| 00:01 | <Dashiva> | I can see the use, but at the same time it might lead to a slippery slope of more styling of titles |
| 00:01 | <othermaciej> | how does attribute parsing normally deal with whitespace in attributes? |
| 00:02 | <othermaciej> | I don't mind if it doesn't have to be a parser special case |
| 00:02 | <Hixie> | html5 preserves everything |
| 00:02 | <Hixie> | (html4 did per-attribute special casing through the fancy sgml rules) |
| 00:13 | <hsivonen> | and XML normalizes line feeds to spaces unless escaped |
| 00:21 | <Hixie> | ok so people can subscribe to the commit-watchers mailing list, and i can post to it... |
| 00:21 | <Hixie> | so why doesn't it work |
| 00:21 | <Hixie> | grr |
| 00:55 | <a-ja> | Hixie, hsivonen: details' content model says legend element required first child (which is what the conformance checker wants), but details' dom interface text says legend is *optional* 1st element. difference between conforming document and what UA should put in dom in case of non-conforming document, or ? |
| 01:20 | <Hixie> | where does the dom interface say it's optional? |
| 01:29 | <a-ja> | sorry, Hixie....was on another channel |
| 01:29 | <a-ja> | The first element child of a details element, if it is a legend element, represents the summary of the details. |
| 01:29 | <a-ja> | If the first element is not a legend element, the UA should provide its own legend (e.g. "Details"). |
| 01:29 | <Hixie> | that doesn't say it's optional |
| 01:30 | <Hixie> | it just says what happens when the element is missing |
| 01:31 | <a-ja> | gotcha...UA dom requirement vs what makes a conforming document. |
| 01:31 | <Hixie> | right |
| 01:32 | <othermaciej> | there are many cases where UAs are required to handle particular non-conforming cases |
| 01:33 | <Dashiva> | Hixie is too diligent at work. He's impeding my daily lggwg fix :) |
| 01:34 | <Hixie> | heh |
| 01:36 | <a-ja> | dunno if it's just me finding this sorta thing a wee bit confusing...from some of the html-wg comments i've been seeing. almost wondering if separate "conforming document" and "UA behavior" doc specs might help that. |
| 01:38 | <a-ja> | any clue how the vote's going so far? |
| 01:38 | <Dashiva> | These? http://www.w3.org/2002/09/wbs/40318/htmlbg/results |
| 01:39 | <othermaciej> | a-ja: I think it's true that a lot of people don't quite get it |
| 01:40 | <a-ja> | tks...wasn't aware running vote total was anywhere public |
| 01:40 | <Hixie> | a-ja: i think we'll have to have separate tutorial pages that help with that, but if you have suggestions for how to make the spec clearer please let me know |
| 01:40 | <othermaciej> | I think what is needed is a good clear explanation |
| 01:40 | <othermaciej> | not whole separate docs |
| 01:40 | <othermaciej> | for WG members |
| 01:40 | <Hixie> | yeah |
| 01:40 | <Hixie> | that too |
| 01:40 | <Dashiva> | I don't know if UA and author sections are completely separate. If they are, they could be given classes so they could be styled with e.g. different backgrounds |
| 01:41 | <a-ja> | yeah....maybe just a "how to read this spec" intro section or somesuch |
| 01:41 | <Hixie> | there is a "how to read the spec" section |
| 01:41 | <a-ja> | doh! |
| 01:41 | <Hixie> | http://www.whatwg.org/specs/web-apps/current-work/#how-to |
| 01:41 | <a-ja> | hard keeping up with you :) |
| 01:42 | <Hixie> | :-) |
| 01:42 | Dashiva | considers that spec an atheist error |
| 01:42 | <Dashiva> | *that section |
| 01:44 | <a-ja> | Hixie: abstained? sheesh! |
| 01:45 | <Hixie> | hm? |
| 01:46 | <a-ja> | you 'abstained' in the vote? |
| 01:47 | <wilhelm> | Probably the appropriate option to choose..(c; |
| 01:47 | <Hixie> | google abstained from the vote |
| 01:47 | <Hixie> | (i'm one of several people representing google) |
| 01:47 | <a-ja> | ah |
| 01:47 | <Hixie> | i figured that it would be inappropriate for me to vote for myself or for my work |
| 01:48 | <Hixie> | and google, as a company, doesn't have a strong opinion either way |
| 01:48 | <Hixie> | (we have people in the xforms wg too, e.g.) |
| 01:50 | <a-ja> | noticed any WAI-types in html-wg? |
| 01:54 | <a-ja> | my impression is that accessibility specialists are kinda conspicuous by their absense.....wonder why. |
| 02:13 | <zcorpan> | does anyone know which spec defines that <!-- is ignored by JS engines? couldn't find anything in ecma262 |
