| 21:09 | <gsnedders> | Lachy: If you look through all my posts going back to March last year you might start to build up a picture |
| 21:09 | <gsnedders> | (disclaimer: a lot of it contains a lot of quite impolite language) |
| 21:11 | gsnedders | stares at some more search strings |
| 21:11 | <gsnedders> | "im horny" |
| 21:11 | <gsnedders> | How did _THAT_ find my site? |
| 21:12 | <gsnedders> | "how to be hetrosexual" |
| 21:12 | <gsnedders> | "how to menage a trois" |
| 21:13 | <Philip`> | This is why I disable Referer |
| 21:14 | <gsnedders> | 'fucked "not literally"' |
| 21:15 | <Lachy> | gsnedders, as far as I can tell from that, you're going through the same bad experiences with girls as virtually any other teenager. |
| 21:15 | <gsnedders> | Lachy: That's why I need to properly explain it all |
| 21:16 | <gsnedders> | Well, maybe not it all, but more. |
| 21:16 | <Philip`> | Someone should make a widget thing you can embed in your site, that sends you an IM message every time someone visits your site from a search engine, and if the search terms are interesting then you can reply in real time and your response pops up on the viewer's screen |
| 21:16 | <gsnedders> | Stuff like setting the scene for when it all happened |
| 21:16 | <Lachy> | there was a girl I was in love with back in high school, who I didn't have the guts to ask out, and when the secret finally came out, I was breifly humiliated too. |
| 21:17 | <Lachy> | but, I got over it |
| 21:19 | <gsnedders> | It's ridiculous how many people know who I love now. |
| 21:19 | <gsnedders> | Oh, wow. That's rather profound. "mutability of time" poetry |
| 21:44 | <Hixie> | annevk: what Philip` said (re toDataURL()) |
| 21:46 | <annevk> | ok, we'll go with that |
| 21:47 | <annevk> | clamp or throw? |
| 21:51 | Philip` | imagines people will write quality=85 |
| 21:53 | <Hixie> | clamp |
| 22:04 | <Dashiva> | Lachy: You mean when you let the secret out. ;) |
| 22:06 | <Lachy> | Dashiva, of course. But back then, I blamed all the people I had told. |
| 22:06 | gsnedders | blogs <http://gsnedders.com/love-is-hate>; |
| 22:07 | <gsnedders> | I always intend on keeping such things secret. I fail badly, though, especially because I suck a lying about such things |
| 22:07 | <Dashiva> | And because deep inside, you don't want it to be secret? |
| 22:08 | <Dashiva> | gsnedders: What's the subscripts on she and her about? |
| 22:08 | <Dashiva> | To keep different people apart? |
| 22:08 | <gsnedders> | Dashiva: yeah |
| 22:08 | <gsnedders> | Dashiva: Without using names |
| 22:08 | <Dashiva> | You should call them Fooette and Barette |
| 22:09 | <gsnedders> | Dashiva: But that's only two. I need five, currently. |
| 22:09 | <gsnedders> | (The numbers start in May '06, FWIW) |
| 22:09 | <gsnedders> | (There is no him + subscript number, yet) |
| 22:10 | <Dashiva> | How am I + subscript are there? ;) |
| 22:10 | <Dashiva> | s/am/many |
| 22:11 | <gsnedders> | :P |
| 22:12 | <gsnedders> | Dashiva: That's obvious from who's speaking. Who someone in the third person is is unclear, though, without naming them. |
| 22:17 | <Hixie> | "tbl: concerned about the HTML WG and perhaps we are getting all lost in the weeds" |
| 22:17 | <Hixie> | first step to solving a problem is recognising you have one, i guess |
| 22:30 | <annevk> | tagsoup is the tag's biggest issue |
| 22:30 | <annevk> | i like it |
| 22:34 | <annevk> | http://cafe.elharo.com/category/web/refactoring-html/ |
| 22:34 | <annevk> | via simonw |
| 22:34 | <othermaciej> | they should name that issue "selfsoup" |
| 22:34 | <annevk> | under "Why XHTML" it says "XHTML makes life harder for document authors in exchange for making life easier for document consumers." |
