| 00:01 | <kimus> | hi, can I strip unwanted tags with html5lib? |
| 00:01 | <kimus> | i'm testing it and the max I could get is a escaped tag |
| 00:03 | <annevk5> | you could try filtering the DOM |
| 00:03 | <kimus> | hi annevk5, humm, ... that sucks :-) |
| 00:04 | <annevk5> | I don't know much about the current feature set of html5lib though |
| 00:05 | <kimus> | I just wanted a python module to 'normalize' user entered HTML |
| 00:05 | <gsnedders> | I don't think there is any easy way, kimus |
| 00:05 | <gsnedders> | I could be totally wrong |
| 00:06 | <kimus> | gsnedders: http://htmlpurifier.org/ |
| 00:06 | <kimus> | works fine for php :-) |
| 00:06 | <gsnedders> | I mean in python html5lib |
| 00:07 | <annevk5> | writing a function that filters the DOM cannot be that hard |
| 00:08 | <annevk5> | though it depends on how complex you want to make it of course :) |
| 00:09 | <kimus> | annevk5: it works fine like it's know... I just dont want to escape not allowed tags |
| 00:10 | <kimus> | not allowed attributes are removed |
| 00:11 | <kimus> | html5lib removes not allowed attributes... but there's no option to remove not allowed tags :-S |
| 00:16 | <kimus> | i'm going to change the sanitize_token |
| 00:17 | <kimus> | in my class implementation |
| 00:18 | <kimus> | annevk5 and gsnedders: done :-D |
| 00:19 | <gsnedders> | :P |
| 00:24 | <kimus> | annevk5 and gsnedders: http://paste.ubuntu.com/146546/ |
| 00:24 | <kimus> | was not hard... but... |
| 00:51 | <gsnedders> | Hixie: When are you planning on doing ref now? |
| 07:03 | <MikeSmith> | Hixie: <keygen> needs to be added to the list of void elements in 8.1.2 |
| 07:05 | <Hixie> | fixed |
| 07:08 | <MikeSmith> | k |
| 07:08 | <MikeSmith> | Hixie: there are not additional constraints on the value of the challenge attribute? any string? |
| 07:10 | <MikeSmith> | s/not additional constraints/no additional constraints/ |
| 07:16 | <Hixie> | MikeSmith: no, why would there? |
| 07:17 | <MikeSmith> | Hixie: dunno. because I have no idea what it really is or what normal values people/apps would have for it |
| 07:18 | <Hixie> | it's just a string used to salt the SPKAC structure |
| 07:18 | <Hixie> | presumably to prevent replay attacks, though i don't really understand what the replay attack would be |
| 07:18 | <Hixie> | since you wouldn't have the private key... |
| 07:19 | <MikeSmith> | I see |
| 07:20 | MikeSmith | reads blooberry's http://www.blooberry.com/indexdot/html/tagpages/k/keygen.htm now |
| 07:20 | <Hixie> | that isn't particularly detailed |
| 07:20 | <MikeSmith> | I'm looking at the links |
| 07:20 | <Hixie> | good luck! |
| 07:20 | <MikeSmith> | it really sucks that AOL shut down developer.netscape.com without redirecting anything |
| 07:21 | <MikeSmith> | idiots |
| 07:21 | <MikeSmith> | ah great |
| 07:21 | <MikeSmith> | all of wp.netscape.com too |
| 07:21 | <MikeSmith> | they get a gold star for consistency at least |
| 07:25 | <Hixie> | bed time |
| 07:25 | <Hixie> | nn |
| 07:26 | <zcorpan> | morning! |
| 07:27 | <zcorpan> | <p>Alternative content</p> - http://nord-trondelag.sv.no/ |
| 07:27 | <zcorpan> | yay for copy-paste |
| 07:29 | <MikeSmith> | hsivonen: (or anybody) what ARIA attribute set would be appropriate for keygen? |
| 07:29 | <zcorpan> | MikeSmith: tabindex? |
| 07:31 | <MikeSmith> | zcorpan: ARIA has a tabindex attribute? |
| 07:31 | <MikeSmith> | tabindex is a global attribute anyway, right? |
| 07:33 | <zcorpan> | don't remember if tabindex is global in aria |
| 07:33 | <zcorpan> | but yeah |
| 07:34 | <zcorpan> | hmm maybe they removed tabindex from aria |
| 07:34 | <zcorpan> | "An implementing host language MUST provide support for the author to make arbitrary elements focusable, such as the tabindex attribute in HTML." |
| 07:34 | <MikeSmith> | that makes sense |
| 07:35 | <MikeSmith> | zcorpan: I meant from among the sets that hsivonen already has in the schema |
| 07:35 | <MikeSmith> | http://www.w3.org/html/wg/markup-spec/#implicit-aria |
| 07:35 | <MikeSmith> | maybe http://www.w3.org/html/wg/markup-spec/#common.attrs.aria.implicit.input |
| 07:35 | <zcorpan> | earlier tabindex was specified in aria because they wanted to use html4 and html5 was so long away from rec so not relevant |
| 07:35 | <MikeSmith> | ah |
| 07:36 | <MikeSmith> | so as far as keygen, isn't it essentially a specialized input control for submitting a public key? |
| 07:37 | <zcorpan> | wasn't "rude" removed from aria? |
| 07:37 | <MikeSmith> | zcorpan: rude? |
| 07:37 | <zcorpan> | aria-live = "off" | "polite" | "assertive" | "rude" |
| 07:38 | <MikeSmith> | wow |
| 07:53 | <MikeSmith> | zcorpan: per spec, <keygen keytype=""> should be valid, right? (because it defaults to "RSA") |
| 07:54 | <MikeSmith> | http://www.whatwg.org/specs/web-apps/current-work/multipage/forms.html#attr-keygen-keytype |
| 07:55 | <MikeSmith> | hmm |
| 07:55 | <MikeSmith> | "If an enumerated attribute is specified, the attribute's value must be an ASCII case-insensitive match for one of the given keywords that are not said to be non-conforming, with no leading or trailing whitespace." |
| 07:56 | <MikeSmith> | so keytype="" is not conforming |
| 08:01 | <hsivonen> | MikeSmith: not really knowing: would the closest analogue to keygen ARIA-wise input type=file? |
| 08:01 | <MikeSmith> | hsivonen: yeah, I suppose |
| 08:01 | <hsivonen> | MikeSmith: in any case, keygen is so special that overriding the role makes no sense |
| 08:02 | <MikeSmith> | hsivonen: OK, so no need to put any ARIA attributes on it at all? |
| 08:05 | <MikeSmith> | Philip`: about https://bugzilla.mozilla.org/attachment.cgi?id=371594 |
| 08:05 | <MikeSmith> | oops |
| 08:05 | <MikeSmith> | https://bugzilla.mozilla.org/show_bug.cgi?id=477564 |
