| 00:00 | <and> | I have discovered them gradually, so I never really got shocked. ;) |
| 00:00 | <and> | It is somewhat impressive, though. |
| 00:28 | <annevk2> | dbaron, the link you put in twitter seems broken |
| 00:28 | <dbaron> | annevk2, hmmm... maybe I should stop letting twitter do the URL shortening |
| 00:28 | <dbaron> | of course, the URL fit in 160 characters, but twitter shortened it anyway |
| 00:28 | <dbaron> | I'm not sure what its rule is |
| 00:29 | <annevk2> | oh, it seems to be a bug in twitter's redirect service |
| 00:29 | <dbaron> | anyway, it was http://lists.w3.org/Archives/Public/www-tag/2009Aug/0054.html |
| 00:29 | <annevk2> | ta |
| 00:29 | <dbaron> | the link twitter made works for me, though |
| 00:32 | <and> | annevk2: The table does not seem to match IE8 (apparently installed on Windows Server 2008). |
| 00:33 | <annevk2> | dbaron, if I copy the link it works, I encountered some bug by directly clicking on it from the web interface |
| 00:34 | <annevk2> | dbaron, not sure what's up |
| 00:34 | <annevk2> | and, well that sucks |
| 00:35 | <annevk2> | though presumably the positive side of sucks is progress |
| 00:36 | <annevk2> | (refresh of the twitter interface fixes the issue) |
| 00:49 | <annevk2> | FYI: http://broadcast.oreilly.com/2009/09/draft-html-5-no-longer-a-marku.html |
| 00:50 | <annevk2> | I'm going to bed, but people with some stamina left might want to clarify a thing or two. E.g. how misnested stuff is non-conforming and such... |
| 00:54 | <Hixie> | hsivonen: yt? |
| 01:20 | <Lachy> | Hixie, I contacted ALA about publishing the article about sectioning, and they said they're interested |
| 01:20 | <Hixie> | cool |
| 01:20 | <Lachy> | so I've been working on it yesterday and today. It's coming along nicely, but not yet finished |
| 01:21 | <Lachy> | the only problem is that it's so long already, and it's not even half complete. |
| 01:22 | <Hixie> | i've been poking at the examples, to see if that helps too |
| 01:23 | <Hixie> | just passed r3800 |
| 01:25 | <Lachy> | it's longer than I thought. Already up to 4 pages already |
| 01:25 | <Lachy> | s/already// - not sure why I said that twice |
| 01:35 | <Lachy> | Hixie, in #the-article-element: <p><time pubdate datetime="2009-10-09T14:28-08:00"></time></p> |
| 01:35 | <Lachy> | you should add a human readable date to hat |
| 01:35 | <Lachy> | *that |
| 01:36 | <Lachy> | same with all the other <time> elements in the example |
| 01:38 | <Lachy> | I suppose the example you added is good enough to satisfy this bug of mine too http://www.w3.org/Bugs/Public/show_bug.cgi?id=7552 |
| 01:38 | <Lachy> | I'll close it. |
| 01:39 | <AryehGregor> | annevk2, I left a comment, but it didn't show up. I guess they're premoderated or something? |
| 01:41 | <Hixie> | Lachy: if there's no contents, the UA will render the datetime appropriately |
| 01:43 | <Lachy> | oh, really? I forgot about that |
| 01:46 | <Lachy> | Hixie, from #the-time-element <footer>Published <time pubdate datetime="2009-08-30T07:13Z">lt;/time>.</footer> - s/lt;/</ |
| 01:47 | <Lachy> | that's the last example, just above The code Element |
| 02:06 | <MikeSmith> | Hixie: I saw you mentioned that |
| 02:07 | <MikeSmith> | Hixie: erp, nm |
| 02:07 | <MikeSmith> | was just going to ask you something about the bug list but I figured it out |
| 04:54 | MikeSmith | reads the diff for r3803 and is glad to see that it includes removing an instance of the phrase "the document must be authored such that" from the spec |
| 05:30 | <Hixie> | wow, html4's examples for <bdo> are terrible |
| 05:30 | Hixie | tries to come up with better ones |
| 05:30 | Hixie | fails |
| 06:38 | <hsivonen> | Hixie: yes, async penalizes document.write() |
| 06:38 | <hsivonen> | Hixie: I'm here now. |
| 06:38 | <hsivonen> | and: Sorry. CVSDude shuffled repos around and I forgot to update docs |
| 06:41 | <Hixie> | hsivonen: i changed the xpath stuff |
| 06:41 | hsivonen | looks |
| 06:41 | <Hixie> | i had to actually learn about xpath and find something that would be a compromise between you and johnathan |
| 06:41 | <Hixie> | so going by my track record, i (a) misunderstood xpath, and (b) wrote something neither of you like |
| 06:41 | <hsivonen> | Hixie: compromise in behavior or compromise in wording? |
| 06:42 | <Hixie> | wording was my intent |
| 06:43 | <hsivonen> | Hixie: hmmkay. I *think* that's equivalent. |
| 06:44 | <hsivonen> | Hixie: I can live with that wording, although I think my wording was clearer |
| 06:44 | <Hixie> | once i actually went and read xpath, i found both you and johnathan were proposing wording that didn't use the xpath 1.0 terminology. |
| 06:44 | <hsivonen> | Hixie: well, it's equivalent in WebKit and Opera. In Gecko, it will be equivalent after fixing another bug. |
| 06:44 | <hsivonen> | Hixie: OK. not surprising |
| 06:45 | <Hixie> | (there's no "name expression" in xpath1, and "NameText" or whateven he was using was just a grammar production with no defined meaning as far as i could tell) |
| 06:45 | <hsivonen> | Hixie: thanks |
| 06:45 | <Hixie> | np |
| 06:45 | <Hixie> | hopefully he'll like it too |
| 06:49 | <hsivonen> | and: documentation fixed |
| 06:49 | <hsivonen> | and: and you can get just the htmlparser if you like: http://about.validator.nu/htmlparser/ |
| 06:51 | <boblet> | I have a nav use question: it’s for site and in-page nav, but is it also appropriate for a series of related articles? Example usage: http://boblet.tumblr.com/post/141239118/html5-structure4 |
| 06:51 | <boblet> | (top right box listing HTML5-related articles) |
| 06:52 | <boblet> | I’d guess aside might be better, but this is the only thing on the site I’d like to put in nav |
| 06:53 | <hsivonen> | and: the 7-bit Norwegian/Danish/Swedish/Finnish modifications of ASCII are obsolete and not Web-relevant |
| 06:59 | <boblet> | hrm, on second reading I think this is a “major navigation block”, so nav is appropriate. nicely worded |
| 07:12 | <hsivonen> | Is iTunes LP using WebKit as the renderer? |
| 07:12 | <hsivonen> | If so, are we going to see the scroll bar styling stuff in spec form on www-style? |
| 07:13 | <MikeSmith> | hsivonen: http://www.satine.org/archives/2009/09/09/does-itunes-9-use-webkit/ |
| 07:15 | <hsivonen> | ok. So how are the non-native-looking scroll bars in iTunes LP content implemented? |
| 07:18 | MikeSmith | has no idea |
| 07:19 | <MikeSmith> | Hixie: I'm looking at recentish pubdate attribute changes, and I notice that the current draft doesn't seem to actually specify any constraints on the value of pubdate |
| 07:20 | <Hixie> | doesn't it say it's a boolean attribute? |
| 07:21 | <MikeSmith> | Hixie: erp.. sorry man. I read right past that |
| 07:21 | MikeSmith | needs more tea |
| 07:22 | MikeSmith | inches ever closer to his quota of dumb questions for the week |
| 07:25 | <Hixie> | MikeSmith: :-) |
| 07:25 | Hixie | gives up on <bdo> examples |
| 07:25 | <MikeSmith> | Hixie: I can ping r12a about that. If he has not good examples for it, all hope is lost |
| 07:25 | <MikeSmith> | *no good examples |
| 07:34 | MikeSmith | looks at twitter and sees that r12a is already looking |
| 07:39 | <boblet> | MikeSmith: will you be in Tokyo for Oct 2-3? there’s a html5-dev-jp group event on then that I’m planning to attend… |
| 07:40 | <Hixie> | MikeSmith: yeah, already asked him :-) |
| 07:41 | <MikeSmith> | boblet: yeah, I should be here. I had been asked to go do a talk outside of Japan at that time but had to turn it down. I'm hoping not to have to travel again at all until Novembeer. |
| 07:41 | <MikeSmith> | Hixie: ah, OK |
| 07:45 | <boblet> | MikeSmith: are you planning to attend? http://bit.ly/3CHUhH it’s gonna be presentations on canvas and offline Thurs/Fri night (both on both nights) then hackathon all day Sat. Held at Google |
| 07:47 | <MikeSmith> | boblet: yeah, I should go. I guess I should talk to Oikawa-san. |
| 07:47 | <MikeSmith> | or who's organizing it? |
| 07:50 | <boblet> | MikeSmith: not sure—being done through the html5-dev-ja list and was announced by Shiroishi-san, who’s also one of the presenters (other is Hatano-san) |
| 07:51 | <boblet> | I’ll ask… |
| 08:23 | <hsivonen> | allowing Web authors acquire explicit locks scares me |
| 09:21 | <annevk3> | AryehGregor, left a comment where? |
| 09:59 | <Philip`> | hsivonen: http://ax.itunes.apple.com/htmlResources/CFA7/dt-storefront.cssz uses things like ::-webkit-scrollbar-track:vertical |
| 09:59 | <annevk3> | Hixie, in /TR/REC-xml-names/ you can drop the REC- |
| 10:02 | <hsivonen> | Philip`: well, if iTunes causes the widget styling problem to get solved in a cross-platform and cross-browser way, yay |
| 10:12 | <annevk3> | http://www.html5patch.org/ |
| 10:23 | <Philip`> | http://happyworm.com/blog/2009/08/24/html-5-the-revolution-will-not-be-televised/ - "it supports audio, video, vector based graphics and animation, geolocation and drag and drop" - why does *everyone* think it includes geolocation? |
| 10:23 | <Hixie> | google said it did at google i/o |
| 10:33 | <jgraham> | Because, as much as people might like it to be something else, HTML5 has become the brand name for "cool cutting edge web technologies" in the same way that Ajax, XHTML and DHTML have been in the past |
| 10:33 | <Hixie> | it's rather flaterring |
| 10:33 | <Hixie> | flattering, even |
| 10:34 | <heycam> | wouldn't have happened if it was still called "Web Applications 1.0", i suppose |
| 10:35 | Hixie | is baffled by this "accessibility task force" stuff |
| 10:35 | <annevk2> | hard to say, back then we called it HTML5 too |
| 10:36 | <annevk2> | Hixie, yeah, better becomes a really successful TF to be worth all this debating |
| 10:36 | <Hixie> | why do we need a TF at all? |
| 10:36 | <Hixie> | what is the TF for? |
| 10:37 | <Hixie> | how can you even _have_ a TF for something as fundamental as accessibility? |
| 10:37 | <Hixie> | it'd be like having a task force for performance |
| 10:37 | <Hixie> | or a task force for making sure the spec is implementable in a multiprocess architecture |
| 10:38 | <annevk2> | there's several WGs for accessibility so that certainly seems like something that's possible |
| 10:38 | <Hixie> | the WGs for accessibility have done two things: |
| 10:38 | <hsivonen> | Hixie: as far as I can tell, the purpose of the TF is to put accessibility of HTML5 under WAI in the org chart |
| 10:39 | <hsivonen> | my assessment could be wrong, though |
| 10:39 | <Hixie> | 1. produced self-standing technologies or documentation, separate from anything else |
| 10:39 | <Hixie> | 2. commented on other specs |
| 10:39 | <annevk2> | hsivonen, I think Sam Ruby said on the list it is that regardless of the whether the TF happens |
| 10:39 | <annevk2> | hsivonen, something about the PFWG having to sign off anything we do before we go to LC |