| 22:37 | Philip` | wonders if he can set Apache to do 30x redirects on certain subdirectories in a Subversion repository, to hide those files from users while still allowing them to successfully checkout the parent directory |
| 22:39 | <annevk> | http://www.w3.org/2001/tag/2008/05/20-minutes.html has some bits on RDF and XML |
| 22:43 | <othermaciej> | wow, consuming XHTML is a walk in the park |
| 22:44 | <othermaciej> | I'm glad to learn this new fact |
| 22:44 | <annevk> | besides that it is not really true, shifting the burden from author to implementor is the wrong move |
| 22:45 | <gsnedders> | Well, it's easier in an interpreted language where you have an XML parser built-in but not a HTML one :P |
| 22:45 | <othermaciej> | that's why mobile browsers all only accept XHTML and parse it with a strict XML parser, right? |
| 22:45 | <othermaciej> | because only processing strict XML is so much easier on a lightweight mobile device than arsing tag soup |
| 22:45 | <annevk> | pretty much |
| 22:45 | <othermaciej> | *parsing |
| 22:45 | <othermaciej> | (for those who need it: </sarcasm>) |
| 22:46 | <mpt> | gsnedders, the place to post those is <http://www.disturbingsearchrequests.com/>; |
| 22:49 | <annevk> | also http://www.w3.org/2001/tag/2008/05/21-minutes.html on distributed extensibility |
| 22:59 | <Hixie> | what happened to that guy who was looking at the two-way networking thing |
| 23:00 | <othermaciej> | I don't know |
| 23:00 | <othermaciej> | he seemed to be on a good track |
| 23:01 | <annevk> | someone with a nick of closure asked about that earlier today |
| 23:03 | <annevk> | he was going gog work on http://code.google.com/soc/2008/esos/about.html |
| 23:03 | <annevk> | http://krijnhoetmer.nl/irc-logs/whatwg/20080605#l-298 |
| 23:03 | <othermaciej> | that might be him (can't remember) |
| 23:04 | <annevk> | s/going/doing/ |
| 23:05 | <annevk> | apparently the ES Operating System is implementing HTML5 as well |
| 23:05 | <annevk> | per http://code.google.com/p/es-operating-system/ |
| 23:06 | <Hixie> | well i'm sure the extra 3 users we'll get from that will be great :-P |
| 23:06 | <annevk> | i guess it has some potential, otherwise google wouldn't sponsor it, right? |
| 23:06 | <othermaciej> | Hixie: zomg you're ignoring implementors!!!111 |
| 23:06 | <gsnedders> | What on earth? |
| 23:07 | <gsnedders> | That seems crazy. |
| 23:07 | <Hixie> | othermaciej: hey if they send feedback i'll be all over it :-) |
| 23:07 | <Hixie> | annevk: it's a very interesting research project, for sure |
| 23:07 | <annevk> | maybe we should get gog money for html5lib next year |
| 23:08 | <Hixie> | i can certainly push for you to be accepted if you do a SoC submission |
| 23:08 | <annevk> | s/money/resources/ |
| 23:08 | <Hixie> | i have no idea if the people who make these decisions put any weight on my opinion, though |
| 23:08 | gsnedders | ponders |
| 23:08 | <Hixie> | :-) |
| 23:08 | <gsnedders> | No, I'm not 18 by then |
| 23:08 | <gsnedders> | Which makes any thoughts entirely irrelevant |
| 23:08 | <Hixie> | gsnedders: can you partake in GHOP? |
| 23:09 | <annevk> | a new version of html5lib written in C/C++ with Python bindings :) |
| 23:09 | <gsnedders> | Hixie: If it weren't for the fact I've never even heard of GHOP :) |
| 23:10 | <gsnedders> | Hixie: Yeah, seemingly, if it runs again |
| 23:11 | <gsnedders> | Hixie: Now, if it weren't for the fact that the dollar sucks so badly atm :( |
| 23:12 | <Philip`> | annevk: Should a C version provide all the features that html5lib does, particularly in terms of the varieties of treebuilder? |
| 23:13 | <gsnedders> | Now, g'nite y'all |
| 23:14 | <Hixie> | Philip`: depends if it's a c version exposed to python or a c version exposed to c |
| 23:14 | <Hixie> | Philip`: presumably, it would expose whatever is conventional for that platform |
| 23:17 | <Hixie> | onbeforeprint and onafterprint |
