| 03:32 | <ezyang> | I think I have more complaining hsivonen to do again :-) |
| 03:33 | <ezyang> | Oh hey, no, it looks like Hixie just didn't label the appropriate commits |
| 03:39 | <ezyang> | No... I totally can yell at him ^_^ |
| 03:39 | <ezyang> | I'm pretty sure the spec doesn't deal with </x x x> yet |
| 03:45 | <ezyang> | nvm again |
| 05:58 | <ezyang> | I'm confused... does that mean that we can have naughty characters in our HTML5 documents? |
| 05:59 | <ezyang> | Huh, I guess so |
| 06:03 | <ezyang> | Oh god, 105 test-cases to fix |
| 06:05 | <ezyang> | Ruh oh... can't actually express \u10FFFF in JSOn |
| 06:09 | <ezyang> | Haha, whoo surrogates |
| 06:10 | <ezyang> | numericEntities: aggh |
| 06:10 | <ezyang> | I wonder where the script we used to generate this was |
| 07:28 | <ezyang> | wtf � |
| 07:30 | <MikeSmith> | ezyang: you testing surrogate handling? |
| 07:31 | <ezyang> | Uhh, sort of |
| 07:31 | <ezyang> | So Hixie landed a change to the spec which implies that crazy stuff like that should get passed through without harm |
| 07:31 | <ezyang> | But that doesn't have any legal surrogates... |
| 07:34 | <ezyang> | I think I should send him mail about this |
| 07:38 | <hsivonen> | ezyang: aside from lone surrogates, the weirdo characters that can now be used as numeric references were already available if you used them as UTF-8 in the raw stream |
| 07:39 | <ezyang> | UTF-8 doesn't even have a mechanism for building something that long... |
| 07:39 | <hsivonen> | ezyang: what were you about to complain to me about? |
| 07:40 | <ezyang> | It was my fault; after double checking with the spec all of your changes were correct. |
| 07:40 | <hsivonen> | ezyang: UTF-8 can build code points until 0x10FFFF just fine |
| 07:40 | <hsivonen> | ok |
| 07:40 | <ezyang> | On an unrelated note, have you looked at r3374? |
| 07:40 | <hsivonen> | of the spec? |
| 07:40 | <ezyang> | Yeah |
| 07:40 | hsivonen | looks |
| 07:41 | <hsivonen> | ezyang: the changes to the Windows-1252 table are questionable, but I'm not sure if they are what existing browsers do |
| 07:42 | <hsivonen> | "If the number is in the range 0x0001 to 0x0008" ... "error" is good |
| 07:43 | <hsivonen> | so, the C1 range stuff I'd have to double-check, but the rest is good |
| 07:46 | <hsivonen> | what's the right way to convert a single-character string to the number corresponding to the code point in JS_ |
| 07:46 | <hsivonen> | ? |
| 07:49 | <hsivonen> | charCodeAt |
| 07:51 | <hsivonen> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/159 |
| 07:52 | <hsivonen> | have to wait for Windows to boot to have a verdict... |
| 07:58 | <ezyang> | hsivonen: The interesting bit is how error handling changed |
| 07:59 | <ezyang> | Instead of replacing it with U+FFFD, we pass it through. |
| 07:59 | <ezyang> | This results in hilarity with large numeric entities |
| 07:59 | <hsivonen> | ezyang: larger than 0x10FFFF? |
| 07:59 | <ezyang> | yep |
| 07:59 | <Philip`> | ezyang: I think numericEntities.test and unicodeChars.test were generated by me with a Perl script that I might have not uploaded anywhere |
| 08:00 | <hsivonen> | oh. those should still becode U+FFFD |
| 08:00 | <ezyang> | Yep |
| 08:00 | <hsivonen> | if the spec now says otherwise, it's a bug |
| 08:00 | <ezyang> | I wrote a script to convert them, so don't worry about it |
| 08:00 | <ezyang> | Well, Hixie very deliberately changed it to match browser behavior |
| 08:00 | <hsivonen> | ezyang: what's browser behavior for > 0x10FFFF? |
| 08:00 | <ezyang> | Firefox simply ignores the character |
| 08:01 | <hsivonen> | that's not good |
| 08:01 | <ezyang> | Don't have access to other browsers at the moment |
| 08:01 | <Hixie> | oops, i thought i'd already made > 10FFFF get not parsed as a charref |