| 10:39 | <Hixie> | so i don't think the TF is comparable to the WGs, since a TF presumably wouldn't do either of those |
| 10:40 | <hsivonen> | annevk2: is the TF a telecon structure then? |
| 10:40 | <hsivonen> | i.e. people signed up for attending telecons? |
| 10:40 | <annevk2> | hsivonen, I think it is the PFWG HTML5-subgroup that wants people from the HTML WG involved |
| 10:41 | <annevk2> | hsivonen, but I might miss something, this has been going on for a while after all |
| 10:41 | <Hixie> | i'm so tired of this "us vs them" nonsense |
| 10:41 | <Hixie> | why can't they just participate in the wg's work like everyone else? |
| 10:42 | <hsivonen> | maybe I'm misunderstanding this all, but I don't like all the "mandate" and process stuff compared to just posting to public-html |
| 10:43 | <jgraham> | I think the accessibility folks think they should have special veto powers above and beyond "normal" WG participants |
| 10:43 | <hsivonen> | jgraham: it sure looks like it |
| 10:43 | <hsivonen> | anyway, lunch time |
| 10:43 | <jgraham> | And they want a process that will ensure that they get that |
| 10:48 | <Hixie> | i guess i'll ask maciej about the accessibility tf tomorrow |
| 10:49 | <Hixie> | and depending on what he says, i'll figure out what to do |
| 10:49 | <Hixie> | participate, try to change the "charter", ignore them, try to get the tf killed, whatever |
| 10:52 | <Hixie> | http://en.wikipedia.org/wiki/MathML#Example is funny |
| 10:53 | <annevk2> | and somewhat sad |
| 10:53 | <Lachy> | Hixie, because the MathML example is 10 times longer than the other languages? |
| 10:53 | <Hixie> | yeah |
| 10:53 | <jgraham> | Funny as in "look how XML turns half a line of readable LaTeX into a screen full of unintelligable XML" |
| 10:54 | <Hixie> | it's also an unfair comparison, the mathml is longer than it needs to be |
| 10:55 | <Lachy> | it's sad how we ended up with an XML based format for maths, instead of simply a <math> element who's contents were interpreted as LaTeX and rendered appropriatley. |
| 10:56 | <jgraham> | Lachy: I argued for that at the time but it is worth noting that "interpreted as LaTeX" isn't a well defined concept |
| 10:56 | <jgraham> | Since TeX is a Turing-complete language |
| 10:56 | <Hixie> | what we ended up with was the least bad of a bad bunch of options |
| 10:57 | <Hixie> | as often has been the case in html5's development |
| 10:57 | <Lachy> | I didn't know it was turing-complete |
| 10:57 | <Hixie> | the continual race to suck the least we can |
| 10:57 | <Hixie> | "HTML5: It could suck more" |
| 10:57 | <Lachy> | I just know a lot of mathematicians actually use latex and it would have made sense to let them use it directly on the web |
| 10:58 | <jgraham> | Yes |
| 10:59 | <Philip`> | If you want to base it on usage, it would probably be better to use Microsoft Office's textual equation format |
| 10:59 | <Hixie> | which exports to mathml |
| 10:59 | <Hixie> | which is one reason we used mathml |
| 10:59 | <jgraham> | We could have defined a subset of LaTeX and some magic to turn it into DOM. I admit this would not have been as conceptually simple as using MathML directly |
| 11:00 | <jgraham> | But it would have been way easier to author |
| 11:00 | <Hixie> | way easier to hand-author |
| 11:01 | <Philip`> | jgraham: If you write a document in e.g. Word's equation editor, which has some textual format and is MathML internally, how would you export that to HTML when HTML requires a crazy syntax like LaTeX? |
| 11:02 | <Philip`> | Parsing LaTeX is only half the problem |
| 11:03 | <jgraham> | It's not obvious to me that going from MathML->LaTeX subset is a hard problem |
| 11:03 | <jgraham> | especially if you design the subset with that in mind |
| 11:04 | <Hixie> | ok if anyone is bored they can come up with examples for good tables that show off all html's table elements and attributes |
| 11:04 | <Hixie> | in the meantime i'm gonna sleep |
| 11:04 | <Hixie> | later |
| 11:05 | <Philip`> | What if you e.g. make one of the variables in your equation red, using CSS? |
| 11:06 | <Philip`> | (in the MathML representation) |
| 11:06 | <jgraham> | Yeah that is a real problem |
| 11:06 | <jgraham> | ALthough a rather uncommon one |
| 11:35 | <jgraham> | To whom it may concern: a manual shouldn't read like a sales pitch |
| 11:41 | Lachy | wonders what sales pitch jgraham is reading |
| 12:18 | <Dashiva> | So now there are complaints that HTML5 removes deprecated attributes from HTML4... isn't that the whole point of deprecation? |
| 12:35 | <erlehmann> | Dashiva, where? |
| 12:36 | <Dashiva> | http://lists.w3.org/Archives/Public/public-html/2009Sep/0450.html |
| 14:00 | <hsivonen> | tree builder flushes a text node before it is complete. a timeout appends another text node as the next sibling of the parser-inserted text node. Then the tree builder flushes the rest of the text. |
| 14:00 | <Lachy> | http://html5doctor.com/the-section-element/ |
| 14:00 | <hsivonen> | Which text node does the later tree builder flush go to? |
| 14:04 | <hsivonen> | Mercurial Queues are great. How do svn and cvs users manage without them? |
| 14:05 | <TabAtkins> | Hixie: I had problems with bidirectional text a while back that I had to solve by explicitly inserting directionality characters. Lemme dig up what the hell it was (it's obsolete now, so the example itself may be gone), so I can give you something for <bdo>. |
| 14:13 | <TabAtkins> | Hixie: Ah, I remember now. No clue why I was doing this, but I was editting a Hebrew text and adding inline English annotations. Very annoying to deal with, and I didn't know <bdo> existed to help me. |
| 14:13 | <jgraham> | hsivonen: Can you give me a really short summary of mq? I haven;t quite worked out if I should be using it all the time or if it is something that I can live without |
| 14:14 | <hsivonen> | jgraham: I regret postponing my starting of using it |
| 14:14 | <hsivonen> | jgraham: basically, you have a pushable and poppable queue of malleable hg revs |
| 14:14 | <hsivonen> | jgraham: that you can revise or pop |
| 14:15 | <hsivonen> | jgraham: then you can freeze the bottommost one to a real rev that you push when you get review |
| 14:15 | <hsivonen> | jgraham: so if your bottommost patch is being reviewed, you can already build on it but go back to address review comments |
| 14:18 | <jgraham> | Interesting. Are there also advantages in review-free environments? |
| 14:19 | <hsivonen> | jgraham: I'm not aware of benefits in environments where you can commit a real rev before you start working on your next bug |
| 14:20 | <hsivonen> | jgraham: maybe hg 1.3 makes rebasing easier if you are rebasing an mq patch. dunno |
| 14:26 | <jgraham> | hsivonen: Thnaks |
| 14:26 | <jgraham> | *Thanks |
| 14:41 | <jgraham> | http://www.marcozehe.de/2009/09/11/youre-a-table-and-i-dont-care-what-lies-underneath/ seems interesting if anyone is bored |
| 14:42 | <jgraham> | Also jQuery is licensed under MIT /and/ GPL WTF? |
| 14:43 | <hsivonen> | jgraham: I raised my eyebrows, too, when I first saw the jQuery license |
| 14:48 | <erlehmann> | i dont get that opera does <video> but not <audio> |
| 14:48 | annevk2 | finally realizes what AryehGregor commented on |
| 14:48 | <annevk2> | erlehmann, I thought it was the other way around |
| 14:49 | <erlehmann> | annevk2, how come ? |
| 14:49 | <erlehmann> | opera had a build that did theora video. i believe they were one of the first. |
| 14:49 | <annevk2> | erlehmann, because we haven't shipped our actual impl yet? |
| 14:49 | <annevk2> | erlehmann, sure, we demonstrated it could be done |
| 14:49 | <annevk2> | erlehmann, the actual API is vastly more complicated though |
| 14:50 | <erlehmann> | annevk2, where does the Audio() object come from that does only WAV ? was it in spec some time in the past ? |
| 14:51 | <annevk2> | it was there quite a while before <video> |
| 14:51 | <annevk2> | also, <audio> is compatible with it |
| 14:51 | <annevk2> | it has a similar Audio() constructor |
| 15:05 | <takkaria> | jgraham: it's not licenced under MIT and GPL, it's licenced under either MIT or GPL, which is slightly different |
| 15:05 | <takkaria> | and not quite as insane |
| 15:06 | <jgraham> | takkaria: In what way is it not insane? |
| 15:07 | <takkaria> | oh, it's insane |
| 15:08 | <takkaria> | just not as insane as trying to have two contradictory licences at once |
| 15:10 | <hsivonen> | is the purpose to waive the MIT notice when someone is using it under the GPL? |
| 15:59 | <[wito]> | So I'm sitting all up in #html, just minding my own business |
| 15:59 | <[wito]> | And suddenly, this guy comes in, asking about Client Side SQL storage in HTML5 |
| 15:59 | <[wito]> | So I think to myself, "that can't be right, can it?" |
| 15:59 | <[wito]> | And so I do a little checking around |
| 15:59 | <Lachy> | [wito], he must have been talking about Web Databases |
| 15:59 | <[wito]> | Whatever thing Webkit has gotten to implementing |
| 16:00 | <[wito]> | so yeah, probably, whatever |
| 16:00 | <Lachy> | it was formerly part of HTML5, till it got split out |
| 16:00 | <[wito]> | I see... |
| 16:00 | <[wito]> | Well, that's good. |
| 16:00 | <Lachy> | http://dev.w3.org/html5/webdatabase/ |
| 16:00 | <annevk2> | and everything new, including CSS3, is talked about as some HTML5 thing these days |
| 16:00 | <[wito]> | Now let me just make my god damned point. |
| 16:01 | <Rik|work> | annevk2: becoming a buzzword, like ajax |
| 16:01 | <annevk2> | you think? |
| 16:01 | <gsnedders> | Rik|work: "becoming"? |
| 16:01 | <hsivonen> | s/some thing/family of technologies/ |
| 16:01 | <[wito]> | HTML5, in case you haven't noted, is bloated |
| 16:01 | <[wito]> | and silly |
| 16:01 | <annevk2> | that's a great point |
| 16:01 | <[wito]> | and all up in #html we've had it |
| 16:01 | <annevk2> | sorry about the sarcasm, ehm, what's your argument? |
| 16:02 | <annevk2> | "bloated" and "silly" is hardly constructive |
| 16:02 | <[wito]> | so we're hereby giving notice of intent to redirect all questions regarding HTML 5 to this channel, no matter how inane they are |
| 16:02 | <[wito]> | just a heads up |
| 16:02 | <annevk2> | and sound somewhat ill-informed |
| 16:02 | <AryehGregor> | Go ahead. |
| 16:02 | <Lachy> | what things are in HTML5 that you think shouldn't be there? |
| 16:02 | <annevk2> | [wito], that's certainly cool |
| 16:02 | <Lachy> | [wito], great! |
| 16:02 | <[wito]> | annevk2: I'll concede that "silly" isn't the best word |
| 16:02 | <jgraham> | Awesome. It sounds like you are just the sort of people I would liek to be directed away from |
| 16:02 | <hsivonen> | it could suck more :-) |
| 16:02 | <erlehmann> | [wito], if with bloat you mean the space between the L and the 5, i can agree |