| 08:07 | <MikeSmith> | might be useful to have some data on what percentage of forms have hundreds of checkboxes |
| 08:07 | <MikeSmith> | https://bugzilla.mozilla.org/show_bug.cgi?id=477564#c14 |
| 08:08 | <MikeSmith> | "Thankfully, pages with 1000 checkboxes are few and far between." |
| 08:08 | <MikeSmith> | suspect that may not be so true |
| 08:08 | <MikeSmith> | at least I'd think there are a significant number with several hundreds of checkboxes |
| 08:16 | <Philip`> | MikeSmith: The most I see is 230 |
| 08:17 | <Philip`> | MikeSmith: and there's 5 with >100 |
| 08:17 | <MikeSmith> | Philip`: OK |
| 08:17 | <MikeSmith> | thanks |
| 08:17 | Philip` | wonders if it's worth commenting on that in that bug |
| 08:18 | <MikeSmith> | Philip`: worth it if you care about the bug at all, not worth it if you don't |
| 08:18 | <MikeSmith> | I guess |
| 08:18 | <MikeSmith> | not sure how much I care about |
| 08:19 | <MikeSmith> | seems like anybody designing forms with 500 or more checkboxes shouldn't be doing that anyway |
| 08:19 | <MikeSmith> | if they care about usability at all |
| 08:21 | <Philip`> | MikeSmith: A UI like http://www.sciencedirect.com/science/journals doesn't seem too unusable |
| 08:22 | <MikeSmith> | Philip`: hmm, and loads much faster than the HTML5 spec, at least |
| 08:25 | <hsivonen> | MikeSmith: either no ARIA or only states and properties that make sense if some of them make sense |
| 08:25 | <hsivonen> | MikeSmith: the real answer depends on what the Task Force is doing |
| 08:25 | <zcorpan> | i could imagine a single-page survey with a table and lots of checkboxes |
| 08:26 | <hsivonen> | I wonder if the TF has keygen on their tables |
| 08:26 | <zcorpan> | Statement | Fully agree | Partially agree | Neither agree nor disagree | Partially disagree | Fully disagree | Don't know |
| 08:27 | <zcorpan> | 6 times the number of statements, and surveys often have hundreds of statements |
| 08:27 | <MikeSmith> | hsivonen: how about I omit any ARIA attributes from keygen schema for now, but put a "#REVISIT: ARIA attributes" comment in? |
| 08:27 | <MikeSmith> | zcorpan: yeah, true |
| 08:27 | <hsivonen> | MikeSmith: seems ok |
| 08:28 | <MikeSmith> | zcorpan: but better design is to put them into multiple pages |
| 08:28 | <MikeSmith> | I would think |
| 08:28 | <MikeSmith> | hsivonen: thanks |
| 08:28 | <zcorpan> | MikeSmith: why? |
| 08:29 | <zcorpan> | MikeSmith: it might be on the same page but on different slides implemented with javascript show/hide |
| 08:29 | <MikeSmith> | yeah, I suppose |
| 08:30 | <MikeSmith> | but I would hate to be in the middle of a hundreds-question survey and have my browser crash |
| 08:30 | <MikeSmith> | for one thing |
| 08:30 | <MikeSmith> | so that I have to go back and re-answer them all |
| 08:30 | <zcorpan> | MikeSmith: that's why firefox implemented the feature to restore the answers, no? |
| 08:30 | <MikeSmith> | or have some problem during form submission |
| 08:30 | <MikeSmith> | yeah, but I mean in general |
| 08:31 | <MikeSmith> | as far as designing across browsers |
| 08:31 | <zcorpan> | the page could do ajaxy autosave |
| 08:31 | <Philip`> | zcorpan: Those would be radio buttons, not checkboxes |
| 08:32 | <zcorpan> | Philip`: right |
| 08:34 | <MikeSmith> | hsivonen: after I added assertions-checking for img@usemap as a descendant of a|bb|button, the following test case is no longer valid: |
| 08:34 | <MikeSmith> | http://syntax.whattf.org/relaxng/tests/html5core/valid/043.html |
| 08:36 | <MikeSmith> | hsivonen: seems like it's not necessary for the img instances in that particular test to have a usemap attribute |
| 08:37 | <MikeSmith> | so may I should just remove the usemap attributes there? |
| 08:37 | <zcorpan> | or add tests that use usemap but not <a> ancestor |
| 08:38 | <MikeSmith> | zcorpan: yah, was now just thinking the same |
| 08:38 | <MikeSmith> | but not add, just change that existing test to not use <a> ancestors |
| 08:38 | <zcorpan> | not sure what it tests, the comments seem to be using pre-content model overhaul terminology |
| 08:39 | <zcorpan> | MikeSmith: it also has ismap which requires <a> ancestor |
| 08:39 | <MikeSmith> | ah |
| 08:41 | <zcorpan> | <a href><img ismap><img ismap></a> seems to be allowed per spec |
| 08:43 | <hsivonen> | MikeSmith: ok to remove the usemap |
| 08:44 | <MikeSmith> | hsivonen: k |
| 08:48 | <MikeSmith> | hmm, should not object@usemap also be classified as interactive content? |
| 08:51 | <hsivonen> | MikeSmith: topic notwithstanding, yes if img@usemap is classified as interactive |
| 08:52 | <MikeSmith> | OK, I'll raise in bugzilla |
| 08:53 | <MikeSmith> | hsivonen: for the img@usemap test, I'll delete from the existing test and just add another new test that has img with usemap but not with ismap |
| 08:54 | <hsivonen> | MikeSmith: OK. the existing tests often test too many things in one file |
| 08:54 | <MikeSmith> | yeah |
| 08:55 | <MikeSmith> | I guess I could just reduce the new test to a simple test that has a map and an img with usemap |
| 09:28 | jgraham | has a page with several thousand checkboxes |
| 09:28 | <jgraham> | Although it is not finished yet so the idea might not work in practice |
| 09:46 | zcorpan | is having difficulties understanding what html5 says should happen for <frameset rows=",86"> |
| 09:46 | <MikeSmith> | hsivonen: any clues as to why the same test case is in both tests/html5full-xhtml/valid/043.xhtml and in tests/html5core/valid/043.xhtml ? |
| 11:30 | <MikeSmith> | hsivonen: patch for htmlparser handling for <keygen> - |
| 11:30 | <MikeSmith> | http://bugzilla.validator.nu/attachment.cgi?id=76 |
| 11:31 | <MikeSmith> | tested and works as expected (my local v.nu recognizes <keygen> as a void element) |