| 08:01 | <Philip`> | I thought Firefox wrapped values larger than 2^32 |
| 08:01 | <ezyang> | I could have missed that cluedump |
| 08:01 | <Hixie> | webkit treats � as &#x110000; |
| 08:01 | <hsivonen> | not great, either |
| 08:02 | <ezyang> | Hixie: I dropped you mail about it, ftr |
| 08:02 | <hsivonen> | hrm. my test case isn't working in IE at all |
| 08:02 | <Hixie> | thanks |
| 08:03 | <ezyang> | It's way past my bedtime. Let me know what we decide :-) |
| 08:03 | <Hixie> | it's on my pile |
| 08:03 | <Hixie> | my latency is about a month right now |
| 08:04 | <hsivonen> | how do I get the DOMString out of a text node in IE? |
| 08:04 | <hsivonen> | it's been a long time since I've written IE workaround JS |
| 08:04 | <Hixie> | .data |
| 08:04 | <hsivonen> | thanks |
| 08:05 | <hsivonen> | no, that's not it |
| 08:05 | <hsivonen> | nodeValue is supposed to be a prop on the object... |
| 08:05 | <hsivonen> | hmm. does IE go crazy if a node value contains a bad character? |
| 08:06 | <hsivonen> | nope. that's not the reason |
| 08:07 | <hsivonen> | something is going badly wrong in my testing in IE |
| 08:07 | <hsivonen> | I feel clueless |
| 08:10 | <hsivonen> | anyway, whatever IE does, it's not mapping the unmapped characters of Windows-1252 C1 range to U+FFFD |
| 08:11 | <Hixie> | .data definitely gets you the text out of Text node in IE last i checked |
| 08:11 | <hsivonen> | Opinions on https://bugzilla.mozilla.org/show_bug.cgi?id=503210 for Web DOM Core? |
| 08:12 | <Hixie> | re bug, no opinion. what's the use case? |
| 08:12 | <hsivonen> | Hixie: I'm failing to get even innerHTML out of an element or .length out of a text node in Live DOM Viewer in IE |
| 08:12 | <Hixie> | have you tried with just regular text? |
| 08:12 | <Hixie> | maybe their serialiser is barfing |
| 08:12 | <hsivonen> | Hixie: serializing any DOM as a JS script that recreates the DOM |
| 08:13 | <hsivonen> | Hixie: I tried with 'a' |
| 08:13 | <hsivonen> | Hixie: same result, which is weird |
| 08:13 | <Hixie> | what's the use case for serializing any DOM as a JS script that recreates the DOM? |
| 08:13 | <Hixie> | odd |
| 08:13 | <hsivonen> | Hixie: fuzzing |
| 08:13 | <Hixie> | so testing? not a platform feature? |
| 08:13 | <Hixie> | does the innerHTML part of the live dom viewer itself work? |
| 08:14 | <hsivonen> | Hixie: yes |
| 08:14 | <Hixie> | then i dunno what's going on |
| 08:19 | <Hixie> | hsivonen: did you implement the spec or webkit behaviour for table taint in the html5.parser? |
| 08:24 | <hsivonen> | Hixie: WebKit behavior |
| 08:25 | <Hixie> | anything to report? |
| 08:25 | <hsivonen> | nothing regarading Web compat regarding taint to report |
| 08:25 | <jgraham> | hsivonen: which version of IE? |
| 08:26 | <hsivonen> | the only problems have been with Gecko-only test suites |
| 08:26 | <hsivonen> | the only taint problems that is |
| 08:26 | <hsivonen> | jgraham: IE8 on Windows XP SP3 |
| 08:26 | <jgraham> | hsivonen: AFAICT IE8 is broken in the live dom viewer |
| 08:26 | <jgraham> | At least for me on windows 7 |
| 08:26 | <hsivonen> | jgraham: well, it's broken in a really subtle way |
| 08:27 | <hsivonen> | because I can w() the text node object |
| 08:27 | <Hixie> | hsivonen: |
| 08:27 | <Hixie> | er |
| 08:27 | <Hixie> | hsivonen: k |
| 08:27 | <jgraham> | hsivonen: oh. Well the whole w() function is broken for me |
| 08:27 | <Hixie> | i guess i'll make <table><b>do</b> <b>do</b> do <b>do</b> <b>do</b></table> parse as "dodo do dodo" instead of "do do do do do" tomorrow then |
| 08:28 | <hsivonen> | Hixie: nice |
| 08:28 | <Hixie> | not really, but it's hard to argue with your points :-) |
| 08:28 | <Hixie> | the list of things in html5 i disagree with is growing quite long |