| 16:02 | <Lachy> | it would help if you could elaborate on why you think it's silly |
| 16:02 | <gsnedders> | It could be called XHTML 2. |
| 16:03 | <AryehGregor> | The people who idle here are the best situated to answer HTML 5 questions, so it would be good to have the questions come here instead of #html. |
| 16:03 | <[wito]> | erlehmann: It's HTML5? |
| 16:03 | <[wito]> | No space? |
| 16:03 | <AryehGregor> | erlehmann, that's being fixed, isn't it? |
| 16:03 | <annevk2> | yes |
| 16:03 | <[wito]> | oh dear... |
| 16:03 | <annevk2> | [wito], http://blog.whatwg.org/spelling-html5 |
| 16:03 | <Lachy> | http://blog.whatwg.org/spelling-html5 |
| 16:03 | <erlehmann> | ahaha |
| 16:03 | <beowulf> | some of us are best situated to ask silly and inane questions, too |
| 16:03 | <[wito]> | Well |
| 16:03 | <[wito]> | If there are no protests |
| 16:03 | <AryehGregor> | The <title> of current-work at whatwg.org still seems to say "HTML 5". |
| 16:03 | <[wito]> | We'll just start shoveling them over by the cartload. |
| 16:03 | <annevk2> | [wito], http://annevankesteren.nl/2007/04/html-red-pill ;) |
| 16:03 | <Lachy> | oops, we should fix that |
| 16:03 | <AryehGregor> | [wito], only hearty agreement. |
| 16:04 | <jgraham> | Silly and inane questions are some of the best. For example: party food; jelly or cheese on sticks? |
| 16:04 | <Lachy> | AryehGregor, no it doesn't. It's "HTML5" |
| 16:04 | <beowulf> | jgraham: cheese on sticks |
| 16:04 | <jgraham> | beowulf: Seriously> |
| 16:04 | <beowulf> | jelly-- |
| 16:04 | <gsnedders> | jgraham: Jelly. We concluded that a few weeks ago, with ice-cream. |
| 16:04 | <[wito]> | annevk2: Very nice; but "no versioning" isn't a good thing |
| 16:04 | <da3d> | jgraham: cheese in jelly |
| 16:04 | <beowulf> | jelly with ice-cream is an abomination |
| 16:04 | <AryehGregor> | Lachy, <title>5 Microdata — HTML 5</title> on http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html#microdata |
| 16:04 | <gsnedders> | [wito]: Now, how is it bloated and silly? |
| 16:04 | <jgraham> | Jelly on sticks? |
| 16:05 | <annevk2> | [wito], well see, I think it is |
| 16:05 | <AryehGregor> | The TOC has HTML5. |
| 16:05 | <[wito]> | annevk2: Then we will have to agree to disagree |
| 16:05 | <hsivonen> | Hixie: the spec splitter has the old title with space |
| 16:05 | <jgraham> | [wito]: Not necessarily. You could have a fight to the death |
| 16:05 | <AryehGregor> | [wito], have you actually read any of the many, many pages of discussion on whatwg and public-html about the versioning question? |
| 16:05 | <[wito]> | AryehGregor: I must admit I have not. |
| 16:05 | <annevk2> | [wito], I'd love to hear your arguments |
| 16:05 | <Lachy> | AryehGregor, oh, that must be a problem with the splitter script |
| 16:05 | <Lachy> | I think that's Philip`s thing |
| 16:05 | <erlehmann> | RTFML then try again |
| 16:06 | <AryehGregor> | [wito], then maybe it would be more appropriate to say you can't see why it's a good thing, or somesuch. |
| 16:06 | <erlehmann> | AryehGregor, just do a s/HTML\ 5/HTML5/g ? |
| 16:06 | <hsivonen> | Philip`: the spec splitter has the old title |
| 16:06 | <beowulf> | [wito]: what's #html going to be answering questions on now? |
| 16:06 | <jgraham> | Maybe about party food? |
| 16:06 | <[wito]> | beowulf: XHTML 1.1, XHTML 1.0 and HTML 4.01 |
| 16:06 | <dbaron> | Hixie, I could come up with some decent uses of <bdo> in describing Chinese signs to English speakers, but I can't find any such examples on the Web... |
| 16:06 | <jgraham> | I think party food would have been better |
| 16:07 | <[wito]> | the "versioned, strictly defined shit with few tags" as you seem to call it |
| 16:07 | <[wito]> | anyway |
| 16:07 | <[wito]> | I'll write a blog entry about the bloat in HTML5 or something, let me get back to you on that. |
| 16:07 | <Lachy> | [wito], ok, I look forward to reading it |
| 16:07 | <beowulf> | me too |
| 16:07 | <annevk2> | [wito], please include the pro-versioning arguments |
| 16:07 | <[wito]> | I just found the idea of Web databases completely and utterly not at all relevant to HTML in any way. |
| 16:07 | <AryehGregor> | [wito], k. |
| 16:07 | <[wito]> | And what have you |
| 16:07 | <Lachy> | [wito], where's your blog? |
| 16:07 | <AryehGregor> | It's not part of the HTML spec. |
| 16:07 | <[wito]> | annevk2: 10-4 |
| 16:08 | <AryehGregor> | At least, not anymore. |
| 16:08 | <[wito]> | Lachy: I'll also have to get me one of those. |
| 16:08 | <annevk2> | [wito], sorry? |
| 16:08 | <Lachy> | hah |
| 16:08 | <AryehGregor> | It's part of a JavaScript API. |
| 16:08 | <annevk2> | [wito], -6? |
| 16:08 | <beowulf> | lol |
| 16:08 | <AryehGregor> | HTML5 does define some JavaScript interfaces in addition to the actual HTML. |
| 16:08 | <Lachy> | annevk2, 10-4 is 6, not -6! |
| 16:08 | <[wito]> | ... |
| 16:08 | annevk2 | was doing rtl |
| 16:08 | <[wito]> | annevk2: ten-four |
| 16:08 | <beowulf> | (ok, i woke the dog laughing at annevk2's response there) |
| 16:09 | <[wito]> | as in, roger |
| 16:09 | <jgraham> | Hey, we could add an <sql> tga that did queries against HTML tables! |
| 16:09 | <AryehGregor> | annevk2, RTL usually does Arabic numerals LTR, sorry. :P |
| 16:09 | <AryehGregor> | jgraham, LOL |
| 16:09 | <[wito]> | Well, I don't have a blog, so I'll just slap up an article |
| 16:09 | <jgraham> | AryehGregor: :) |
| 16:09 | <[wito]> | expect it ~sunday, I'll get back to y'all with the URL |
| 16:09 | <Lachy> | [wito], Ten-codes don't work too well in written communication |
| 16:09 | <[wito]> | Anyway, I have some HTML questions to go asnwer |
| 16:09 | <AryehGregor> | Is there a page that addresses common objections to HTML5 in the vein of "The way we used to do it is better"? |
| 16:10 | <Lachy> | [wito], is #html a channel on freenode where you're getting all these questions? |
| 16:10 | <[wito]> | Occasionally. |
| 16:10 | Lachy | joins to see if any HTML5 questions pop up |
| 16:10 | <AryehGregor> | [wito], you do realize that none of the HTML versions you think should be handled in #html are actually going to survive for much longer? Given that HTML5 is the only HTML version that actually has any resources being put into anymore? |
| 16:10 | <AryehGregor> | I mean, all other HTML specs are obsolescent. |
| 16:10 | <hsivonen> | on versioning, please take another look at http://hsivonen.iki.fi/doctype/ie8-mode.png before wanting versioning |
| 16:10 | <[wito]> | AryehGregor: obsolescent |
| 16:10 | <[wito]> | that's an awsome word |
| 16:11 | <[wito]> | But I disagree |
| 16:11 | <AryehGregor> | A handy one too. |
| 16:11 | <[wito]> | and XHTML 1.1 parsers are stable |
| 16:11 | <AryehGregor> | So are HTML 1 parsers, I'm sure. |
| 16:11 | <[wito]> | and will continue to benefit from improvements to SVG rendering engines |
| 16:11 | <annevk2> | actually, XML 1.0 5Th edition just changed them |
| 16:11 | <annevk2> | though fixes might not have been deployed to your favorite parser library just yet |
| 16:11 | <Lachy> | AryehGregor, there was no HTML 1 |
| 16:12 | <AryehGregor> | Rats. |
| 16:12 | <AryehGregor> | Foiled again. |
| 16:12 | <beowulf> | it was called Tags, no? |
| 16:12 | <[wito]> | and that's the glory of versioning |
| 16:12 | <AryehGregor> | Well, I can fairly say that all HTML 1 parsers are stable, then. |
| 16:12 | <Lachy> | there's an informal language that was developed prior to HTML 2.0 being published as RFC 1866 in 1995 |
| 16:12 | <[wito]> | basically, when you've got a version down; you freeze development, branch and have at the next version |
| 16:12 | <annevk2> | [wito], XML 1.0 has no subversioning |
| 16:12 | <AryehGregor> | (they're also all unstable, and made out of Swiss cheese, and so forth) |
| 16:12 | <annevk2> | [wito], they changed parsing rules as "errata" |
| 16:13 | <annevk2> | [wito], it was quite controversial |
| 16:13 | <AryehGregor> | [wito], or you could just make sure everything is backwards-compatible and not need versioning. |
| 16:13 | <[wito]> | annevk2: I'm sorry, whatnow? |
| 16:13 | <annevk2> | [wito], see w3.org/tr/xml |
| 16:13 | <[wito]> | AryehGregor: Except for the fact that this will keep you back later |
| 16:13 | <AryehGregor> | Which is what HTML5 does. |
| 16:13 | <[wito]> | backwards-compatible is a dream |
| 16:13 | <erlehmann> | [wito], how will it ? |
| 16:13 | <AryehGregor> | [wito], it will keep you back less than pretending you don't need to be completely backward compatible, which is what XHTML 2 did. |
| 16:13 | <[wito]> | and in *my* programming experience, it's not always a good thing |
| 16:13 | <annevk2> | much like versioning on the Web |
| 16:13 | <[wito]> | sometimes, you just need to go forward |
| 16:14 | <AryehGregor> | [wito], it's not a good thing. It's inevitable on the web, however. |
| 16:14 | <jgraham> | backwards compatible is a necessary reaslity for the web |
| 16:14 | <jgraham> | *reality |
| 16:14 | <AryehGregor> | You can't fix up all the content out there. |
| 16:14 | <erlehmann> | use the XML serialization, use namespaces. voila. |
| 16:14 | <AryehGregor> | That's just fact. |
| 16:14 | <[wito]> | AryehGregor: Which is where versioning comes into play |
| 16:14 | <AryehGregor> | Except then users can't use the new features without rewriting their site. |
| 16:14 | <AryehGregor> | Which few will do. |
| 16:14 | <jgraham> | AryehGregor: There is only one browser vendor in the world that is prepared to ship a verioned engine |
| 16:15 | <jgraham> | *versioned |
| 16:15 | <[wito]> | AryehGregor: "Oh no! I don't get nuthin' for nuthin'!" |
| 16:15 | <annevk2> | jgraham, and even they hate it |
| 16:15 | <Steve^> | If you look at Python, they are breaking backwards-compatibility |
| 16:15 | <jgraham> | And hsivonen just posted you an informative link as to why that is a clusterfuck |
| 16:15 | <Steve^> | everyone is free to use the old version if they want |
| 16:15 | <[wito]> | anyway |
| 16:15 | <jgraham> | Steve^: Not comparable |
| 16:15 | <[wito]> | I didn't come here to discuss versioning |
| 16:15 | <AryehGregor> | [wito], you can be contemptuous all you like. If you think versioning is a valid substitute for backwards compatibility, you should have convinced the major browser vendors of it when you still had a chance. Too late now. |
| 16:16 | <jgraham> | I can't follow a link from a python 2.5 program to a python 2.6 program |
| 16:16 | <AryehGregor> | XHTML 2.0 didn't work, for whatever reason. HTML5 is working. |
| 16:16 | <[wito]> | I came here to say that sql databases is too much even for HTML5 |
| 16:16 | <[wito]> | and it seems I was right |
| 16:16 | <AryehGregor> | SQL databases are part of a different spec, technically. |
| 16:16 | <beowulf> | [wito]: evidence? |
| 16:16 | <[wito]> | AryehGregor: which is my point |
| 16:16 | <AryehGregor> | HTML5 does define a bunch of JavaScript APIs, but mostly they're closely linked to the elements it defines. |