| 02:13 | <othermaciej> | zcorpan: I don't think it is in any spec |
| 02:15 | <Hixie> | it's not defined in any spec but it should be in ecma262 |
| 02:16 | <zcorpan> | http://forums.whatwg.org/viewtopic.php?t=38#179 |
| 02:18 | <zcorpan> | is ecma262 being developed? |
| 02:19 | <Hixie> | yejs |
| 02:19 | <Hixie> | yes |
| 02:19 | <Hixie> | even |
| 02:20 | <Dashiva> | As far as I know, <!-- counting as // in JS is simply a result of browsers not wanting two different JS parsers for external and internal scripts |
| 02:25 | <zcorpan> | ah. i thought it was only special on the first line. it is indeed equivalent to // |
| 02:25 | <Hixie> | correct |
| 02:25 | <Dashiva> | It's an outrage, I use the valid JS a<!--b all the time (not) :) |
| 02:26 | <Hixie> | heh |
| 02:27 | <othermaciej> | zcorpan: I'm not sure it is fixed in the latest ecma262 |
| 02:27 | <othermaciej> | zcorpan: I need to make a list of JS compat issues that are missing from or contrary to the spec |
| 02:28 | <othermaciej> | like for..in enumeration order, reserved keyword leniency, <!--, etc |
| 02:28 | <Dashiva> | othermaciej: regexp has some nasty ones |
| 02:28 | <zcorpan> | ok |
| 02:29 | <othermaciej> | Dashiva: good point |
| 02:30 | <Hixie> | hm |
| 02:30 | <Hixie> | in cross-document messaging |
| 02:31 | <Hixie> | do we want to be passing Documents around, or Windows? |
| 02:31 | <Hixie> | if we pass Documents around we have to let XSS access postMessage() |
| 02:31 | <Hixie> | rigth now i think Documents are just locked down totally, right? |
| 02:31 | <Dashiva> | Windows, please |
| 02:32 | <Dashiva> | Unless you can recovery window from document |
| 02:32 | <Dashiva> | -y |
| 02:32 | <Hixie> | you can go both ways these days |
| 02:32 | <Hixie> | window.document.defaultView == window |
| 02:33 | <Dashiva> | Then it's not critical which way. I'd rather prefer window, though. It's the natural root element when scripting is concerned |
| 02:33 | <Hixie> | need to get a hold of the opera guys |
| 02:33 | <Hixie> | since this changes their implementation |
| 02:35 | <othermaciej> | Hixie: in Safari we don't allow xss access to the document at all |
| 02:51 | <zcorpan> | http://simon.html5.org/temp/author-view-of-html5.css -- this was how i planned to hide (or mark) parts of the spec that don't apply to authors (as an alternative stylsheet)... it seems to be the simplest way to me |
| 02:53 | <othermaciej> | Hixie: what do you think about adding a toggle() operation to classList? adds the token if not present, removes it if present |
| 03:25 | <Hixie> | othermaciej: seems reasonable; send a mail? |
| 03:30 | <othermaciej> | Hixie: sure |
| 03:31 | <Hixie> | thanks |
| 03:31 | <othermaciej> | done |
| 03:49 | <Hixie> | horrah, the commit-watchers mailing list works! |
| 03:49 | <zcorpan> | yay! |
| 03:50 | <Hixie> | turns out i'd forgotten to add the e-mail address it was sending from to the commit-watchers list |
| 03:50 | <Hixie> | so the list was like "hey stop spamming me" |
| 03:50 | <Hixie> | but since i'd turned off moderation... |
| 04:13 | <Hixie> | why the fuck does mozilla support target="_new" |
| 04:13 | <Hixie> | where did they get that from |
| 04:15 | <othermaciej> | Hixie: netscape legacy I'd guess? |
| 04:15 | <Hixie> | IE doesn't do it |
| 04:15 | <othermaciej> | that's why I think it is just netscape legacy, and not a real compat issue |
| 11:17 | <annevk> | For the internal subset only entity and attribute declarations matter right? |
| 11:17 | <annevk> | Attribute declarations for attributes of type ID and entity declarations (and DTD level entity declarations) for entities used in the document |
| 11:21 | <Lachy> | I assume parameter entities would matter too |
| 11:22 | <annevk> | Default attributes too :( |
| 11:22 | <annevk> | And entities can actually contain markup? |
| 11:23 | <annevk> | Lachy, yeah, that's what I meant with DTD level entity decl... |
| 11:23 | <Lachy> | oh |
| 11:23 | annevk | is not too good with the terminology |
| 11:24 | <Lachy> | I assume external entities don't need to be fetched, though |
| 11:24 | <annevk> | no, never |
| 11:24 | <Lachy> | good |
| 11:24 | <annevk> | that concept is dropped from XML5 |
| 11:24 | <annevk> | no need to be backwards compatible with that |
| 11:36 | <annevk> | the internal subset legacy is painful |
| 11:37 | <annevk> | see the last example in http://www.w3.org/TR/xml/#sec-entexpand for instance |