| 09:02 | <Mrmil> | Hi, imagine a category tree structure and you are to select a category for your article, how would you do that? <select> with nested <optgroup>s seems to be an interesting idea but you cannot select XYZ category marked up with label="XYZ category" because it is, well, a label. I looked into the archives, saw some mails about it, but the spec is silent about this... |
| 09:26 | <Philip`> | Mrmil: <select>/<option> with indentation of the option text to indicate nesting? |
| 09:26 | <Philip`> | (That doesn't seem very nice but I guess it sort of works) |
| 09:31 | <Mrmil> | Philip`: Yeah, we're currently using that, but I don't think it's too accessible. If the optgroup mechanism was capable of doing it, it would be awesome |
| 09:42 | Philip` | wonders if something like <details><legend><input type=radio>Category</legend><details><legend><input type=radio>Sub-category</legend><input type=radio>Sub-sub-category</details></details> would be appropriate semantics |
| 09:42 | <Philip`> | (plus some <label>s or whatever) |
| 09:59 | <Mrmil> | looks quite interesting... |
| 10:07 | hsivonen | wonders if "slut" in Shane's blog URL triggered the "courtesy preview" on tinyarro.ws |
| 10:29 | <hsivonen> | http://blog.halindrome.com/2009/07/w3c-you-ignorant-slut.html?showComment=1247069810064#c2703275792180082243 |
| 10:31 | <takkaria> | Shane seems to think that walled garden |
| 10:31 | <takkaria> | +s are worth supporting explicitly |
| 10:33 | <Philip`> | and presumably worth supporting as first-class citizens of the language, rather than using <div class> and data-* attributes or similar extensibility mechanisms |
| 10:36 | <jgraham> | I'm not sure I take people seriously who come out with statements like "[HTML 5 is a] bloated monstrosity that already kills the battery on your phone" |
| 10:37 | <takkaria> | I'm not sure I take people seriously who answer "why modularise a spec?" with "because modularisation itself is the benefit" |
| 10:38 | <takkaria> | though I guess the answer was also "because of the mindset behind it", but that seems to be to be someone who aims for theoretical purity and would probably be unsuited to actually editing HTML5 in a way that wasn't theoretically pure |
| 10:39 | hsivonen | finds https://www.psd2html.com/order-now.html ISO 9001-certified. |
| 10:39 | <Philip`> | jgraham: Seems to be a strange and not entirely coherent mixture of complaining about the size of the language and complaining about the size of the single-page specification |
| 10:40 | <othermaciej> | oh, I saw that post and I didn't notice that was Shane McCarron's blog |
| 10:40 | <Philip`> | hsivonen: Someone should tell them they're allowed to write more than one line of code per <script> |
| 10:42 | <Philip`> | takkaria: It sounds to me like the mindset is about allowing extensibility, not about theoretical purity |
| 10:43 | hsivonen | notes it was jax from Opera who brought up phone batteries |
| 10:43 | <takkaria> | you could split HTML5 up into multiple modules and not have any more extension points than you have now, though |
| 10:44 | <hsivonen> | takkaria: or you could have prescribed extension points as in XHTML M12N but the wrong points |
| 10:44 | <hsivonen> | (see XHTML MP) |
| 10:44 | Philip` | wonders who jax is, then finds http://my.opera.com/jax/blog/2009/07/07/an-xhtml-2-far |
| 10:46 | <Philip`> | takkaria: Indeed, which is why the important thing is the mindset behind M12N (of providing well-defined extensibility points that are on an equal footing to the HTML markup itself) and not the mechanism of splitting the spec into modules |
| 10:48 | <MikeSmith> | Philip`: jax is Jonny Axelsson |
| 10:48 | <MikeSmith> | Opera employee |
| 10:48 | <MikeSmith> | or former one now, maybe |