| 16:16 | <[wito]> | beowulf: sql databases aren't part of the HTML5 specc |
| 16:17 | <pengo> | client side database? seriously? |
| 16:17 | <[wito]> | I remain confident that XHTML 1.1 will continue to work for me for decades to come |
| 16:17 | <[wito]> | so yeah |
| 16:17 | <Steve^> | jgraham, a 2.5 can execute a 2.6 one though. Its not quite the same as HTML isn't compiled/interpreted in the same way. Newer browsers can simply ignore deprecated tags |
| 16:17 | <[wito]> | have fun |
| 16:17 | <AryehGregor> | Steve^, not if they want the web to work. |
| 16:17 | <beowulf> | [wito]: no, what's your evidence that's it's too much, i'd like to see it all in one 'document' actually |
| 16:17 | <AryehGregor> | Or if they want to be HTML5-compliant. |
| 16:17 | <AryehGregor> | pengo, sure. It seems like a good idea to me. Mozilla apparently doesn't want to implement it, though. |
| 16:18 | <AryehGregor> | At least not now. |
| 16:18 | <beowulf> | i think spec splitting is unhelpful, but, what do i know |
| 16:18 | <AryehGregor> | (I don't know, I haven't paid much attention to it) |
| 16:18 | <[wito]> | beowulf: Sure, let me sit down tomorrow and write a comprehensive article detailing my point of view on HTML5 |
| 16:18 | <Rik|work> | [wito]: XHTML1.1 ? is it working for you on IE ? |
| 16:18 | <beowulf> | [wito]: ok |
| 16:18 | <[wito]> | I'll publish it on early sunday (GMT+2) and post the URL here |
| 16:18 | <[wito]> | Rik|work: Hint: I don't care. |
| 16:18 | <annevk2> | [wito], the part of XHTML 1.1 that's a subset of HTML5 should remain to work fine |
| 16:18 | <AryehGregor> | [wito], that sets you apart from those of us who actually want our websites to be useful, I guess. |
| 16:18 | <annevk2> | [wito], pretty much like it does now, indeed |
| 16:18 | <[wito]> | right now it's friday, and I'm working my way through a few beers |
| 16:18 | <Rik|work> | [wito]: lucky you |
| 16:19 | <[wito]> | AryehGregor: http://totlandweb.info/reiko |
| 16:19 | <[wito]> | I don't need useful. |
| 16:19 | <[wito]> | I've got AWSU |
| 16:19 | <[wito]> | M |
| 16:20 | <Steve^> | I have a question about the correct use of svg. It only works when the page is sent as xml? |
| 16:20 | <AryehGregor> | Steve^, no. |
| 16:20 | <AryehGregor> | It works in HTML serialization as well. |
| 16:20 | <[wito]> | except that it shouldn't. |
| 16:20 | <[wito]> | That's another thing |
| 16:21 | <[wito]> | which I'll detail in my article |
| 16:21 | <Steve^> | AryehGregor, I need to change the markup for that? |
| 16:21 | <jgraham> | Steve^: (assuming an HTML5 parser) |
| 16:21 | <AryehGregor> | [wito], you realize we've all heard these opinions a hundred times, I hope? |
| 16:21 | <jgraham> | (which only FIrefox nightlies have) |
| 16:21 | <[wito]> | Oh, yeah |
| 16:21 | <AryehGregor> | I'm going to guess you're not going to make a single point in your article that we haven't all considered. But go ahead, I'm sure some of us will respond. |
| 16:21 | <AryehGregor> | Steve^, no, you don't in theory. |
| 16:21 | <AryehGregor> | However, of course most browsers don't yet support <svg>. |
| 16:22 | <Steve^> | does Opera 10? |
| 16:22 | <jgraham> | [wito]: I love your democratic notion that SVG "shouldn't" work in HTML. |
| 16:22 | <[wito]> | AryehGregor: wut |
| 16:22 | <AryehGregor> | Steve^, no. (AFAIK) |
| 16:22 | <Steve^> | Oh |
| 16:22 | <jgraham> | Steve^: No, sadly |
| 16:22 | <Lachy> | [wito], there was a lot of pressure for us to add support for SVG in HTML |
| 16:22 | <annevk2> | AryehGregor, seems your comment got passed validation btw |
| 16:22 | <Steve^> | It works when the file extension is .xml, rather than .htm |
| 16:22 | <[wito]> | Lachy: And you went ahead like populistic pigs and pissed over the fact that svg is mother fucking XML |
| 16:22 | <[wito]> | I'm sorry, that was a bit harsh |
| 16:22 | <Lachy> | [wito], in fact, there's still a lot of pressure for us to go further with regards to what we support and how |
| 16:22 | <[wito]> | s/pigs/dogs/ |
| 16:22 | <Lachy> | woah, calm down |
| 16:23 | <jgraham> | Steve^: Yeah in Opera if you serve well formed XML then the SVG will render |
| 16:23 | <[wito]> | Lachy: I calm down in much the same fashion as Samuel L. Jackson does: Not. |
| 16:23 | <[wito]> | As is my god-given right. |
| 16:23 | <[wito]> | Or somesuch |
| 16:23 | <jgraham> | If you serve common or garden HTML (i.e. anything with a text/html mime type) then it won't |
| 16:23 | <Steve^> | jgraham, I was under the impression Opera had implemented svg properly. Since it hasn't, everything is fine, I can wait |
| 16:24 | <jgraham> | This doesn't have anything to do with our SVG implementation |
| 16:24 | <jgraham> | It's a HTML parser issue |
| 16:24 | <Lachy> | [wito], luckily, we have a right to ignore trolls too. |
| 16:24 | <[wito]> | Oh, yes |
| 16:24 | <[wito]> | Indeed. |
| 16:24 | <annevk2> | [wito], maybe you should get one of those tasty burgers |
| 16:24 | <[wito]> | annevk2: Maybe I should. |
| 16:24 | <Steve^> | Yea thats fine |
| 16:25 | <[wito]> | That, or get laid. |
| 16:25 | <[wito]> | I really need to get laid. |
| 16:25 | <jgraham> | We haven't updated our HTML parser to recognise SVG and put it in the rioght namespace so that our SVG code kicks in at the right time |
| 16:25 | <da3d> | I figure inline MathML is a lot more important than inline SVG in HTML, actually... |
| 16:25 | <[wito]> | Lachy: Unfortunately, I was called out on being [undemocratic] |
| 16:26 | beowulf | passes round the cheese on sticks |
| 16:26 | <jgraham> | I want jelly dammit |
| 16:26 | <[wito]> | and ignoring people outright is also rather undemocratic |
| 16:26 | <[wito]> | do, you know. |
| 16:26 | <Steve^> | I know where I've confused myself. Opera and friends support .svg but not the HTML5 version :) |
| 16:26 | <jgraham> | Steve^: Right |
| 16:27 | <Dashiva> | Although I still don't know why anyone would put SVG in <html> instead of in .svg |
| 16:27 | <[wito]> | Also, I don't mean to troll, I'm merely stating my (admittedly low) opion of HTML5 |
| 16:27 | <[wito]> | and telling you that we will forward questions much more agressively. |
| 16:27 | <[wito]> | So, um. |
| 16:27 | <[wito]> | Yeah |
| 16:28 | <[wito]> | Cool cakes. |
| 16:28 | <Steve^> | Dashiva, so it can be generated on the fly by PHP, etc? |
| 16:28 | <jgraham> | [wito]: The thing about opinions is that everyone has them. Therefore they're not really very unique or interesting unless they are backed up with reasoned argument |
| 16:28 | <[wito]> | True |
| 16:28 | <jgraham> | Which so far yours haven't been |
| 16:28 | <[wito]> | but while I would *like* to present a coherent and reasoned argument; it's just too late in the week for me for that kind of thing. |
| 16:28 | <jgraham> | So I guess you should write that article or something |
| 16:28 | <Dashiva> | Steve^: People have been generating images on the fly a long time, so it seems to work fine as it is |
| 16:28 | <[wito]> | That's the plan· |
| 16:28 | <AryehGregor> | [wito], please do forward people here. Obviously we don't want them getting your view of HTML5. :) |
| 16:29 | <[wito]> | <below_belt>Then maybe you should stop making it suck. ;) |
| 16:29 | <Steve^> | Dashiva, but you then need to save it somewhere, keep track of it, clean it up. Why not send it once and forget? |
| 16:29 | <[wito]> | or is the <burn> tag more appropriate? |
| 16:29 | <[wito]> | I can never remember. |
| 16:29 | <[wito]> | </below_belt> |
| 16:30 | <Dashiva> | Steve^: Besides, it seems that use case would be served just as well by a data URL |
| 16:30 | <annevk2> | honestly, I'm glad there's still people that are violently opposed |
| 16:30 | <annevk2> | otherwise we'd be doing something wrong |
| 16:30 | <[wito]> | annevk2: You know what we should do? |
| 16:30 | <TabAtkins> | Dashiva: Yeah, a data url *would* work well. And wouldn't give us any lip about namespaces and such in HTML. |
| 16:30 | <[wito]> | Get some boxing gloves on, duke it out, man to (wo?)man |
| 16:31 | <[wito]> | trust me |
| 16:31 | <[wito]> | it'd be hot. |
| 16:31 | <TabAtkins> | Anne's a guy. |
| 16:31 | <Steve^> | Dashiva, embed the svg data in get variables on an image request? Has its limits |
| 16:31 | <Dashiva> | get variables? |
| 16:31 | <[wito]> | we could sell tickets to that shindig |
| 16:31 | <Dashiva> | Why would you be using those? |
| 16:31 | <Steve^> | you know, ?foo=bar&so=on |
| 16:31 | <Steve^> | Ok, what do you mean? |
| 16:31 | <Dashiva> | data:application/svg+xml,nonsensegoeshere |
| 16:32 | <annevk2> | Steve^, fwiw, all browsers support generation of SVG on the fly through JavaScript/DOM |
| 16:32 | <annevk2> | Steve^, also in HTML |
| 16:32 | <Steve^> | Dashiva, where does that go? |
| 16:32 | <Dashiva> | In @src |
| 16:32 | <[wito]> | anyway, I'll get back to you with an article |
| 16:32 | <[wito]> | see you all on sunday |
| 16:32 | <[wito]> | [wito] out |
| 16:32 | <Steve^> | have a nice day |
| 16:33 | <annevk2> | that was amusing |
| 16:33 | <Steve^> | if you don't have crazy people shouting at you, you're doing it wrong |
| 16:33 | <TabAtkins> | Fun diversions. At least the guy recognized when he jumped from crazy arguing to trolling. |
| 16:33 | <Dashiva> | But surely that's not the only reason for inline svg |
| 16:34 | <Steve^> | Ok, didn't realise you could do that. |
| 16:34 | <Steve^> | Readability? |
| 16:34 | <TabAtkins> | SVG isn't very readable once you go beyond trivial hand-authored examples. ^_^ |
| 16:34 | <TabAtkins> | Though that may be the fault of the tools. |
| 16:35 | <jgraham> | Dashiva: It makes sense for cases where you have a lot of SVG and HTML all together and don't want strict parsing, no? |
| 16:35 | <Philip`> | jgraham: In review-free environments, it lets you keep a clean repository history by e.g. adding a feature in one patch, then adding another feature on top of it in another patch, then going back and editing the first one, and then committing everything without any of the back-and-forth mess |
| 16:35 | <Philip`> | hsivonen: Hmm, yes, I guess I should fix the spec-splitter :-( |
| 16:35 | <TabAtkins> | I think it makes sense to let SVG stay strict in its parsing, because it's already established as being so. |
| 16:35 | <jgraham> | Philip`: Yeah. I was reading about it a bit more. I guess I should try it out |
| 16:36 | <Dashiva> | jgraham: Lots of dynamic SVG and non-strict HTML? |
| 16:36 | <Philip`> | (Also probably ought to fix it to split into more even-sized sections too, since Hixie irritatingly has done work on the spec and it's all different now) |
| 16:36 | <jgraham> | TabAtkins: I disagree because I think that strict parsing is a hinderance to SVG |
| 16:36 | <jgraham> | I wouldn't deploy a mainly-dynamic-svg interface to amazon.com with strict parsing |