| 11:32 | <MikeSmith> | but I realize maybe some handling also needs to be added to cases other than IN_BODY |
| 11:33 | <MikeSmith> | (though that case is the only one for which the spec makes specific mention of <keygen>) |
| 12:14 | <annevk42> | Philip`, is it easy to find more pages with broken Location headers? |
| 12:16 | <annevk42> | Philip`, it would be cool to have a "out of X amount of links Y amount of links had a Location header and of those Z amount of links had a broken Location header (relative, invalid character). List links that with broken headers: ..." |
| 12:29 | <jgraham> | http://intertwingly.net/blog/2009/04/08/HTML-Reunification |
| 12:30 | <annevk42> | Browser vendors no longer like HTML5? |
| 12:33 | <hsivonen> | I'm very annoyed by the proposal to break DOM Consistency for "language extensions". |
| 12:33 | <hsivonen> | why do "language extensions" have to go poke the "here be dragons" area? can't extensions simply not dabble with colons? |
| 12:34 | <annevk42> | And how does ubiquity-xforms qualify as language extension? Surely non-browsers would have to implement it natively short of implementing an AI-complete algorithm to figure out how it works from the attached ECMAScript library? |
| 12:35 | <jgraham> | Heh, I thought hsivonen would like that :) |
| 12:36 | <hsivonen> | annevk42: as far as I can tell, ubiquity-xforms needs to use colons only to keep up the appearances of the XForms syntax going strong |
| 12:36 | <jgraham> | annevk42: The presence of js libraries seems to blur the distinction between langauge extension and platform extension considerably |
| 12:36 | <jgraham> | e.g. <canvas> can be implemented in js for IE |
| 12:37 | <annevk42> | The inevitable result of this is everyone sending proprietary markup over the wire. |
| 12:38 | <hsivonen> | it seems to me that it's unsafe to assume that successful Firefox extension functionality wouldn't migrate to the core feature sets of browsers |
| 12:38 | <annevk42> | That "works" because of a bunch of ECMAScript libraries. |
| 12:38 | <jgraham> | I am kindof saddened that we have resorted to "consenus" politics being more important than the technical arguments that have been made |
| 12:38 | <hsivonen> | jgraham: oh, indeed |
| 12:39 | <annevk42> | hsivonen, true, but we're far way from something interoperable I think |
| 12:39 | <hsivonen> | annevk42: I mean the argument is that RDFa is a "language extension" that Firefox extensions could be sensitive to |
| 12:40 | <annevk42> | ah yeah, you mean it would make it a platform extension down the road? |
| 12:40 | <hsivonen> | annevk42: if it that setup becomes successful, why wouldn't the client-side feature migrate into Firefox itself or into the native feature set of other browsers that don't support Firefox extensions but need to support successful features? |
| 12:40 | <jgraham> | hsivonen: Indeed |
| 12:41 | <jgraham> | Although I think that is in some sense the goal |
| 12:41 | <jgraham> | Albeit not explicitly stated |
| 12:41 | <hsivonen> | if it's the goal, then the extension should play by the rules that are needed to avoid sucky browser core code down the road |
| 12:42 | <annevk42> | I hadn't reached the part around "It needs to be noted that such nodes are placed into the DOM today differently by HTML and XML parsers." yet. That's seems like a terrible solution. |
| 12:42 | <jgraham> | hsivonen: Sure. But W3C people are being told that their baby is ugly and they are trying to find ways to route around the objections rather than consider their merits |
| 12:43 | <jgraham> | It seems to me that one possible outcome of this is to break XML on the web entirely |
| 12:45 | <zcorpan> | "ARIA introduces elements and attributes ..." - aria introduces elements? |
| 12:45 | <hsivonen> | no |
| 12:46 | <hsivonen> | unless <access> is making its way into ARIA while people aren't looking. |
| 12:48 | <annevk42> | "lesser-distributed extensibility" |
| 12:48 | <annevk42> | You got to love the terms all these architecture astronauts come up with. I wonder if any normal author can still follow what this is about. |
| 12:51 | <takkaria> | I thought we had implementor consensus on <q> these days |
| 12:53 | <jgraham> | Yeah, I thought we had converged on the insane HTML4 requirements |
| 12:54 | <annevk42> | We have. |
| 12:56 | <zcorpan> | chris has still not replied about their ua style sheet |
| 12:56 | <zcorpan> | although i guess Philip` reverse engineered it anyway |
| 13:03 | <rubys1> | re <q>, that was a result of my discussions with Steven where he claimed that the platform portions of XHTML2 had consensus. He said that what is reflected in XHTML2 was based on input from the CSS working group. |
| 13:08 | <rubys> | @jgraham, re: "I am kindof saddened that we have resorted to "consenus" politics being more important than the technical arguments that have been made" |
| 13:08 | <rubys> | author conformance requirements are codified personal opinion. |
| 13:09 | <annevk42> | platform portions of XHTML2 had consensus with whom? |
| 13:09 | <jgraham> | rubys: Sure if you're talking about things like "should I quote my attributes" |
| 13:10 | <annevk42> | <q> has quotation marks in IE8, Safari, Firefox, Opera, Chrome, etc. |
| 13:10 | <rubys> | Apparently the change in the way the Q element was defined was done jointly between the (then) HTML working group and the CSS Working group. |
| 13:10 | <jgraham> | Not if you're talking about "can I invent new names with a colon in them" |
| 13:10 | <annevk42> | rubys, for the HTML <q> element or the XHTML2 <q> element? |
| 13:10 | <annevk42> | anyway, all user agents render quotation marks |