| 10:48 | <Philip`> | MikeSmith: I figured that out after finding his blog :-) |
| 10:49 | <hsivonen> | former? |
| 10:49 | MikeSmith | hasn't read the blog yet |
| 10:49 | <MikeSmith> | hsivonen: I'm not sure if he still works for Opera |
| 10:49 | hsivonen | sees "now that I don't work for Opera anymore" |
| 10:49 | <MikeSmith> | hadn't heard much about/for him for a long time |
| 10:50 | <MikeSmith> | about/from |
| 10:50 | <MikeSmith> | I wonder who he does work for |
| 10:50 | <MikeSmith> | he was living in Prague still, I think |
| 10:50 | <MikeSmith> | is living in Praque, has been for a few years now |
| 10:57 | <Lachy> | JohnResig, yt? |
| 10:57 | <Philip`> | http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0D%0A%3Cstyle%3E%0D%0A.a%20%7B%20-o-link%3A%20attr(href)%3B%20-o-link-source%3A%20next%3B%20%7D%0D%0A.b%20%7B%20-o-link%3A%20attr(href)%3B%20-o-link-source%3A%20current%3B%20%7D%0D%0A%3C%2Fstyle%3E%0D%0A%3Cp%20class%3D%22a%22%20href%3D%22javascript%3Aalert(1)%22%3EClick%20me%0D%0A%3Cp%20class%3D%22b%22%20href%3D%22javascript%3Aalert(2)%22%3EClick%20me%0D%0A |
| 10:57 | <Philip`> | Uh, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/160 |
| 10:57 | <Philip`> | I never knew Opera did that |
| 10:57 | <Lachy> | JohnResig, I'm just wondering if you've had a chance to work on the selectors api test suite yet. |
| 10:59 | <Lachy> | Philip`, what does -o-link-source do? |
| 11:00 | jgraham | guesses it sets rel or something |
| 11:00 | <othermaciej> | I can report that, of the bits of HTML5 we have implemented so far, none seem to be particularly bad for phone battery |
| 11:00 | <Philip`> | Lachy: It makes the element into a link, using the -o-link of the current or next element (or something like that) |
| 11:01 | <Philip`> | (In that example, both lines use the href from the second line) |
| 11:01 | <Lachy> | ok |
| 11:02 | <Philip`> | Hmm, maybe 'next' chooses the next element (in tree order) that has -o-link |
| 11:03 | <Philip`> | so you could e.g. say <td style=-o-link-source:next><img src><a href=... style=-o-link:attr(href)>...</a></td> or whatever |
| 11:03 | <Philip`> | (to make the whole cell into a link) |
| 11:14 | gsnedders | thinks he ought to set up some wiki for himself |
| 11:14 | <gsnedders> | But I already treat my blog like a wiki in a lot of ways |
| 11:16 | <Philip`> | gsnedders: Do you have edit wars with yourself? |
| 11:16 | <gsnedders> | Philip`: Rarely, just constantly building upon what I wrote earlier |
| 11:18 | <Philip`> | gsnedders: Do you mindlessly vandalise your pages? |
| 11:19 | <Philip`> | It's not sounding much like a wiki at all |
| 11:20 | <jgraham> | Philip`: Have you /read/ gsnedders poetry? |
| 11:22 | <Philip`> | jgraham: No :-p |
| 11:25 | <gsnedders> | jgraham: in parts I agree with the mindless. I'm not sure whether it's really vandalism though. |
| 11:31 | <jgraham> | I like to think of it as mindless vandalism of the mind |
| 11:32 | <gsnedders> | Oh, but my mind was already mindlessly vandalised before writing them. |
| 12:43 | <hsivonen> | grr. why have my GWT Eclipse launch configs become unlauchable... |
| 12:44 | <JohnResig> | Lachy: hey, I haven't had a chance yet - I've needed to finish up some other things before moving on to that |
| 12:44 | <JohnResig> | Lachy: it's on my list, hoping to tackle it very soon |
| 12:46 | <Lachy> | ok |
| 12:48 | <hsivonen> | does anyone happen to know how to make Eclipse re-acknowledge the ".launch" file name extension? |
| 12:54 | <hsivonen> | Safari with <!-- a -- !> is interesting |
| 12:55 | <pesla> | hsivonen, i *think* you can just assign file associations in General -> Editors -> File Associations |
| 12:58 | <hsivonen> | heh. IE8 has the same comment buffer trimming strategy I implemented as opposed to what Hixie specced, except they haven't been careful enough with edge cases to hide the impl. details |