| 16:37 | <Dashiva> | Wouldn't a mostly-dynamic-svg interface be done in JS? |
| 16:37 | <TabAtkins> | Shrug. I still like the idea of XML strictness, even if it's totally unworkable on the public web. |
| 16:37 | <jgraham> | Dashiva: "dynamic" in the sense of "generated by server-side scripts" |
| 16:37 | <Steve^> | I think XSLT taught me that XML isn't always the way to go |
| 16:38 | <TabAtkins> | I think XSLT just teaches that XSLT isn't the way to go. Like, ever. |
| 16:38 | <beowulf> | Steve^: XSLT taught me that XSisn't that way to go |
| 16:38 | <Philip`> | jgraham: "an <sql> tga that did queries against HTML tables!" - if I remember correctly, tantek talked about adding something for SQL queries on tables and <csv>s recently, though maybe I'm just imagining that |
| 16:38 | <Dashiva> | Philip`: Sure that wasn't Dmitry Turin? |
| 16:38 | <beowulf> | i lost an LT\s in that sentence |
| 16:39 | <TabAtkins> | Philip`: Yeah, tantek was free-associating with treating tables like SQL. |
| 16:40 | <TabAtkins> | I'm pretty sure he meant it as a javascript api, though. |
| 16:41 | <Steve^> | Due to bandwidth, I download the HTML5 spec as a PDF. In two weeks, how out of date would it be? |
| 16:41 | <TabAtkins> | Philip`: Oh, and we settled on just embedding csv into a <pre> with a @separator attribute being the flag for "treat this as csv". |
| 16:41 | <jgraham> | But with an <svg> tag users of legacy browsers would be able to see the SQL so they could run the queries manually, giving them equivalent functionaility albeit with reduced performance |
| 16:41 | <TabAtkins> | Steve^ you can keep track of that exactly by following WHATWG on twitter. |
| 16:41 | <TabAtkins> | I think at the point that users see and understand SQL they could just read the js and be equally enlightened. |
| 16:42 | <annevk2> | Steve^, at this point changes are mostly non-normative |
| 16:42 | <annevk2> | Steve^, though a lot of the changes could help you understand the spec better, such as the addition of markup examples |
| 16:42 | jgraham | was being less than entirely serious |
| 16:42 | <Steve^> | ok, cool |
| 16:42 | <TabAtkins> | jgraham: Noted, but still wanted to respond. ^_^ |
| 16:44 | <Steve^> | The spec talks of an outlining algorithm. Has this been implemented anywhere? (Can I see what it would produce for my site?) |
| 16:45 | <TabAtkins> | Damn, having a hard time finding if anyone has put together a js library for executing sql against html tables, because of all the noise about people wanting to display sql results as tables. |
| 16:45 | <TabAtkins> | Steve^ gsnedders has something for that. |
| 16:45 | <da3d> | Steve^: http://gsnedders.html5.org/outliner/ |
| 16:45 | <Steve^> | thanks |
| 16:47 | <Steve^> | uh oh, I have untitled sections ;) |
| 16:48 | <Philip`> | TabAtkins: There's stuff like http://code.google.com/p/trimpath/wiki/TrimQuery for doing SQL in JS |
| 16:49 | <Philip`> | There must be a better syntax for this kind of thing than SQL, though |
| 16:49 | <TabAtkins> | Philip`, awesome thanks. |
| 16:49 | <TabAtkins> | Aw, I like SQL though. |
| 16:49 | <TabAtkins> | Plus, everyone knows it. |
| 16:50 | <Philip`> | Everyone knows one dialect of it, which is slightly incompatible with every other dialect |
| 16:51 | <TabAtkins> | But only in minor ways. The essence of SQL is pretty universal. |
| 16:51 | <TabAtkins> | Especially simple stuff like SELECTing. |
| 16:51 | <TabAtkins> | A lot of the differences I notice are in CREATE and complex things like triggers. |
| 16:51 | <AryehGregor> | Like LIMIT vs. TOP vs. some convoluted thing involving stuff I've never even heard of? |
| 16:52 | <AryehGregor> | Or in how when doing GROUP BY you can select unrelated columns with aggregation operators in MySQL, but like no other DBMS? |
| 16:52 | <AryehGregor> | Or like how SQLite requires an ESCAPE clause after LIKE if you want to escape anything? |
| 16:52 | <TabAtkins> | Bah, you're no fun. |
| 16:52 | <Steve^> | I agree with TabAtkins |
| 16:53 | <Steve^> | The core commands are transferable |
| 16:53 | <TabAtkins> | Anyway, modifying TrimQuery actually looks like it could be done solely by parsing a table into a json resultset. |
| 16:53 | <Steve^> | which is extremely helpful when you're in an unfamiliar system |
| 16:53 | <AryehGregor> | Or like INSERT IGNORE, REPLACE, INSERT ... ON DUPLICATE KEY UPDATE from MySQL? |
| 16:53 | <AryehGregor> | Or multivalue INSERT? |
| 16:54 | TabAtkins | plugs his ears. |
| 16:54 | <TabAtkins> | Notice that I said "the essence of SQL". |
| 16:54 | <Philip`> | Or concatenating strings |
| 16:54 | <AryehGregor> | Or the exact syntax for UNION. |
| 16:54 | <Philip`> | Or whether columns are statically typed |
| 16:55 | <AryehGregor> | Or . . . well, look in includes/db/ in MediaWiki for our DB abstraction layer. :) |
| 16:55 | <TabAtkins> | So hey, your point has long since been driven into the ground. Congratulations. |
| 16:55 | <AryehGregor> | Lots of fun stuff in there, and most of it doesn't work too well regardless. |
| 16:55 | <AryehGregor> | Also, SQL is a terrible language on its merits. |
| 16:55 | <Philip`> | TabAtkins: The essence of SQL is just relational algebra, and you can do that in any syntax, e.g. a nicer JS-like one :-) |
| 16:55 | <AryehGregor> | It's unreasonably verbose, unpredictable, and hard to understand for newcomers. |
| 16:56 | <Philip`> | (I guess that's kind of what LINQ does too, though I know nothing about it) |
| 16:56 | <TabAtkins> | AryehGregor: And yet, it exists everywhere. It's ubiquitous. It's a very successful horrible technology. |
| 16:56 | <Philip`> | TabAtkins: Just like HTML |
| 16:56 | gsnedders | wants to implement outlining in JS |
| 16:56 | <TabAtkins> | *Exactly*. |
| 16:57 | <TabAtkins> | gsnedders: That's easy. |
| 16:57 | <Philip`> | gsnedders: Just write a Python interpreter in JS |
| 16:57 | <TabAtkins> | gsnedders: If you can include a jquery dependency... |
| 16:57 | <gsnedders> | TabAtkins: It's not hard without it |
| 16:58 | <TabAtkins> | gsnedders: Yeah, but it's easier. And I can't actually program in DOM. ;_; |
| 17:00 | <jgraham> | gsnedders: Why are you telling us about it? you could be *doing* it |
| 17:01 | <TabAtkins> | Doing it is way funner. |
| 17:01 | TabAtkins | already knows what this weekend's project will be. |
| 17:03 | <gsnedders> | TabAtkins: It's no easier, as far as I can tell. |
| 17:03 | <gsnedders> | jgraham: Waiting for phone calls… |
| 17:04 | <Steve^> | oh, its 9/11 |
| 17:05 | <Steve^> | trending topics seem to be my main source of news nowadays |
| 17:05 | <gsnedders> | Steve^: That's not a valid ISO8601 date. |
| 17:06 | <Philip`> | Steve^: The date is considered news? |
| 17:06 | <jgraham> | Philip`: Well it changes every day |
| 17:07 | <Philip`> | Dates don't seem particularly rare, or unpredictable |
| 17:07 | <jgraham> | You obviously never lived through a calendar change |
| 17:07 | <Philip`> | Well, no, because I'm not five hundred years old |
| 17:09 | <beowulf> | yeah, but with all the kids running around you feel it, don't you? |
| 17:10 | <gsnedders> | scrollIntoView is DOM0, right? |
| 17:14 | <gsnedders> | And isn't in CSSOM View… |
| 17:14 | <gsnedders> | scroll(el.offsetLeft, el.offsetTop) should be the same as el.scrollIntoView(true), right? |
| 17:16 | <Steve^> | I wonder if I can change wordpress to automatically add sections to my posts |
| 17:17 | <Steve^> | Although, if it can be automatic, you could say its pointless to do |
| 17:39 | <annevk2> | and, is http://web.archive.org/web/20080204211015/http://www.hitachi-to.co.jp/prod/prod_2/inter/emk/help/TextEncoder/CodePage.htm more accurate for Internet Explorer? |
| 17:39 | <annevk2> | and, it does include UTF8 as label, for instance |
| 17:47 | <Steve^> | the definition of flow content feels weak |
| 17:59 | <Lachy_> | Steve^, how would you suggest improving it? |
| 18:01 | <Steve^> | I don't really understand what it is |
| 18:02 | <Lachy_> | it's a catch-all category for things that can be used almost anywhere |
| 18:02 | <Lachy_> | the name of it is inherited from the HTML4 concept of %Flow; from the DTDs |
| 18:03 | <Steve^> | You list the things that can be used almost anywhere in Flow content, rather than listing the few things that can't be? |
| 18:07 | <Lachy_> | yes |
| 18:08 | <Steve^> | Flow feels like the odd one out, but I can't think of a better solution |
| 18:12 | <Lachy> | Steve^, I once tried to come up with a better description for it, but didn't really succeed |
| 18:12 | <Lachy> | http://dev.w3.org/html5/html-author/#flow-content |
| 18:13 | <Steve^> | is there a list somewhere of things that aren't flow? |
| 18:14 | <Lachy> | Steve^, there will eventually be a table that lists all the elements and their categories here http://dev.w3.org/html5/html-author/#index-of-elements |
| 18:29 | <Lachy> | so apparently, HTML 5's proposed basis in DOM/JS skews web control and monetization towards corporations and away from individual authors/researchers, to the detriment of society, and we should incorporate ODRL into HTML5. |
| 18:29 | <Lachy> | http://lists.w3.org/Archives/Public/www-tag/2009Sep/0028.html |
| 18:30 | <Lachy> | if anyone can actually comprhend what that guy is ranting about, let me know |
| 18:33 | <TabAtkins> | Lachy: As far as I can tell, he's going something like this (you've probably already gotten this far in comprehension): |
| 18:33 | <hober> | Lachy: seriously. Also, makes an interesting constrast with http://happyworm.com/blog/2009/08/24/html-5-the-revolution-will-not-be-televised/ |
| 18:33 | <Steve^> | "when there are high barriers to entry in science and |
| 18:33 | <Steve^> | arts publishing, publishers can achieve monopolistic or oligopolistic |
| 18:33 | <Steve^> | returns of scale." |
| 18:33 | <TabAtkins> | "I don't understand javascript or the DOM. Web 2.0 things use javascript, and make a lot of money. I am also anti-corporate, so anything that makes money must be bad for me, especially since I don't understand it." |
| 18:34 | <Steve^> | So, the document is so complicated that only Microsoft can profit from it? |
| 18:34 | <hober> | Steve^: that would explain how IE8 has more HTML5 features in it than all of the other browsers combined. Oh, wait. |
| 18:35 | <Lachy> | TabAtkins, I got the "I don't understand javascript or the DOM" bit myself, and your summary of the rest seems to make sense |
| 18:35 | <Steve^> | I don't know Javascript and yet the HTML5 is quite readable to me |
| 18:35 | <Steve^> | they seem to be quite separate |
| 18:37 | <Steve^> | FYI, I am NOT Steven Rowat :P |
| 18:37 | <TabAtkins> | Then there's some vague hand-waving about it being impossible for an individual to understand html5, thus hurting work-at-home efforts and presumably benefitting corporations. I think this is just tying back into the "I don't understand" angle. |