| 13:11 | <annevk42> | I'm fine with changing it, but I don't really see it happen |
| 13:11 | <rubys> | IE renders quotation marks? |
| 13:11 | <annevk42> | I just said so |
| 13:11 | <rubys> | wow, I thought it was otherwise |
| 13:12 | <annevk42> | IE8 supports "all" of CSS 2.1 and could therefore implement support for <q> |
| 13:12 | <annevk42> | and they did |
| 13:12 | <annevk42> | (against what HTML5 suggested at the time, but we changed HTML5 after that happened) |
| 13:13 | <rubys> | I just ran a test with IE8 and don't see the quotes. |
| 13:14 | <rubys> | ah, it was in IE7 standards mode |
| 13:16 | <rubys> | http://www.w3.org/MarkUp/2009/ED-xhtml2-20090205/mod-text.html#sec_9.8. |
| 13:17 | <rubys> | http://dev.w3.org/html5/spec/Overview.html#the-q-element |
| 13:17 | <annevk42> | I know XHTML2 contradicts HTML4 |
| 13:18 | <annevk42> | XHTML2 originally had a <quote> element |
| 13:18 | <annevk42> | But since their original bold design choices they're tweaking it more and more to be just like HTML4 with weird incompatible changes. See e.g. the <img> element. |
| 13:45 | <Philip`> | annevk42: Something like http://philip.html5.org/data/invalid-location.txt ? (Feel free to munge it into a nicer format) |
| 13:46 | <Philip`> | zcorpan: Like http://philip.html5.org/data/ie8-quotes.txt ? It's not really reverse engineering (since I couldn't figure out where the quoting rules were stored), it's just testing a lot of possible values :-) |
| 13:48 | <annevk42> | Philip`, cool, though that doesn't tell how many pages attempted to do a redirect in total |
| 13:49 | <Philip`> | annevk42: I don't have that information, since I only got the headers that HttpClient finally returned (or the exceptions it threw instead of returning a response) |
| 13:49 | <annevk42> | ok |
| 13:51 | <Philip`> | (http://www.dotnetdotcom.org/ says about 16% of some completely different set of URLs returned a redirect status code) |
| 13:54 | <Philip`> | (Numbers a bit meaningless anyway, because e.g. half of the errors were fixed by Yahoo Maps being fixed) |
| 13:54 | <hsivonen> | perhaps <q> isn't the most important part of the blog post and Steven isn't the best source on implmentor consensus on <q> |
| 13:55 | <annevk42> | that's a bit too logical for my taste |
| 13:56 | <gsnedders> | rubys2: Tinyurls always do that |
| 13:59 | gsnedders | needs to buy a suit… today |
| 14:01 | <Philip`> | Does it need to fit? |
| 14:07 | <gsnedders> | Philip`: I think a suit that doesn't fit isn't formal enough |
| 14:08 | <mpilgrim> | hsivonen: indeed. let's not all waste time discussing the exact shade of blue painted on the side of the nuclear bomb |
| 14:09 | gsnedders | still wants pink with blue spots |
| 14:10 | <Philip`> | "I want to explore the idea of dropping the assumption that the current HTML working group has the sole responsibility for, and absolute dominion over, authoring guidelines." - it seems more like validator developers have absolute dominion in practice |
| 14:11 | <Philip`> | (to the same extent that browser developers have absolute dominion over e.g. parsing requirements) |
| 14:12 | <Dashiva> | Frontpage and dreamweaver might have even more sway than validators |
| 14:12 | gsnedders | goes crazy |
| 14:14 | <beowulf> | re: tinyurl, what about a @alt for <a> ... |
| 14:15 | <hsivonen> | Philip`: shh! don't tell people about the absolute dominion! |
| 14:17 | <hsivonen> | https://twitter.com/jreschke/statuses/1476412112 |
| 14:18 | Philip` | agrees with jgraham about not seeing a clear distinction between platform and language features |
| 14:18 | <Philip`> | because the platform is already extensive enough that you can do pretty much anything on it |
| 14:19 | <Philip`> | e.g. you could implement <canvas> using lots of coloured <div>s, with XHR and a PNG decoder written in JS so you can implement drawImage |
| 14:20 | <Philip`> | but it's better (for performance) to move that into the platform |
| 14:20 | <mpilgrim> | don't spend too much time trying to dissect it |
| 14:20 | <mpilgrim> | it's a false dichotomy |
| 14:21 | <mpilgrim> | just a way of trying to change the subject |
| 14:21 | <Philip`> | and similarly XForms could be implemented with scripts, but that's not going to help users of ELinks or Opera Mini or NoScript or WWW::Mechanize::FormFiller |
| 14:24 | <rubys2> | *shrug* TV Raman suggested it, and a room full of people seemed to grok it |
| 14:25 | <Philip`> | mpilgrim: I'm not sure what it's trying to change the subject away from |
| 14:32 | <takkaria> | lots of intuitively simple distinctions turn out to be somewhat false, though |
| 14:32 | <takkaria> | or at least, not as simple as they at first seem |
| 14:33 | <jgraham> | It seems to be trying to be an attemp to create a false divison such that each group of interested parties can be told that their desires are being met by one part of the divison |
| 14:35 | <hsivonen> | jgraham++ |
| 14:38 | <Philip`> | rubys2: Maybe those people weren't aware that 2D graphics has been implemented as a scripting library, or that ubiquity-xforms violates fundamental assumptions of XML (e.g. if I change the prefix from "xf" to "xf2" then it stops work entirely) |
| 14:40 | <mpilgrim> | still not the most important part |
| 14:40 | <Philip`> | (and if I create new elements through scripts then it doesn't work) |
| 14:41 | <mpilgrim> | (hint: the only sentence that matters is the one with the word "king" in it. everything else is smokescreen.) |
| 14:41 | <Philip`> | (so it's kind of an interesting hack but not an adequate substitute for a real feature) |
| 14:42 | <Philip`> | s/work/working/ a few lines ago |
| 14:50 | <hsivonen> | mpilgrim: what's your take on specs that say "foo is not allowed"? |