| 12:58 | <hsivonen> | Upon >, they trim the accumulated buffer by two from the end |
| 12:59 | <hsivonen> | pesla: can't find a way to associate it with launching as opposed to an editor :-( |
| 13:27 | <hsivonen> | aargh. I broke livedom.validator.nu |
| 13:39 | <MikeSmith> | wow. Glazman blog rendering prked in latest Minefield |
| 13:39 | <MikeSmith> | http://www.glazman.org/weblog/ |
| 13:39 | <MikeSmith> | on Linux at least, in my environment |
| 13:43 | <MikeSmith> | sayrer: you awake? |
| 14:01 | <MikeSmith> | in my minefield, glazou looks like this: |
| 14:01 | <MikeSmith> | http://people.w3.org/mike/gecko/glazou.png |
| 14:02 | <jgraham> | MikeSmith: WFM using a recentish build with html5.enable |
| 14:02 | <jgraham> | on linux |
| 14:03 | <MikeSmith> | OK |
| 14:03 | <jgraham> | Oh, my build seems to be way more out of date than I thought |
| 14:03 | <jgraham> | (2009-07-03) |
| 14:03 | <MikeSmith> | this is with very latest minefield -- today or yesterday |
| 14:03 | <MikeSmith> | and same thing with html5.enabled false |
| 14:13 | <Lachy> | jgraham, doesn't Minefield do automatic updates for you every day? |
| 14:14 | <Philip`> | MikeSmith: I see the problem - the text is yucky and un-antialiased :-( |
| 14:14 | <MikeSmith> | Lachy: you can turn that auto-update thing off |
| 14:14 | <MikeSmith> | Philip`: good eye! |
| 14:14 | <Lachy> | MikeSmith, ok. Why would you, though? |
| 14:15 | <MikeSmith> | Lachy: because it's stupid? |
| 14:15 | <Lachy> | is that a question or a statement? |
| 14:15 | <jgraham> | Lachy: It offers to upgrade every day but I often ignore it |
| 14:15 | <MikeSmith> | Lachy: are you asking or making a suggestion? |
| 14:16 | <jgraham> | Are we playing questions? |
| 14:16 | <Lachy> | Mine automatically downloads the updates and installs them whenever I restart. |
| 14:16 | gsnedders | tries to write Ruby |
| 14:17 | <Philip`> | gsnedders: Which kind of Ruby? |
| 14:17 | <Lachy> | gsnedders, that's a very short word to write, but well done. You succeeded. |
| 14:18 | <jgraham> | Didn't Lachy just lose? |
| 14:18 | <gsnedders> | ph |
| 14:18 | <gsnedders> | Philip`: The shiny one |
| 14:18 | <Lachy> | jgraham, we were playing a game? |
| 14:19 | <Philip`> | gsnedders: Animal, vegetable, or mineral? |
| 14:19 | <Lachy> | Philip`, Animal |
| 14:20 | <gsnedders> | Philip`: The shiny one. |
| 14:21 | <Philip`> | So gsnedders is writing a Ruby that is a shiny animal? |
| 14:21 | <gsnedders> | Yes. |
| 14:21 | <gsnedders> | (And I'm saying that just because Hixie can't, because he says he never lies) |
| 14:22 | <jgraham> | Couldn't that be a lie? |
| 14:28 | <gsnedders> | Hah! No wonder he's doing so well at controlling the future of the intarwebs! |
| 14:39 | <eighty4> | is <b> and <i> actually going to be left in html5? |
| 14:44 | <jgraham> | Why wouldn't they be? |
| 14:46 | <eighty4> | seems they is all ready represented by <strong> for example. And it doesn't seem to be a good example for when to use them. Or why would the specs say this: "In the previous example, the correct element to use would have been strong, not b."? |
| 14:47 | <eighty4> | uhm... nevermind that quote :) |
| 15:00 | <Lachy> | eighty4, http://lachy.id.au/log/2007/05/b-and-i |
| 15:01 | <eighty4> | Lachy: thanks :D |
| 15:06 | <eighty4> | Lachy: but given there definition how is the names of them motivated? <aside>, <article> and so on describes their intended use. How does <b> describe it's use? Or isn't that important? |
| 15:08 | <beowulf> | eighty4: how does b not describe it's intended use? |
| 15:08 | <eighty4> | Sorry if I'm bringing up already discussed stuff. Haven't read up on my html5 :/ |