| 18:37 | <Lachy> | Steve^, I assumed that since nothing you've said reaches the same level of incomprehension as him |
| 18:38 | <Steve^> | awww, thanks |
| 18:38 | <Lachy> | TabAtkins, have you figured out how the ODRL thing at the end fits in with the rest? |
| 18:38 | <Steve^> | Oh and which bit of HTML 5 is complicated? |
| 18:39 | <TabAtkins> | Wow, I actually skipped right past that. One sec. |
| 18:39 | <Steve^> | Sure, the content models are new and require some thought, but nothing the odd blog post won't fix |
| 18:40 | <TabAtkins> | All right, looks like the odrl thing is "I don't understand ODRL, but it has 'Open' in its name and thus is probably bad for corporations. Use that instead of javascript." |
| 18:40 | <Steve^> | TabAtkins, you could have your own TV show to rival the likes of Dr Phil |
| 18:40 | <TabAtkins> | Presumably for expressing rights, but it's not made clear how that relates to anything else in the email. |
| 18:41 | <TabAtkins> | Steve^, what, where I ridicule people who are possibly schizophrenic on the internet? |
| 18:41 | <Steve^> | rather you translate their shit into real words |
| 18:41 | <TabAtkins> | Well, if I were on the TAG mailing list I might try to do that. |
| 18:42 | <Steve^> | If you were to switch every mention of Javascript to Silverlight, the document would make an equal amount of nonsense |
| 18:42 | <TabAtkins> | But as far as I can tell I don't want to be on the TAG mailing list. |
| 18:43 | <Steve^> | The conspiracy theory must be that the head of Mozilla and Opera are undercover Microsoft operatives |
| 18:43 | <TabAtkins> | Gah, I think my brain is locking in its attempt to understand how rights expression has anything to do with the complexity of javascript. |
| 18:44 | <Lachy> | Steve^, shh! Don't tell everyone |
| 18:44 | <TabAtkins> | Steve^, I've seen the pictures with Sylvain, Anne, and Hakon having drinks together. Clear evidence of conspiracy! |
| 18:44 | <TabAtkins> | Lachy, are you planning to respond to that email? |
| 18:44 | <Lachy> | no |
| 18:44 | <Lachy> | I'm not on the tag mailing list |
| 18:44 | <TabAtkins> | k. ^_^ |
| 18:45 | <Steve^> | the only response it: |
| 18:45 | <Steve^> | TLDR |
| 18:45 | <Steve^> | *is |
| 18:45 | <Lachy> | WTF? |
| 18:45 | <Steve^> | Too long, didn't read |
| 18:45 | <TabAtkins> | That would be a valid response, but I think actually calling out the nonsensical bullshit would be amusing. |
| 18:45 | <TabAtkins> | tl;dr |
| 18:46 | <Steve^> | if you say that the documents are of a monopolistic nature, then his arguments may make some sort of sense |
| 18:46 | <Steve^> | but there is no proof of the former |
| 18:47 | <Lachy> | Steve^, ah, but in my case, it would be TMCDCAAWTFDODRLHTDW?I |
| 18:47 | <Lachy> | --> Too much crap, didn't comprehend anything, and wtf does ODRL have to do with it?! |
| 18:48 | <Dashiva> | So many interesting people |
| 18:49 | <waqas> | Dashiva, just what I was thinking. |
| 18:49 | <TabAtkins> | So, full explanation of his email: "I don't understand javascript, so it must be too complicated for *anyone* to understand and only groups can use it. Corporations have groups of people, so they're trying to influence the language to make it better for them and worse for startups. This is a tragedy because then nobody can work from home. Apropos of nothing, I like ODRL and think it... |
| 18:49 | <TabAtkins> | ...should be given language-level support." |
| 18:51 | <Steve^> | you have invested too much thought into this |
| 18:51 | <Steve^> | the crazy people have won another round |
| 18:52 | <TabAtkins> | Indeed. |
| 18:52 | <Philip`> | Steve^: You are ignoring the return on the investment, which is a reasonable amount of entertainment and a continued feeling of superiority, making it worthwhile |
| 18:53 | <Steve^> | this is true |
| 18:53 | <Steve^> | I am superior |
| 18:54 | <cardona507> | is <meta name=""> for keywords and description? |
| 18:56 | <Lachy> | cardona507, http://www.whatwg.org/specs/web-apps/current-work/multipage/#standard-metadata-names |
| 18:56 | <cardona507> | <meta name="description" content="foo">? |
| 18:56 | <cardona507> | thanks lachy |
| 18:58 | <Philip`> | Why does http://html5.googlecode.com/svn/trunk/incoming/ exist, containing two SVN repositories inside the SVN repository? |
| 19:08 | <Philip`> | Would people prefer multipage sections to be generally larger or smaller than they currently are? |
| 19:09 | <TabAtkins> | Smaller. |
| 19:09 | <Steve^> | smaller |
| 19:09 | <Philip`> | Why? |
| 19:09 | <TabAtkins> | Because they're very large? |
| 19:10 | <Steve^> | My internet is through a mobile broadband dongle and its expensive to keep loading so much information |
| 19:10 | <Steve^> | it would be nice for the default spec page to not be the single page version |
| 19:10 | <TabAtkins> | And I just plain have an old computer that takes a bit to process even the multipage sections. |
| 19:10 | <Philip`> | Steve^: Which default? http://whatwg.org/html5 is the multipage one |
| 19:10 | <TabAtkins> | Steve^ - just go to whatwg.org/html5. That's the multipage version. |
| 19:10 | <Steve^> | yes, I know that now |
| 19:11 | <Steve^> | but if you click Specs on the frontpage |
| 19:11 | <and> | annevk2: I have no reason to believe that it should be more accurate, but it would be a good idea to have a look at the labels which are listed either there or in the .NET reference but not in the current wiki table. |
| 19:11 | <Philip`> | Steve^: Hmm, that's a good point - maybe you should mail Hixie asking him to change it |
| 19:12 | <Philip`> | Anyway, since the votes are currently 2 vs 0, I'll try to err on the side of making smaller sections :-) |
| 19:12 | <Steve^> | the first bit of the spec detailing the options available should be displayed instead |
| 19:13 | Philip` | could aim for <100KB |
| 19:13 | <TabAtkins> | That'd be nice, Philip`. |
| 19:13 | <and> | annevk2: The change from ‘may’ to ‘must’ this morning is not really a ‘minor’ ‘[w]ording’ edit. |
| 19:14 | <Steve^> | Philip`, the current version has a page for each second-level menu item? |
| 19:14 | <Steve^> | you would split that into several pages or split it by third-level? |
| 19:14 | <and> | hsivonen: Thanks a lot! |
| 19:15 | <annevk3> | and, I didn't make that change, but I do agree with it (well, that was the intent of the original wording anyway) |
| 19:16 | <and> | annevk3: I noticed that, and was not sure you would agree. |
| 19:16 | <annevk3> | and, ah ok |
| 19:16 | <annevk3> | and, btw, you said the IE data didn't match, I totally missed http://html5.org/temp/2009/ie-encodings.htm (copy of webarchive) earlier which also gives IE data |
| 19:17 | <Philip`> | Steve^: It's split on <h2>s (the ~dozen top-level sections), plus a list of human-selected points that attempt to make it sensibly split (no huge pages, no tiny pages, etc) |
| 19:17 | <annevk3> | and, ah, you added that |
| 19:19 | <and> | Yes, it was the best reference I found a long time ago, and it still seems to list labels not on the wiki, but I have not checked whether or not IE8 actually support those. |
| 19:20 | <annevk3> | my IE situation is still pretty crappy because my laptop does not have enough ram for running several OSs at the same time |
| 19:22 | <and> | I plan to test the labels on the wiki in IE8 tonight and let you know what the results are. |
| 19:23 | <Steve^> | http://spoon.net/browsers/ lets you run any browser via the web |
| 19:23 | <Steve^> | though it is Windows only |
| 19:23 | <TabAtkins> | Hrm, IE8 won't run click handlers on display:none inputs. |
| 19:24 | <Dashiva> | How would you click them? |
| 19:24 | <TabAtkins> | <label> |
| 19:24 | <annevk3> | and, I'll try make a table for Safari/ICU this weekend |
| 19:24 | <TabAtkins> | And damn, visibility:hidden suppressed the handlers too. I guess I'll have to position them offscreen. >_< |
| 19:24 | <annevk3> | and, shouldn't be too hard |
| 19:25 | <Steve^> | TabAtkins, why are you clicking invisible buttons? |
| 19:25 | <TabAtkins> | Steve^: I have a list of checkboxes and radios. For styling purposes I'm hiding the inputs themselves and just styling the label when they're :checked. |
| 19:29 | <Steve^> | how do you know if they are checkboxes or radio buttons? |
| 19:29 | <TabAtkins> | Regardless, that seems like an interop bug (well, it *is* an interop bug - I just got a bug filed against me by QA because of it). Anyone know what the correct behavior is? |
| 19:29 | <TabAtkins> | Steve^ doesn't matter for this . |
| 19:29 | <Dashiva> | TabAtkins: Why don't you just have onclick on the label? |
| 19:30 | <TabAtkins> | Dashiva: a few reasons. For one, that's just silly - why not just remove the inputs entirely? For two, this code was originally designed assuming visible inputs, so I've got a lot of onclick handlers on the inputs. Moving them all to the label wouldn't be trivial. |
| 19:31 | <TabAtkins> | I do believe in semantics here, after all. ^_^ |
| 19:31 | <Dashiva> | It seems like a prime candidate for <span role="checkbox"> :) |
| 19:31 | <Steve^> | when you turn off the CSS, you would get nice checkboxes, true |
| 19:34 | <Philip`> | Argh, <div class=impl> breaks the splitter :-( |
| 19:34 | <AryehGregor> | TabAtkins, IE8 sounds like it should be correct . . . the input isn't receiving the click, after all. |
| 19:35 | <Steve^> | the label is though and it is responsible for passing the click to the input |
| 19:35 | <TabAtkins> | AryehGregor: Isn't it, though? Everyone treats it as receiving the click. I'm not sure which is the correct behavior. |
| 19:35 | <TabAtkins> | If you're saying that clicking the <label> should not, in general, fire click handlers for the associated input, then *everyone* disagrees with you. |
| 19:36 | TabAtkins | sighs. |
| 19:36 | <TabAtkins> | Our SEO company is getting uppity again about wanting server access. |
| 19:37 | <Steve^> | they want your dataz |
| 19:38 | <Steve^> | "For example, on platforms where clicking a checkbox label checks the checkbox, clicking the label in the following snippet could trigger the user agent to run synthetic click activation steps (page 123) on the input element, as if the element itself had been triggered by the user:" |
| 19:38 | <TabAtkins> | Yeah, everyone does that. |
| 19:38 | <Steve^> | someone chose the word "could". So you're screwed :) |
| 19:39 | <AryehGregor> | Yay. |
| 19:39 | <TabAtkins> | The question is just whether it should continue to do on inputs which have been hidden via CSS. |
| 19:40 | <Steve^> | just push them off the screen |
| 19:40 | <Steve^> | then everyone is happy |
| 19:41 | <AryehGregor> | TabAtkins, it looks like the spec is trying to defer to platform conventions here. |
| 19:41 | <TabAtkins> | Yeah, that's what I've done. But that's stupid and I hate it. |