| 14:52 | jgraham | wonders why it is made out to be a big deal whether XHTML2 and HTML5 converge or not |
| 14:54 | <Dashiva> | I could guess that XHTML2 failing would be bad PR for the XML world |
| 14:54 | <Philip`> | Is "Reunification" meant to refer to unification of HTML5 and XHTML2? |
| 14:55 | <annevk42> | Philip`, I think so, yes |
| 14:55 | <hsivonen> | Dashiva: does Sam care about bad PR for the XML world? |
| 14:55 | <annevk42> | Philip`, I'm not sure when that became a goal |
| 14:55 | <MikeSmith> | where is the word "reunification" used? |
| 14:56 | <Philip`> | MikeSmith: In the title |
| 14:56 | <Philip`> | (but nowhere else) |
| 14:56 | <MikeSmith> | which title ? |
| 14:56 | <Philip`> | MikeSmith: http://intertwingly.net/blog/2009/04/08/HTML-Reunification |
| 14:57 | <Philip`> | MikeSmith: That's literally hours old; surely you must have seen it already, via RSS and Planets and Twitter and IRC |
| 14:58 | <jgraham> | It is interesting that no one expects the same dynamics to apply in other fields. Like no one tries to reunify different programming languages |
| 14:58 | <mpilgrim> | Philip`: rubys2 talks about "unification" in the same way the romulans talk about it (c.f. http://memory-alpha.org/en/wiki/Unification_II_(episode) ) |
| 14:58 | <Philip`> | jgraham: Sure they do |
| 14:58 | <MikeSmith> | Philip`: sorry, i was too busy having dinner and drinking with a beautiful woman |
| 14:58 | <Philip`> | jgraham: C99 and C++0x have been moving towards more commonality |
| 14:58 | <MikeSmith> | I suppose it would be judged wrong of me to say that I think this whole thing was a lost cause |
| 14:59 | <MikeSmith> | (if I were to actualy say that) |
| 14:59 | <jgraham> | MikeSmith: What whole thing? |
| 14:59 | <MikeSmith> | jgraham: the marriage of heaven and hell |
| 14:59 | <Philip`> | MikeSmith: Even if it's a lost cause we'll still continue searching for it, because it might turn up in the last place we look |
| 14:59 | <MikeSmith> | proposed marriage |
| 14:59 | <jgraham> | Philip`: In some sense C++ is supposed to be a superset of C |
| 15:00 | <jgraham> | So that makes more sense |
| 15:00 | <Philip`> | Also we might find some other interesting things that fell down the back of a cupboard while we're searching |
| 15:00 | <MikeSmith> | the heaven and hell parts being different depending on your political beliefs |
| 15:00 | <jgraham> | This is more like C and lisp unifying :) |
| 15:01 | <hsivonen> | jgraham: comparing XHTML2 to lisp seems insulting to lisp |
| 15:01 | <MikeSmith> | I'm finding that sam's blog doesn't read quite the same after consuming a full bottle of wine |
| 15:01 | <Philip`> | jgraham: C++ started that way, but changed incompatibly, and C changed too (though much less), and now both languages are evolving to remove some of the differences; similarly XHTML2 and HTML5 started from the existing (X)HTML language, and evolved differently |
| 15:02 | <mpilgrim> | i find it's much more understandable if you read it out loud in the voice of the comic book guy from the simpsons |
| 15:02 | <MikeSmith> | but I do notice the word "XForms" in there |
| 15:02 | <hsivonen> | Philip`: the difference is that people actually use C and C++. A lot. |
| 15:03 | <Philip`> | hsivonen: Perhaps jgraham should have said "Like no one tries to reunify different programming languages when one is used by a billion people and the other is used by two people" |
| 15:03 | <MikeSmith> | "Any additional features required of browsers in support for XForms would have to meet a very high bar indeed" |
| 15:03 | <MikeSmith> | emphasis on the "indeed" |
| 15:04 | <MikeSmith> | damn, hsivonen, that's quite a comment |
| 15:05 | <Philip`> | MikeSmith: Indeed, since ubiquity-xforms requires browsers to be infinitely fast so you don't get an ugly flash of not-yet-munged content appearing while the scripts load |
| 15:05 | mpilgrim | feels a "translation from ruby-speak to english" coming on |
| 15:05 | <MikeSmith> | speak of the devil |
| 15:05 | <Philip`> | (Or, I suppose, browsers could implement something like XBL2) |
| 15:06 | <Philip`> | (but for text/html) |
| 15:06 | <mpilgrim> | nah, not even worth my time |
| 15:06 | <Philip`> | (though I know nothing about XBL2 except that it's the magic solution to all problems that is sometimes promoted by WHATWG people) |
| 15:06 | <MikeSmith> | rubys: we are performing hermeneutics on your blog entry |
| 15:06 | <hsivonen> | MikeSmith: what's quite a comment? |
| 15:06 | <MikeSmith> | long |
| 15:06 | <MikeSmith> | lots of words |
| 15:07 | <MikeSmith> | i think you quoted the most of the original text |
| 15:08 | MikeSmith | arrives at Yokohama and seat frees up so that I don't have to keep typing with one hand |
| 15:08 | rubys | looks up "hermeneutics" |
| 15:09 | <Philip`> | Sounds like a dirty word |
| 15:09 | <MikeSmith> | I don't know that it means either. I overheard it in Vacation Bible School |
| 15:10 | <hsivonen> | MikeSmith: it seems that it's unnecessary to repeat the code for the KEYGEN switch-case. I'll look into it carefully tomorrow. Gotta go now. |
| 15:10 | <Philip`> | MikeSmith: Definitely a dirty word, then |
| 15:10 | <MikeSmith> | hsivonen: wait 2 seconds |
| 15:10 | <MikeSmith> | please |
| 15:10 | <hsivonen> | MikeSmith: ok |
| 15:10 | <MikeSmith> | http://bugzilla.validator.nu/attachment.cgi?id=77 |
| 15:11 | <hsivonen> | MikeSmith: on the face of it, it seems that case INPUT: case KEYGEN: /* old code */ would work |
| 15:11 | <MikeSmith> | patch for the syntax part |
| 15:11 | <MikeSmith> | when you got time |
| 15:11 | <MikeSmith> | hsivonen: OK |
| 15:11 | <MikeSmith> | hsivonen: anyway, no rush |
| 15:11 | <MikeSmith> | just wanted to send you links before I forget |
| 15:11 | <hsivonen> | MikeSmith: looks good. |