| 23:17 | <Hixie> | opinions? |
| 23:17 | <Hixie> | othermaciej? |
| 23:17 | <annevk> | i haven't seen any requests for them |
| 23:17 | <roc> | i've had requests for them |
| 23:17 | <annevk> | well, a few actually |
| 23:18 | <Hixie> | there are definitely requests |
| 23:18 | <Hixie> | the question is whether @media is a better answer |
| 23:18 | <roc> | seems like anywhere a page has a "print" link, onbeforeprint would be useful ... and there are plenty of those |
| 23:18 | <roc> | I don't think we should expect CSS to provide a total solution always |
| 23:18 | <Hixie> | so i should add them |
| 23:18 | <Hixie> | okie dokie |
| 23:19 | <Hixie> | the next question is how should they work |
| 23:19 | roc | ducks out |
| 23:19 | <othermaciej> | Hixie: yes? |
| 23:19 | <Hixie> | if we expect interop with IE, it seems that they require UAs to synchronously snapshot the DOM and styles |
| 23:19 | <Hixie> | othermaciej: see two lines above where i said your name, to here |
| 23:19 | <othermaciej> | oh the print events? |
| 23:19 | <othermaciej> | I have no idea what they do |
| 23:20 | <annevk> | yeah, and Access control review while you're at it :) |
| 23:20 | <roc> | why do these events require DOM snapshotting? |
| 23:20 | <Hixie> | it appears they fire just before and just after the DOM is snapshotted |
| 23:20 | <Hixie> | roc: window.print() fires the events but isn't synchronous (i.e. it returns straight away and the printing UI is async) |
| 23:20 | <Hixie> | in IE, at least |
| 23:21 | <roc> | in Gecko window.print() is synchronous |
| 23:21 | <roc> | so I don't see that onbeforeprint is introducing a new requirement |
| 23:21 | <Hixie> | hm, fair enough |
| 23:22 | <roc> | except for the requirement that we not crash when onbeforeprint does something insane, but that's normal |
| 23:27 | <annevk> | in Opera print() doesn't work unless invoked through some user generated thingie |
| 23:28 | <roc> | so? That's why onbeforeprint is useful |
| 23:29 | <annevk> | I mean that you need onclick=print() to make it work |
| 23:30 | <annevk> | user initiated event is probably a better word (not sure if event types are restricted) |
| 23:31 | <annevk> | both events are probably useful in case the user uses UI to activate the print dialog |
| 23:33 | <Hixie> | trusted interaction event i think is the most accurate term |
| 23:33 | <Hixie> | since you don't want an author-dispatched click event to do it either |
| 23:49 | <Dashiva> | What. How is debugging JS and CSS easier in XHTML? It's the exact opposite if you want to do the proper XML way... |
| 23:50 | <jgraham_> | Would anyone like to try a test package of html5lib-0.11? http://james.html5.org/temp/html5lib-0.11.zip I think it might be broken in some obscure way but it might just by some sort of icky setuptools residue on my system |
| 23:53 | <Philip`> | jgraham_: It unzips and installs and runs the README example alright for me |
| 23:53 | <jgraham_> | Can you run the tests? |
| 23:53 | <smedero> | same here. os x 10.5.3... running some of the tests now |
| 23:54 | <Philip`> | jgraham_: "python setup.py test" gives 2 errors |
| 23:54 | <jgraham_> | Ah, it is broken then. OK, I'll have to sort that out tomorrow |
| 23:54 | <Philip`> | ERROR: runTest (test_treewalkers.TestCase) ---- TypeError: runTest() takes exactly 6 arguments (1 given) |
| 23:54 | <smedero> | same. |
| 23:54 | <Philip`> | ERROR: test_chardet (test_encoding.Html5EncodingTestCase) ---- IOError: [Errno 2] No such file or directory: './testdata/encoding/chardet/test_big5.txt' |
| 23:54 | <smedero> | (i only get that one error actually) |
| 23:55 | <jgraham_> | Thanks for checking (I got the same errors) |
| 23:56 | <jgraham_> | (but I'm not sure why, especially the runTest() takes exactly 6 arguments one) |