| 19:41 | <AryehGregor> | I'm not sure that's meaningful in this case, though. |
| 19:41 | <TabAtkins> | Yeah, because there's probably not a sense of "hidden via CSS" in the normal platform. |
| 19:42 | <TabAtkins> | So, time to file a bug! |
| 19:44 | <Philip`> | Is there some way to stop Opera Turbo being stupid and popping up a little thing saying it thinks I'm on a slow network, when actually I'm not? |
| 19:44 | <Steve^> | click arrow, configure |
| 19:44 | <AryehGregor> | TabAtkins, more than that, platform conventions are only supposed to be relevant to UI, not development . . . |
| 19:46 | <Steve^> | Technically, you should be using some onCheck callback |
| 19:46 | <Philip`> | Oh, I didn't see the 'notify' option in the configure dialog |
| 19:49 | <TabAtkins> | Steve^: I suppose I could maybe use the change event, but I doubt that'll help here. |
| 19:53 | <Philip`> | Hixie: Is there any chance you could split the <div class="impl"> in "Parsing HTML documents" so there's one div around each subsection in there, instead of just one around the whole section? |
| 19:54 | <Philip`> | (other the spec-splitter can't split it while preserving the implness of all the subsections, and I'm not going to bother trying to add lots of fancy code to make it work) |
| 19:54 | <Philip`> | s/other/otherwise/ |
| 20:05 | <TabAtkins> | Man, upgrading wordpress was stupid easy. Two clicks. One to upgrade files, and one to upgrade the db when I logged into the admin again. |
| 20:05 | <Hixie> | Philip`: is that the only occurance of that problem? |
| 20:06 | <Hixie> | wait, aren't there like eighteen bazillion subsections in that one? |
| 20:06 | <Philip`> | Hixie: It's the only serious one (it ends up with a 380KB page) |
| 20:07 | <Philip`> | (I think "Rendering" is next most serious unsplittable one, but that's only 140KB which isn't really a problem) |
| 20:07 | <Hixie> | would it work if i just brough the divs down one level? |
| 20:07 | <Hixie> | rather than all the way |
| 20:08 | <Philip`> | Hixie: I just meant the 8 subsections, not the eighteen bazillion sub(sub)*sections |
| 20:08 | <Philip`> | so that would be fine |
| 20:08 | <Hixie> | ah ok excellent |
| 20:08 | <Hixie> | sure, i could do that |
| 20:14 | <Philip`> | http://www.whatwg.org/specs/web-apps/current-work/multipage/ should hopefully be better split now (into smaller pages), and should become betterer when the parsing impl is split up |
| 20:15 | <Hixie> | cool |
| 20:15 | <Hixie> | checked in the changes |
| 20:17 | <TabAtkins> | Philip`, I like it. |
| 20:17 | TabAtkins | was just reading something in 10.4, and is happy to see it shrunken. |
| 20:17 | <Steve^> | Hixie, when you follow the link to the latest draft from the specs page, would it be possible for that to go to the multiple-page version rather than the one-page version? |
| 20:19 | <Philip`> | TabAtkins: That particular section was just suffering from a bug (section 11 was in <div class=impl> so it got merged into section 10 instead of being split apart) |
| 20:19 | <Philip`> | Other sections were mainly suffering because the hard-coded list of splitting points was out of date |
| 20:20 | <TabAtkins> | Blarg, I meant 4.10 |
| 20:20 | <Philip`> | but now they're up to date again, temporarily |
| 20:20 | <Hixie> | Steve^: yeah, you or someone else e-mailed me about that revently, it's on my list of things to fix |
| 20:20 | <Philip`> | TabAtkins: Oh, okay |
| 20:20 | <Steve^> | thanks |
| 20:20 | <Hixie> | Steve^: most of the links already go straight to the multipage copy iirc |
| 20:22 | <Philip`> | Hixie: Thanks for the parser split, seems to be working fine now |
| 20:22 | <Hixie> | cool |
| 20:23 | <Philip`> | tree-construction is still 200KB but anyone crazy enough to look at that algorithm can surely cope with such a giant page |
| 20:25 | <erlehmann> | Philip`, how come the splitting is hard-coded ? |
| 20:26 | <Hixie> | i wonder what would be a prototypical table |
| 20:26 | <Hixie> | maybe something from The Origin of Species |
| 20:26 | <Philip`> | erlehmann: Because that's easier than finding a nice automatic approach |
| 20:27 | <Philip`> | (The goals are to make each page ~100KB, and preferably to split along hierarchical lines, and otherwise to split at logical points) |
| 20:28 | <Dashiva> | Hard to solve optimally, but easy with a little cheating :) |
| 20:28 | <erlehmann> | knuth probably solved this back in the 70ies ;D |
| 20:28 | <Hixie> | ...or not, it apparently has no tables. |
| 20:28 | <Philip`> | The problem is with the idea of "logical" |
| 20:28 | <Philip`> | which is rather hard to define |
| 20:29 | <Dashiva> | Philip`: Try lots of different splits, send them out pairwise on mechanical turk |
| 20:29 | <Philip`> | Dashiva: Mechanical Turk costs tens of cents, and I don't think HTML5 is worth such an investment from me |
| 20:30 | <Dashiva> | Or you could use GA |
| 20:30 | <TabAtkins> | with what fitness criteria? |
| 20:30 | <erlehmann> | Philip`, why not use the heading level nesting algorithm or how its called ? split at highest level, look if significantly over >100KB, look at subpart, repeat |
| 20:30 | <Dashiva> | TabAtkins: You provide thumbs up and thumbs down buttons to the user |
| 20:30 | <erlehmann> | that would seem logical to me |
| 20:31 | <Dashiva> | erlehmann: That fails on cases like 110 KB with 10 subsections |
| 20:32 | <Philip`> | erlehmann: Look at e.g. "4.10.4.1 States of the type attribute" |
| 20:33 | <Philip`> | It would seem illogical to me to split it in the middle of the sequence of date/time-related types - it's better to split just after them |
| 20:33 | <Philip`> | but an algorithm couldn't know that |
| 20:34 | <erlehmann> | there are no semantic hooks in the document from which to "get" that, so … |
| 20:34 | <Philip`> | So that's why it's better having a human deciding where to split :-) |
| 20:35 | <Philip`> | Hixie: What do you mean by "prototypical"? |
| 20:36 | <Hixie> | Philip`: maybe "canonical" would be a better term, though it's not quite right either |
| 20:36 | <Philip`> | If you're looking for well-known tables of data, maybe something like census data could be appropriate? |
| 20:36 | <Dashiva> | A multiplication table :) |
| 20:36 | <Lachy> | why doesn't Hixie just insert some special markers into the spec where it is logical to split? Is that harder than maintaining a list of hard coded section ids? |
| 20:36 | <annevk3> | matrix math |
| 20:36 | <Hixie> | Dashiva: already have an addition table |
| 20:36 | <Dashiva> | Lachy: I suppose Hixie has a higher workload than Philip` |
| 20:36 | <Hixie> | i'd like a table with a <tfoot>, spans, colgroups, all kinds of stuff |
| 20:37 | <Philip`> | Lachy: That would require Hixie to do more work |
| 20:37 | <Philip`> | Hixie: It seems unrealistic to have a single table with all the possible features, and needlessly confusing |
| 20:38 | <Philip`> | Hixie: You could do an example with a table of HTML elements and attributes |
| 20:38 | <Dashiva> | Hixie: http://www.pvv.org/~magnusrk/wurm/item.php :) |
| 20:38 | <Dashiva> | I guess I don't use tfoot |
| 20:38 | <Philip`> | As a bonus, it would also be a table of HTML elements and attributes |
| 20:38 | <Hixie> | oh it also has to be small |
| 20:38 | <Philip`> | s/HTML/HTML5/g |
| 20:39 | <Philip`> | Reduce the font size to 6pt |
| 20:39 | <Hixie> | maciej is here, bbl |
| 20:55 | <Lachy> | with all the discussion about pedagogic validation and encourage XML-inspired habits because it supposedly easier for students, I really wish I could remember how I managed to cope with learning back in the pre-XHTML age when HTML 4 was bleeding edge |
| 20:55 | <AryehGregor> | I did okay. |
| 20:55 | <AryehGregor> | I only learned XML later. |
| 20:56 | <Dashiva> | There already was one mail (smylers I think?) that pointed out how many of the XML things are actually counterintuitive |
| 20:56 | <Lachy> | Dashiva, I don't recall that one. Got a link? |
| 20:56 | <AryehGregor> | I strongly suspect that "XML is easier to learn" is a smokescreen for "I don't like XML and am an educator so here's an argument you spec writers/implementors/authors can't really argue with". |
| 20:57 | <AryehGregor> | I think the trailing solidus in text/html is obviously unintuitive, since it doesn't actually do anything but people will think it does. |
| 20:57 | <AryehGregor> | I've made that mistake and know someone else who did, and I'd say we were both at a much higher level of comprehension than your average HTML student. |
| 20:58 | <Dashiva> | Lachy: http://lists.w3.org/Archives/Public/public-html/2009Sep/0456.html |
| 20:58 | <Lachy> | AryehGregor, yeah, I do remember that was somewhat confusing when I tried it on some non-empty elements |
| 20:59 | <AryehGregor> | No comment on quotes. I don't see how omitting quotes is unintuitive, though. |
| 20:59 | <AryehGregor> | Unless maybe you're used to usual programming languages? |
| 21:00 | <Lachy> | one thing that confused the hell out of me was why, if XHTML was supposed to require well-formedness, I didn't understand why browsers didn't report the errors. It wasn't till a little later that I learned about XML mime types that it all made sense |
| 21:00 | <Lachy> | and I realised that XHTML as text/html was actually harmful to my learning |
| 21:01 | <Steve^> | I remember a kid telling me that you could omit the closing body and html tags, which I thought was incorrect and very bad |
| 21:01 | <Steve^> | It didn't help the teacher told us it was herf and scr |
| 21:01 | <Steve^> | (href / src) |
| 21:01 | <Lachy> | at least, now with the proposals for pedagogic validation, students will no longer be mislead about writing XHTML when they're just writing HTML with specific coding conventions |
| 21:03 | Lachy | is contemplating getting up and actually going to the party I said I would go to tonight. |
| 21:03 | <AryehGregor> | Pfft, parties. |
| 21:04 | gsnedders | grumbles something about all his photos from his ball being terrible |
| 21:04 | <Lachy> | as much fun as partying on IRC is and all |
| 21:04 | <Steve^> | I say ignore them, no programming languages are alterred to keep teachers happy |
| 21:04 | <Steve^> | why is HTML? |
| 21:04 | <gsnedders> | Steve^: This is altering the language, though. |
| 21:04 | <AryehGregor> | A lot of programming languages are designed to be easy to learn. |
| 21:05 | <AryehGregor> | Those are usually the most horrifying and evil ones. |
| 21:05 | <Steve^> | sure, Pascal supposedly |
| 21:05 | <Steve^> | its a teaching language |
| 21:05 | <AryehGregor> | I'm thinking like Visual Basic or various macro languages. |
| 21:05 | <Steve^> | Java is fairly straight forward by design, not necessarily so its easier to teach it |
| 21:06 | <Steve^> | I haven't encountered XJava where all variables must be explicitly destroyed or anything |