| 15:11 | <MikeSmith> | k |
| 15:12 | <MikeSmith> | I'll write some tests as well |
| 15:13 | <LeifHS> | zcorpan: regarding <object> etc, what exactly does "fallback-free" mean? Where is the effect of considering object fallback-free explained? |
| 15:14 | <Philip`> | LeifHS: http://www.whatwg.org/specs/web-apps/current-work/multipage/dom.html#fallback-free |
| 15:15 | <MikeSmith> | wow, LeifHS on IRC |
| 15:15 | <LeifHS> | Mike, everyithing can happen ... |
| 15:15 | <MikeSmith> | LeifHS: yeah, that's what makes life worth living |
| 15:16 | <MikeSmith> | damn, 9 minutes of battery left, and 25 minutes to my destination |
| 15:16 | <Philip`> | LeifHS: More usefully: http://www.whatwg.org/specs/web-apps/current-work/#fallback-free and then wait a minute for everything to load and burn a load of CPU power, and then click "fallback-free" to see a list of references, which should be all its effects |
| 15:17 | <Philip`> | MikeSmith: Type slower to conserve battery power |
| 15:17 | <MikeSmith> | heh |
| 15:17 | <MikeSmith> | well, I got keygen patches for v.nu done today |
| 15:18 | <MikeSmith> | so I guess I can claim I've accomplisehd something |
| 15:18 | <MikeSmith> | and it's only 23:23 |
| 15:18 | <MikeSmith> | usually by this time of the day, I've accomplished nothing at all |
| 15:23 | <LeifHS> | Phillip: All I see is that it is said that it is considered "fallback-free". The issue I am after is what it means if an OBJECT with EMBED is considered fallback-free. That, in my mind, means that the definition of how OBJECT works changes. But to be sure, I wanted something that told me what effeect it has. |
| 15:23 | <LeifHS> | Sorry, meant "Philip". |
| 15:26 | <annevk42> | It only means something when it is said to mean something |
| 15:29 | <Philip`> | LeifHS: Clicking the definition to see references should show the effects - it appears to just be involved in document.nameofobjectelement |
| 15:30 | <Philip`> | (so you can't access non-fallback-free object elements via that mechanism) |
| 15:54 | jgraham | wonders why rubys's blog says "This comment has been blocked" if I type anything into the comment box |
| 15:54 | <jgraham> | Not that I had anything in particular to say but it is still quite unfriendly |
| 16:03 | <LeifHS> | Philip, I've read the definition, but if you have external references in min, htne have not seen them. |
| 16:08 | <Philip`> | LeifHS: It's not referred to externally from HTML5 (if that's what you mean), and the only internal references are the previous few paragraphs |
| 16:09 | <Philip`> | (and those paragraphs are describing how document.nameofobjectelement will (or will not) find the object with that name, depending on whether it's fallback-free) |
| 16:10 | zcorpan | wonders what is being unclear |
| 16:13 | <LeifHS> | Scripting is not my strong sidde, but how can we decide what should be available for scripting without a look at how UAs are treating OBJECT fallback as such? And the trouble is that <embed> is treated as a <source> element by Webkit (and partly by IE). In all the OBJECT-with-EMBED examples that zcorpan looked at, any markup beside EMBED will be revelealed if EMBED is revealed. Except in... |
| 16:13 | <LeifHS> | ...Webkit. See http://www.malform.no/html5/object+youtube. To make an OBJECT-with-EMBED be seen as fallback free, would - it seems to me - be to sanction the WebKit behaviour. (See the thread about interopable object in HTMLwg) |
| 16:14 | <rubys> | jgraham: weird |
| 16:14 | <rubys> | it shouldn't do that.... and obviously doesn't for others. |
| 16:14 | <rubys> | is it reproducible? |
| 16:17 | Philip` | wonders if rubys's blog still puts a "submit" button directly where the "preview" button one was about to click on was located at the moment one started to click one's mouse button |
| 16:17 | <zcorpan> | LeifHS: this issue actually has nothing to do with fallback. "fallback-free" is just a term that could be chosen better |
| 16:18 | <Philip`> | rubys: I get "This comment has been blocked" as soon as I type a letter into the comment box |
| 16:18 | <zcorpan> | LeifHS: it's more like "is-intended-for-internet-explorer-and-should-not-be-visible-for-other-browsers" |
| 16:18 | <zcorpan> | LeifHS: for the purposes of document.foo lookup |
| 16:19 | <jgraham> | rubys: Yes, it is reproducable. Based on IP address? |
| 16:20 | jgraham | has no way to verify the IP address thing |
| 16:23 | <jgraham> | Oh, I didn't notice that Philip` had the same problem |
| 16:28 | <annevk5> | LeifHS, <embed> _is_ exactly like <img>, <br>, <source>, etc. in all browsers |
| 16:28 | <LeifHS> | sorry - bye for now. L |
| 16:38 | <zcorpan> | it seems the xml core wg are unaware of google existing |
| 17:16 | <Philip`> | zcorpan: I thought Google was the cause of XML's failure on the web, because AdSense relies on document.write and their business model would be destroyed by XHTML - surely the WG can't be unaware of such facts |
| 17:19 | <zcorpan> | Philip`: if they are aware of google, then they don't seem to know that people use google rather than going to w3.org to find out what <?xml-model?> does upon finding it in view source |
| 17:20 | <zcorpan> | actually "xml-model" is probably not the best name because it gives unrelated results in google |
| 17:21 | <Philip`> | Maybe they think Google will point users at the relevant page on w3.org |
| 17:22 | <Philip`> | in which case somebody could tell them about w3schools.com's pagerank |
| 17:36 | <rubys> | jgraham, Philip`: I believe I've isolated the problem, and disabled that check for now until I can correct it. |
| 18:24 | <Hixie> | zcorpan: rows=",86" is treated like rows="1,86" |
| 18:29 | <zcorpan> | Hixie: not "1*,86"? |