| 11:41 | <Lachy> | yikes! |
| 11:41 | Lachy | wishes annevk luck and runs off to have some dinner |
| 11:45 | <annevk> | it seems doable, but people claiming XML is simple should learn it a bit better |
| 11:46 | <annevk> | also stuff like data:text/xml,<!DOCTYPE html[<!ENTITY test "<">]><html>&test;</html> not being well-formed... |
| 11:49 | <annevk> | but inventing yet another processor is clearly not desirable |
| 12:01 | <annevk> | Because of this silly stuff you also have to think about recursion |
| 12:01 | <annevk> | consider<!DOCTYPE html[<!ENTITY test "&test;">]> |
| 12:02 | othermaciej | shudders |
| 12:03 | <annevk> | or the more complicated <!DOCTYPE html[<!ENTITY test "&test2;"><!ENTITY test2 "&test;">]> |
| 12:08 | annevk | ponders about error handling for that |
| 12:10 | <annevk> | The XML spec says this about it: "A parsed entity MUST NOT contain a recursive reference to itself, either directly or indirectly." |
| 12:13 | annevk | was hoping he could just put the expansion of the entity in the inputstream |
| 12:13 | <annevk> | this seems to prohibit that |
| 12:16 | Dashiva | shakes head at http://forums.whatwg.org/viewtopic.php?t=39 |
| 12:20 | <annevk> | heh |
| 12:20 | <annevk> | that guy is funny |
| 12:21 | <annevk> | data:text/xml,<!DOCTYPE html[<!ENTITY test "<x/>">]><x x="&test;"/> is another funny case |
| 12:22 | <annevk> | (non-well-formed) |
| 12:23 | <annevk> | data:text/xml,<!DOCTYPE html[<!ENTITY test "<x/>">]><x>&test;</x> is well-formed |
| 12:23 | <annevk> | with "x>x" as tree |
| 12:24 | <annevk> | so the value of an entity needs to be parsed in context, but it needs to be parsed in a contrained environment to make sure you don't get infinite loops |
| 12:25 | <Philip`> | data:text/xml,<!DOCTYPE html[<!ENTITY test "<x>">]><x>&test;</x> in Firefox complains of "asynchronous entity" - did they just make that term up? I can't see anybody else calling it that |
| 12:25 | annevk | is not planning to introduce the eternal XML doc |
| 12:25 | <annevk> | their XML parser comes from a library |
| 12:28 | <Philip`> | Ah - in that case, did expat just make that term up? :-) |
| 12:32 | <annevk> | I suppose you just put the used entities within the contrained environment on a stack and once you encounter an entity you've seen before you either ignore it or print it out (without expanding it) |
| 12:33 | <annevk> | maybe it's an idea that if you encounter an entity you do just that... and once you're done you just feed all that into the inputstream so it can take care of <p> sequences |
| 12:33 | <annevk> | and such |
| 14:35 | <krijnh> | Typo in the WA1 draft; 1.4. Terminology, paragraph 8, 'or the node itself is there is none' => if there is no none |
| 14:35 | <krijnh> | Hixie: Ping :) |
| 14:35 | <annevk> | typos are after the spec is done material |
| 14:35 | <krijnh> | Okidoi |
| 14:36 | <krijnh> | *Okidoki |
| 14:36 | <annevk> | maybe we should have a wiki page that lists them :) |
| 14:36 | <krijnh> | Nah |
| 14:37 | <krijnh> | The spec isn't done with typos btw |
| 14:42 | <krijnh> | Is the web forms 2 project for IE still running? |
| 14:43 | <krijnh> | Seems to totally not work in IE7 |
| 14:44 | <annevk> | dunno |
| 15:00 | <SpookyET> | Someone was critical http://www.webdevout.net/tidings/2007/04/23/the-whimzical-world-of-html-5/ |
| 15:03 | <annevk> | he's certainly correct to assert that bugmode is a terrible idea |
| 15:05 | <annevk> | the parts about SGML and DTD seem to be uninformed |
| 15:06 | <annevk> | same for the rest |
| 23:24 | <zcorpan_> | Hixie: "A date or time string is a valid date or time string if the following algorithm, when run on the string, doesn't say the string is invalid." does that mean that authors have to understand the algorithm in order to know how to write their date or time strings? |
| 23:36 | <Dashiva> | "If second is not a number in the range 0 ≤ hour < 60, then fail." |
| 23:36 | <Dashiva> | Poor second, not even inside its own range |
| 23:37 | zcorpan_ | is working on http://simon.html5.org/temp/author-view-of-html5.css |
| 23:39 | <zcorpan_> | it will be a very large ruleset that |
| 23:42 | <zcorpan_> | i hope this will help authors read and review the spec |
| 23:53 | <Philip`> | zcorpan_: Looks like that might be a bit fragile when the spec changes... |
| 23:56 | <zcorpan_> | Philip`: yes, like the status markers |