| 15:13 | <jgraham> | So I have a problem. I have a test harness which loads testcases into an iframe. It is not impossible that those testcases will cause infinite loops or other bad behaviour so it would be nice to be able to set a timeout and kill the script if the timeout is reached |
| 15:14 | <jgraham> | Howwever this doesn't seem to be easy; the event loop seems to prioritise running the existing script above anything else. |
| 15:14 | <jgraham> | (run to completion I guess) |
| 15:14 | <Lachy> | eighty4, b is intended for use cases which are commenly represented typographically in bold text, but for which no specific semantic element exists |
| 15:15 | <Lachy> | eighty4, it would be unreasonable to have to create separate semantic elements to individually cover each of those cases, when it wouldn't gain any benefit over using the generic element |
| 15:15 | <jgraham> | Is there some way I force the source of the iframe to change before the testcase script has finished running? |
| 15:15 | <Lachy> | similarly for the i element |
| 15:15 | <eighty4> | beowulf: from what I understand it's not intended to represent bold text but rather like Lachy just wrote "text usually written in bold" |
| 15:15 | <jgraham> | (or close the window with the iframe in or delete the iframe or something, as long as it causes the script to die) |
| 15:16 | <Lachy> | eighty4, the names are retained for backwards compatibility reasons. It's best not to read too much into just the name of the element. It's the definition that matters mose |
| 15:17 | <eighty4> | Lachy: that explanation was good :) Not sure I'm agreeing with you but still. |
| 15:17 | <takkaria> | jgraham: I'm guessing that setTimeout() isn't being your friend |
| 15:17 | <Lachy> | jgraham, I don't know of any way to achieve that, unless perhaps setting the iframe's window.location to a new page interrupts any scripts running within the iframe |
| 15:17 | <eighty4> | Lachy: I got that backward compatiblility reason :) Just not a big fan of that either. |
| 15:18 | <Lachy> | eighty4, it's really just a matter of being pragmatic over theoretical purity |
| 15:18 | <jgraham> | Lachy: setting the <iframe>.src doesn't |
| 15:18 | <eighty4> | Lachy: have nothing to do today and haven't been reading much about html5 so I'm catching up. |
| 15:18 | <eighty4> | I'm not intending to start an argument |
| 15:18 | <Lachy> | jgraham, what about iframe.contentWindow.location? |
| 15:19 | <jgraham> | Lachy: OK I will try it |
| 15:19 | <Lachy> | eighty4, no worries. You've been much more pleasant and understanding than the person I debated this with over twitter 2 days ago |
| 15:20 | <eighty4> | Lachy: I'm not really bothered with html5 having it, but I can imagine it being missused. |
| 15:20 | <Lachy> | indeed. It will be. |
| 15:21 | <takkaria> | I'm not sure it makes any difference if it's "misused" |
| 15:21 | <takkaria> | the idea that <strong> is more accessible than <b> in any way is a red herring |
| 15:22 | <eighty4> | takkaria: but <b> will still give bold text I assume? |
| 15:22 | <takkaria> | well, yeah |
| 15:22 | <takkaria> | just like <strong> does |
| 15:22 | <eighty4> | true |
| 15:22 | <takkaria> | they are equivalent on the vast mojority of pages |
| 15:22 | <takkaria> | except those select few which change default styling for either |
| 15:23 | jgraham | mutters about how badly designed the 'this' semantics are in javascript |
| 15:24 | <jgraham> | Hmm, deleting the iframe seems to work well |
| 15:24 | <eighty4> | takkaria: but still, <strong> have semantic meaning while <b> doesn't. From what I understand |
| 15:24 | <takkaria> | they both have meaning |
| 15:25 | <eighty4> | * just not as clear semantic meaning then :) |
| 15:25 | <takkaria> | I'm not sure <strong> has a particularly clear meaning either |