| 18:29 | <rubys> | jgraham, Philip: I believe I've fixed the problem. My code was comparing GMT time to local times and falsely concluded that there was a DOS attack in progress. |
| 18:29 | <Hixie> | zcorpan: er yes, sorry |
| 18:29 | <zcorpan> | Hixie: ok, then that's good |
| 18:30 | <zcorpan> | Hixie: thanks for speccing out framesets |
| 18:31 | <zcorpan> | Hixie: it has already saved us the work of reverse engineering ie :) |
| 18:32 | <zcorpan> | s/the/some/ |
| 18:36 | <Hixie> | :-) |
| 18:44 | <gsnedders> | Hixie: When are you planning on doing ref now? |
| 18:48 | <gsnedders> | hsivonen: As for RelaxNG/XSD, AFAIK lxml has built-in support for XSD but not RelaxNG |
| 18:59 | <Philip`> | gsnedders: Like http://www.xml.com/cs/user/view/cs_msg/2576 ? |
| 18:59 | <Hixie> | gsnedders: ref? |
| 18:59 | gsnedders | is n00b'd |
| 18:59 | <gsnedders> | Hixie: references |
| 18:59 | <Hixie> | gsnedders: august |
| 19:00 | Philip` | makes an evidence-free assumption that that feature wasn't removed from lxml |
| 19:00 | <Philip`> | (libxml2 definitely supports RelaxNG, though) |
| 19:06 | <gsnedders> | I know that libxml2 does :P |
| 19:07 | Philip` | decides that offering to proofread someone's PhD dissertation may not have been the most effort-saving thing to do |
| 19:14 | Philip` | is now up to page 94 |
| 19:19 | <Philip`> | http://philip.html5.org/demos/xforms/example-xf.html vs http://philip.html5.org/demos/xforms/example-xf2.html |
| 20:26 | <hsivonen> | http://webbackplane.com/node/47 |
| 20:27 | <gmiernicki> | shouldnt the video on that page be in <video> as a theora .ogg? |
| 20:27 | <Philip`> | No, it should be a transcript |
| 20:27 | <gmiernicki> | some of us don't install flash ;D |
| 20:29 | <hsivonen> | gmiernicki: the publisher isn't an HTML5 proponent |
| 20:29 | <Philip`> | Some of us install Flash but only so that it works in Firefox 2 and not a browser which said person regularly uses, because it tends to crash in Opera |
| 20:30 | <smedero> | the publisher apparently isn't interested in anyone hearing that video either. |
| 20:30 | <Philip`> | (and said person hasn't figured out how to make the plugin work in his copy of Firefox 3) |
| 20:30 | <smedero> | i've got the volume jacked up to eleven... and I can barely make out portions of it |
| 20:32 | <hsivonen> | available in another encumbered format at http://blip.tv/file/897562/ |
| 20:34 | <Philip`> | Does anyone have a pointer to a script library that implements RDFa in a browser? |
| 20:35 | <Philip`> | http://www.w3.org/2006/07/SWD/RDFa/impl/js/ - oh, that |
| 20:35 | <smedero> | perhaps http://code.google.com/p/rdfquery/ |
| 20:36 | <smedero> | http://www.jenitennison.com/blog/node/94 seems to describe the project better than the google project page does |
| 20:40 | <hsivonen> | what's http://lists.w3.org/Archives/Public/public-xml-core-wg/2009Apr/att-0001/1193.htm about? |
| 20:41 | <hsivonen> | did the ISO committee decide to do shemaLocation as PI? |
| 20:41 | <hsivonen> | *schemaLocation |
| 20:41 | <hsivonen> | despite RNG not having it by design |
| 21:00 | <inimino> | so is this HTML Reunification post suggesting that libraries will implement namespace support in JavaScript? |
| 21:00 | <inimino> | including looking for and handling xmlns attributes? |
| 21:00 | <hsivonen> | inimino: so it seems |
| 21:01 | <inimino> | wow |
| 21:01 | <Hixie> | wait what? that's the most ridiculous thing i've heard all day. Who suggested that? |
| 21:02 | <inimino> | I'm trying to make sense of the paragraph that contains " Since HTML parsers are unaware of xmlns attributes, it is entirely too dangerous to encourage the use of default namespaces in conformant HTML." |
| 21:02 | <inimino> | that's the only interpretation I've considered that makes that make any sense, but I'm probably misunderstanding the whole post |
| 21:03 | <Hixie> | what does that even mean |
| 21:03 | <Hixie> | who wrote this? is this a lastweek crazy post or something? |
| 21:03 | <hsivonen> | Hixie: http://intertwingly.net/blog/2009/04/08/HTML-Reunification#c1239205696 |
| 21:04 | <inimino> | "libraries that operate on the resulting DOM will consistently evaluate namespaces after the reparenting operation." |
| 21:04 | <hsivonen> | Hixie: Mr. Last Week homed in on the crux of Sam's post: the gatekeeper and king maker sentence |
| 21:04 | <inimino> | as far as I can tell, the idea as that we can shoe-horn in distributed extensibility by just treating names with colons transparently |
| 21:05 | <Hixie> | oh, it's a blog post |
| 21:05 | <Hixie> | comment i mean |
| 21:05 | <inimino> | and then I guess the idea is that JS libraries will implement something like Namespaces in XML at a layer above that |
| 21:05 | <Hixie> | sam gets lots of people commenting crazy stuff on his blog, i wouldn't pay much attention to them |
| 21:06 | <Hixie> | (i didn't look to see who wrote the comment) |
| 21:06 | <rubys1> | that would be Sam. |
| 21:06 | <inimino> | it's not a comment, it's it Sam's post |
| 21:06 | <inimino> | s/it/in/ |
| 21:06 | <Hixie> | oh |
| 21:08 | <Hixie> | hsivonen: gatekeeper and king maker? |
| 21:08 | <rubys1> | Mr. Last Week homed into what MarkP views as the crux of the post. |
| 21:09 | <Hixie> | did mark blog also? |
| 21:09 | <Hixie> | man i really need to start reading blogs again |
| 21:09 | <rubys1> | mark showed up in IRC. |
| 21:09 | <Hixie> | oh |
| 21:09 | <Hixie> | i had assumed he was talking about lastweek too |
| 21:09 | <Hixie> | d'oh |
| 21:10 | <smedero> | see: http://krijnhoetmer.nl/irc-logs/whatwg/20090408#l-502 |
| 21:10 | <hsivonen> | rubys1: fwiw, I agree with markp on what the crux was but opted to focus on the DOM consistency thing in my comments |