| 21:07 | <Dashiva> | "Besides, you need <html> for lang!" <-- Great argument there :) |
| 21:07 | <Steve^> | oh wait, you should always wrap if statements in { } even if it is a single command |
| 21:07 | <TabAtkins> | Lachy: Your talk of parties just reminded me that it's liquor friday. |
| 21:07 | <Lachy> | Dashiva, my response to that would be you don't need lang="" |
| 21:07 | <Steve^> | ah Friday Beer, those were the days |
| 21:07 | <gsnedders> | Steve^: s/were/are/ |
| 21:07 | <Dashiva> | Lachy: Indeed |
| 21:08 | <gsnedders> | (It's just I'm not allowed to have beer) |
| 21:08 | <TabAtkins> | It's a lot easier to have Liquor Friday when you work from home. It was annoying to wait for management to leave previously. |
| 21:08 | <Lachy> | Steve^, I didn't know you used to work for Opera |
| 21:08 | <Steve^> | Lachy, then you probably didn't know we've met |
| 21:08 | <gsnedders> | LOL |
| 21:08 | <TabAtkins> | Heh. |
| 21:09 | <gsnedders> | Lachy: fail. |
| 21:09 | <Lachy> | Steve^, are you the guy I met a couple of weeks ago at friday beer when Bruce Lawson was here? |
| 21:09 | <Steve^> | indeed |
| 21:09 | <Lachy> | wow |
| 21:09 | <Steve^> | amongst the talk of PDFs |
| 21:09 | gsnedders | should go to the Oslo office sometime |
| 21:10 | <TabAtkins> | What's the drinking age in the UK, gsnedders? 18? |
| 21:10 | <gsnedders> | TabAtkins: Short answer: it depends. |
| 21:10 | <TabAtkins> | That's... interesting. |
| 21:11 | <Philip`> | It depends on what you're drinking, e.g. water has no restriction |
| 21:11 | <Lachy> | gsnedders, I think the question is mean to be, at what age did you start getting drunk? |
| 21:11 | <gsnedders> | To quote wikipedia: |
| 21:11 | <TabAtkins> | That is not what the question is meant to be. >_< |
| 21:11 | <gsnedders> | "Children under 5 must not be given alcohol unless under medical supervision or in an emergency (Children and Young Persons Act 1933, Children and Young Persons (Scotland) Act 1937).[48][49]. However, children aged 5 and over may legally consume alcohol in their own home or someone else's as long as they are under the supervision of an adult. |
| 21:11 | <gsnedders> | The minimum age for the purchase of alcohol is 18. People aged 16 and 17 may consume wine, beer or cider on licensed premises (pubs/bars/restaurants) with a table meal. In England & Wales, an adult must order.[50] In Scotland, no adult is required to be present.[51][52] The legal age for the purchase of alcohol from an off-licence (store/supermarket) is 18 (16 for liqueur chocolates). |
| 21:11 | <gsnedders> | Purchasing alcohol on behalf of a minor is illegal in Scotland, England and Wales. This means acting as the young person's agent.[53][54] |
| 21:11 | <Steve^> | rules? |
| 21:11 | <Steve^> | I know you need to "look over 25" to not get IDed nowadays |
| 21:11 | <TabAtkins> | So, 16 more or less. |
| 21:11 | <Lachy> | what emergency situation would require giving a child alcohol? |
| 21:11 | <Steve^> | 25 is a long way from 17 |
| 21:11 | <gsnedders> | TabAtkins: No, 18 more or less |
| 21:11 | <TabAtkins> | Lachy: He just got divorced. |
| 21:12 | <AryehGregor> | Lachy, no other anesthetic handy? |
| 21:12 | <TabAtkins> | gsnedders: 18 to buy. Looks like 16 can buy in pubs, though in some parts an adult needs to buy it first. |
| 21:12 | <Steve^> | 16 cannot buy |
| 21:12 | <Steve^> | 18+ only can buy |
| 21:12 | <TabAtkins> | Oh, I see. Was reading it differently. |
| 21:12 | <Steve^> | includes cigs and lighters |
| 21:12 | <gsnedders> | Steve^: In a pub in Scotland and NI you can get wine, beer, or cider with food at 16. |
| 21:13 | <gsnedders> | s/pub/pub or restaurant/ |
| 21:13 | <Steve^> | does a pack of crisps count as food? |
| 21:13 | <gsnedders> | Steve^: From memory the law says it has to be a proper meal. |
| 21:14 | <TabAtkins> | Wiki says "with a table meal". |
| 21:14 | <Steve^> | so, pie or bigger? |
| 21:14 | <TabAtkins> | Sure. |
| 21:15 | <Steve^> | I prefer England as I can afford beer on everyday of the week |
| 21:15 | <Steve^> | I don't just rely on the free friday stuff |
| 21:15 | <gsnedders> | Lachy: As for when I started getting drunk, you so don't know how Straight Edge I am :P |
| 21:15 | <TabAtkins> | Caffeine, gsnedders? |
| 21:16 | <gsnedders> | TabAtkins: Almost never. |
| 21:16 | <TabAtkins> | Mang. I likes me the basic legal recreational chemicals. |
| 21:17 | <TabAtkins> | I mean, even sugar alters brain chemistry. |
| 21:17 | <Steve^> | so does oxygen |
| 21:17 | <TabAtkins> | I'd call that a *part* of brain chemistry. |
| 21:18 | <TabAtkins> | You don't really have much chemistry going on there without oxygen. |
| 21:18 | <Philip`> | I cryogenically freeze my brain to prevent its chemistry changing over time |
| 21:18 | <gsnedders> | Philip`: Then how did you construct that sentence. |
| 21:18 | <Steve^> | I'm addicted to oxygen |
| 21:19 | <TabAtkins> | Pretty sure Philip`'s just a longstanding chat macro, hurriedly programmed during his brief moments of thawing. |
| 21:19 | <Philip`> | gsnedders: I occasionally makes clones of my frozen brain, let them act as me, and then destroy them before they diverge too far from the pure original me |
| 21:19 | <Philip`> | s/makes/make/ |
| 21:19 | <Steve^> | these clones aren't very good typists |
| 21:20 | <Philip`> | I blame my keyboard :-( |
| 21:20 | <TabAtkins> | I have an excuse when I type extra "e"s. I fixed my e key in the gimpiest way possible. |
| 21:20 | <gsnedders> | Also: the amusing thing about working for Opera while applying to uni is that at open days anybody in CS who asks me what I'm doing during gap year is they try _really_ hard to convince me to go to their uni. |
| 21:20 | <TabAtkins> | It's currently an n key resting on half a pin head, secured by tape. |
| 21:21 | <TabAtkins> | Haha, I'll bet gsnedders. |
| 21:22 | <gsnedders> | Then when I point out I don't have very good grades, they're more working along the lines of how to work around this, rather than reacting that I have no chance. :P |
| 21:22 | <TabAtkins> | Yeah, actual work is way better at demonstrating competence than the proxy of grades. |
| 21:23 | <gsnedders> | In the UK you're _really_ hard pushed to get into uni without grades. |
| 21:23 | <TabAtkins> | Yeah, but y'all have all the class issues that we mostly dropped. |
| 21:23 | <gsnedders> | Also, FWIW, wrt to what you were saying about for fees in US nowadays, I have my doubts whether that applies to international students too |
| 21:24 | <gsnedders> | Also: why does Google in Opera default to google.se now!? |
| 21:25 | <Steve^> | because you're in Sweden? |
| 21:25 | <gsnedders> | I'm not, I'm in .uk for this month. |
| 21:26 | <Steve^> | you know when you've been on the internet too long when you refer to countries by their TLDs |
| 21:26 | <gsnedders> | :D |
| 21:26 | <gsnedders> | It only gets confusing when you hit things like .za |
| 21:26 | <Steve^> | South Africa |
| 21:26 | <gsnedders> | I had to look that up a few weeks ago :) |
| 21:27 | <AryehGregor> | I'd have guessed Zambia or something. |
| 21:27 | <TabAtkins> | For some reason I know that one. But I didn't know .se |
| 21:27 | <AryehGregor> | I knew .se. |
| 21:27 | <AryehGregor> | I know language codes better, because Wikipedia uses language codes and not TLDs. |
| 21:28 | <gsnedders> | I use a mix of the two-alpha-char country codes and TLDs |
| 21:28 | TabAtkins | needs to go spend the rest of his lunch playing video games. bbl |
| 21:29 | <Steve^> | lunch? |
| 21:29 | <TabAtkins> | I'm GMT-6 (well, -5 right now) |
| 21:29 | <TabAtkins> | It's 3:30pm over here. |
| 22:03 | <Steve^> | what is div short for? |
| 22:05 | <Philip`> | Division? |
| 22:05 | <Steve^> | In a basic HTML5 markup, div is a bit of an eyesore. You have nav, section, article and then.. div |
| 22:05 | <Steve^> | I think that's why some have called for a <content> element |
| 22:06 | <Steve^> | it has an english meaning |
| 22:06 | <Philip`> | You could simply not use div |
| 22:06 | <Steve^> | indeed, the div is only there for styling reasons |
| 22:06 | <Steve^> | and it should be a div |
| 22:07 | <Steve^> | I see people getting carried away with the new semantic elements, thinking they will cover everything |
| 22:07 | <Steve^> | and expecting to remove all divs |
| 22:08 | <Steve^> | if it has a pretty new name, everyone would be happy :) |
| 22:09 | <AryehGregor> | Ian said he thinks <div> shouldn't exist. |
| 22:09 | <AryehGregor> | Although he admitted this was crazy. :) |
| 22:10 | <Steve^> | for a black and white website you'd be ok |
| 22:10 | <inimino> | removing all divs would require fixing CSS |
| 22:11 | <Steve^> | it would be tricky to give your site a full border |
| 22:11 | <inimino> | or settling for simpler designs :) |
| 22:11 | <AryehGregor> | No, it would only require browsers implementing current cutting-edge CSS, mostly. |
| 22:11 | <AryehGregor> | Well, that might not be quite enough. |
| 22:11 | <AryehGregor> | Isn't there an artificial-wrapper-creator in some draft, though? |
| 22:14 | <Steve^> | lets bring back tables \o/ |
| 22:17 | <AryehGregor> | They still exist, you know. |
| 22:17 | <Philip`> | Tables never went away |
| 22:19 | <erlehmann_> | table webdesign at tables.on.nimp.org ;D (dont click) |
| 22:19 | <Steve^> | if table, tr, td, th were actually called layout, row, panel, heading respectively, they'd have been loved forever |
| 22:19 | <jamesr> | i doubt that |
| 22:20 | <erlehmann_> | Steve^, that is a horrible picture you are painting |
| 22:20 | <Steve^> | With the correct CSS, I can make up whatever tags I like? |
| 22:20 | Steve^ | is off to start a revolution |
| 22:25 | <erlehmann> | Steve^, DONT |
| 22:26 | <annevk3> | Philip`, any chance you're going to hack a mini-TOC in each separate page one of these days? |
| 22:27 | <annevk3> | one upon a time books led me to believe custom tags were the holy grail of Web design |
| 22:27 | <annevk3> | s/one/once/ sigh |
| 22:27 | <Dashiva> | Custom tags are like programs generated from spec code |
| 22:28 | <Dashiva> | The functionality has to be created somewhere |
| 22:46 | <TabAtkins> | erlehmann, I hate you. |
| 22:48 | <Steve^> | whoa |
| 22:49 | <Steve^> | someone's harbouring some bad thoughts |
| 23:18 | <TabAtkins> | Woo I got my friend to abuse his connections to get tickets to one of my favorite artists! |
| 23:47 | <and> | annevk2: The table seems to match IE8 apart from the (???) encodings, most of which are supposed to be EBCDIC-based but are decoded as US-ASCII, so perhaps not supported at all. |
| 23:49 | <and> | I have no idea what x-europa is, and there are several new Taiwanese encodings, but it does not seem too bad in general. |
| 23:53 | <and> | (I obviously got off to a bad start yesterday with the (???) ones.) |