| 15:25 | <eighty4> | it have a better meaning for me at least. But that doesn't really say anything |
| 15:26 | <takkaria> | they're basically synonyms, discussing which one is "better" seems to be a waste of everyone's time |
| 15:26 | <takkaria> | I seem to have acquired a stronger view on this than I used to have. :) |
| 15:27 | <beowulf> | i've developed the strong view that it just doesn't matter |
| 15:27 | <eighty4> | takkaria: I'm doing my best on wasting my time on meanless stuff right now. Third day on my "time off" |
| 15:30 | <takkaria> | I really wish people had never used the word "semantic" in conjunction with the web, "meaning" is much less opaque and you couldn't get consturctions like "semantic meaning" |
| 15:32 | <eighty4> | But I'll stop bothering you now |
| 15:33 | <Lachy> | eighty4, without <i>, which element would you use to mark up a ship name, which is conventionally rendered in italics? |
| 15:35 | <eighty4> | Lachy: <title>, <name> or something similar. But yes. It would be a challange to figure out what that element should be named |
| 15:35 | <Lachy> | so you're advocating creating a new element for each case like that? |
| 15:35 | <Lachy> | even though there are potentially dozens of such cases? |
| 15:35 | <eighty4> | no, not really. Haven thought about it that much |
| 15:36 | <eighty4> | It just feels like there should be a better name for <i> then i |
| 15:36 | <eighty4> | but I guess there isn't |
| 15:36 | <takkaria> | a better name for representating traditioanlly italicised text than "italic"? :) |
| 15:37 | <eighty4> | it seems there should ba a name for the italic "meaning" |
| 15:37 | <eighty4> | *should be |
| 15:38 | eighty4 | doesn't like his new espresso. |
| 15:38 | <eighty4> | I should buy new |
| 15:44 | <Lachy> | Reading Shelly's latest on www-archive, it seems she prefers to work under the assumption that Microsoft isn't acting in good faith and has convinced her self that they're going to object to everything as soon as we try to go to LC |
| 19:42 | <takkaria> | I really wish people would stop referring to HTML5 as "the spec to build a browser" |
| 19:43 | <sicking> | takkaria, indeed, it's not just for browsers |
| 19:43 | <sicking> | and you need a lot more to build a browser, like http and css |
| 19:43 | <takkaria> | given that I refer to it fairly often when writing JS, I think it's a wholly inaccurate title given only to rile people |
| 19:49 | <aja> | hsivonen: ping |
| 19:54 | <gsnedders> | takkaria: But do you understand the event loop? |
| 19:54 | <gsnedders> | Does anyone (apart from H) understand the event loop!? |
| 19:59 | <aja> | hsivonen: are you aware of any remaining gecko figure/legend issues with new parser? i'm looking into 1...perhaps just some default stylesheet additions needed still |
| 22:16 | <ezyang> | I'm kind of curious to know if anyone is using php html5lib |
| 22:19 | <Philip`> | ezyang: Add some code to it that pings a URL whenever someone runs the library |
| 22:19 | <ezyang> | Har har har :-P |
| 22:19 | <Philip`> | That's what the Python one does |
| 22:20 | <ezyang> | Wait, are you serious? |
| 22:20 | <ezyang> | That's not actually acceptable. |
| 22:20 | Philip` | is rarely serious :-) |
| 22:23 | Hixie | reads the htmlwg minutes and wonders if maybe it could be a prerequisite that people who ask for changes to html5 actually read it first |
| 22:24 | <Hixie> | or at least the relevant sections |
| 22:30 | <Lachy> | Hixie, which points in the telcon minutes are you referring to? |
| 22:32 | <cying> | Hixie: what fun would that be? |
| 23:35 | <Lachy> | I'm very surprised by Joshue's outright rejection of my rather practical proposal for making that particular use of canvas accessible. |