| 21:10 | rubys1 | happens to think that author conformance requirements is the crux, as I stated up front, but go figure. |
| 21:10 | Hixie | sighs and goes to read sam's blog |
| 21:13 | <Hixie> | it's probably a bad sign that i don't understand what you're saying |
| 21:14 | <Hixie> | though it's news to me that the HTML WG has any power of "which features are permissible and which are not in the open web" |
| 21:14 | <Hixie> | i mean, the html5 and the whatwg is a pretty good example of how that's not true |
| 21:14 | <gsnedders> | But the evil cabal must live on! |
| 21:14 | <hsivonen> | rubys1: don't authoring conformance reqs count as gatekeeping? |
| 21:15 | <Hixie> | rdfa is another great example -- there are people, including sites like myspace, using it quite happily in html |
| 21:15 | <hsivonen> | Hixie: including w3.org! |
| 21:15 | <Hixie> | exactly |
| 21:15 | <Hixie> | anyway |
| 21:16 | <Hixie> | the rdfa discussion will be resolved in a matter of weeks |
| 21:16 | <Hixie> | it's been on the calendar for april for months |
| 21:16 | <Hixie> | and i don't understand the rest of the post! |
| 21:16 | <Hixie> | so... |
| 21:16 | <Philip`> | Depends on your definition of "resolved" |
| 21:16 | <tantek> | Hixie, html5/whatwg and microformats was just the start. now there's a whole Open Web Foundation that is "is an independent non-profit dedicated to the development and protection of open, non-proprietary specifications for web technologies." http://openwebfoundation.org/ |
| 21:16 | Hixie | goes back to trying to get <datagrid> fixed |
| 21:16 | <Hixie> | tantek: indeed |
| 21:17 | <gsnedders> | As opposed to the Web Foundation? |
| 21:17 | <Hixie> | Philip`: "To make a firm decision about" |
| 21:17 | <Hixie> | Philip`: "To find a solution to" |
| 21:20 | <hsivonen> | Hixie: still trying to make <datagrid> a huge replaced element instead of shipping a MIT-licensed JS lib implemented on top of <table> & ARIA? |
| 21:20 | <Hixie> | yes |
| 21:20 | <Hixie> | except not "huge" |
| 21:20 | <Hixie> | :-) |
| 21:20 | <Philip`> | Hixie: Depends on whether your definition of "resolved" means just from your perspective, or from the perspective of the HTML WG and the people who will disagree with your decision and won't let the matter be closed |
| 21:21 | <Hixie> | Philip`: i mean from the perspective of the WHATWG, since this is #whatwg. It's still not clear to me how things are resolved in the HTMLWG. |
| 21:23 | <Philip`> | Most IRC discussion about the HTML WG occurs in here, so it's good to clarify which is meant :-) |
| 21:24 | <Hixie> | Philip`: granted :-) |
| 21:24 | <hsivonen> | fwiw, the comment thread branches into http://edward.oconnor.cx/2009/04/DOM-Consistency |
| 21:25 | <smedero> | rubys1: "I’d like to retain intact the design principles (link to http://www.w3.org/TR/html-design-principles/) for platform design as worked out by the browser vendors." Perhaps I'm being nitpicky but FWIW, I thought these were worked out by the HTML WG as a whole in April 2007 and WG consensus to publish them was polled in Nov 2007. |
| 21:26 | <rubys1> | smedero: that's a fair point |
| 21:28 | <rubys1> | refresh |
| 21:29 | <smedero> | ahh, good |
| 21:50 | gsnedders | stabs Google Books |
| 21:50 | gsnedders | wants to see page 215 of urn:isbn:9780520220805 |
| 21:55 | <Hixie> | buy it |
| 21:55 | <Hixie> | :-) |
| 21:55 | <jcranmer> | meh, torrents are easier |
| 21:56 | <gsnedders> | Hixie: I made the mistake of going to not one, but three book stores today. I think I have too much already. :) |
| 21:56 | <gsnedders> | Hixie: Also, I need it just to check reference from secondary source, so I'm just going to reference the secondary source instead :P |
| 21:56 | <Hixie> | heh |
| 21:57 | <gsnedders> | (Basically it contains Nabokov saying in a letter he's working on something called, "The Kingdom by the Sea" — it would later become "Lolita".) |
| 21:57 | <gsnedders> | (The original title is a reference to Poe's "Annabel Lee".) |
| 21:58 | gsnedders | goes back to working on his English dissertation |
| 21:58 | gsnedders | sighs |
| 22:05 | gsnedders | tries to make sense of his English teachers scribbling on his draft |
| 22:05 | <gsnedders> | This would be simpler if it weren't for the fact that he often ended up scribbling down things about what we had ended up tangentially discussed and thus aren't entirely obviously related. |
| 22:12 | <Philip`> | gsnedders: Can't you just cite Wikipedia? It saves a lot of this bother |
| 22:19 | <Hixie> | blog admins, i changed from wp-cache to wp-super-cache and added wp-spamfree |
| 22:19 | <Hixie> | let me know if it breaks anything |
| 22:19 | <Dashiva> | wp-ultra-cache coming next fall |
| 22:59 | <annevk5> | it's funny that when ubiquity-xforms is put to test it fails |
| 22:59 | <annevk5> | the disturbing thing is of course that some day they get all the details right (if they at all care) and claim we should support it because it's simple |
| 23:02 | <Hixie> | the day they get all the details right will be the day it looks like wf2 :-) |
| 23:02 | Hixie | ducks |
| 23:03 | <Hixie> | (wf2 is what resulted from our attempt to adapt xforms for html, so i'm not really joking) |
| 23:07 | <gsnedders> | Hixie: What you need to fight spam is escape @name and @id on the comment form using entities. Spam bots are dumb. |
| 23:07 | <gsnedders> | (That's basically all the spam protection on my blog. Works awesomely.) |
| 23:08 | <Hixie> | i don't particularly care about fighting spam, i was just enabling the plugins dreamhost asked me to enable |
| 23:09 | <gsnedders> | Hixie: What? Is spam like CSRF and not a problem for the web? |
| 23:56 | <takkaria> | neither of Philip's demos worked for me |
| 23:57 | <takkaria> | (the xforms ones) |