| 00:32 | <Hixie> | annevk: one bug per issue please, btw :-) |
| 01:03 | <GPHemsley> | AryehGregor: Do you have an example of a Wikipedia article that you think warrants a lang attribute it doesn't have? |
| 01:03 | <GPHemsley> | AryehGregor: And also one of a non-standard code? |
| 01:04 | <AryehGregor> | GPHemsley, I don't know what you think warrants a lang attribute. Go to any English Wikipedia article on a French phrase or whatnot and there will be foreign-language content that's not marked up, almost certainly. |
| 01:04 | <GPHemsley> | Fair enough. |
| 01:04 | <AryehGregor> | Or just look at the language links in the lower left. |
| 01:04 | <AryehGregor> | <li class="interwiki-ar"><a href="http://ar.wikipedia.org/wiki/">العربية</a></li>; |
| 01:05 | <GPHemsley> | Ah. Hmm/ |
| 01:05 | <AryehGregor> | As for non-standard codes, view-source:http://zh-yue.wikipedia.org/wiki/頭版 |
| 01:05 | <AryehGregor> | Er. |
| 01:05 | <AryehGregor> | That's Chrome-specific, I think. |
| 01:05 | <AryehGregor> | http://zh-yue.wikipedia.org/wiki/頭版 |
| 01:05 | <AryehGregor> | <html xmlns="http://www.w3.org/1999/xhtml"; xml:lang="zh-yue" lang="zh-yue" dir="ltr"> |
| 01:05 | <GPHemsley> | nah, that's what Firefox uses, too |
| 01:05 | <AryehGregor> | Well, anyway. |
| 01:06 | <AryehGregor> | The correct code there is yue, not zh-yue, as far as I understand. |
| 01:06 | <AryehGregor> | There are others too. |
| 01:06 | <GPHemsley> | Nope, it's zh-yue |
| 01:06 | <GPHemsley> | Type: language |
| 01:06 | <GPHemsley> | Subtag: yue |
| 01:06 | <GPHemsley> | Description: Yue Chinese |
| 01:06 | <GPHemsley> | Added: 2009-07-29 |
| 01:06 | <GPHemsley> | Macrolanguage: zh |
| 01:07 | <GPHemsley> | hmm |
| 01:07 | <AryehGregor> | Pretty sure not. |
| 01:07 | <GPHemsley> | yeah, lemme look that up further |
| 01:07 | <AryehGregor> | ISO 639-3 code is yue. |
| 01:07 | <GPHemsley> | Oh, it's valid... just obsolete |
| 01:07 | <GPHemsley> | Type: extlang |
| 01:07 | <GPHemsley> | Subtag: yue |
| 01:07 | <GPHemsley> | Description: Yue Chinese |
| 01:07 | <GPHemsley> | Added: 2009-07-29 |
| 01:07 | <GPHemsley> | Preferred-Value: yue |
| 01:07 | <AryehGregor> | http://validator.nu/?doc=http://zh-yue.wikipedia.org/wiki/頭版&profile=permissive |
| 01:07 | <GPHemsley> | Prefix: zh |
| 01:07 | <GPHemsley> | Macrolanguage: zh |
| 01:08 | <AryehGregor> | "Bad value zh-yue for attribute lang on element html: Found reserved language extension subtag." |
| 01:09 | <AryehGregor> | http://validator.nu/?doc=http://www.wikipedia.org&profile=permissive |
| 01:09 | <AryehGregor> | That has a bunch more. |
| 01:09 | <AryehGregor> | Some are just warnings, though. |
| 01:09 | <AryehGregor> | And some might be bogus. |
| 01:10 | <GPHemsley> | Yeah, it seems that the parser doesn't allow extlang values |
| 01:10 | <GPHemsley> | s/allow/account for/ |
| 01:11 | <AryehGregor> | I filed a bug. |
| 01:12 | <GPHemsley> | k |
| 01:12 | <GPHemsley> | URL? |
| 01:13 | <AryehGregor> | http://bugzilla.validator.nu/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=Simetrical%2Bvalidator⊙gc&emailtype1=exact&emailassigned_to1=1&emailreporter1=1 |
| 01:19 | <AryehGregor> | Hixie, so, if standards shouldn't allow anything that isn't itself standard, why does <object> validate? (I assume it does, anyway.) |
| 01:23 | <GPHemsley> | AryehGregor: Alright, I've CC'd myself on those related bugs |
| 01:23 | <GPHemsley> | AryehGregor: http://meta.wikimedia.org/wiki/List_of_Wikipedias |
| 01:24 | <AryehGregor> | GPHemsley, what about it? |
| 01:24 | <GPHemsley> | That helps to sort out which of the errors are real errors |
| 01:24 | <AryehGregor> | How? |
| 01:24 | <GPHemsley> | as it gives the English name of what the Wikipedia is for |
| 01:24 | <GPHemsley> | to help match against the IANA registry |
| 01:24 | <AryehGregor> | Ah. |
| 01:25 | <GPHemsley> | zh-yue is Cantonese |
| 01:25 | <AryehGregor> | Well, yes, I knew that. |
| 01:25 | <GPHemsley> | sh is inded Serbo-Croatian |
| 01:25 | <GPHemsley> | indeed |
| 01:25 | <GPHemsley> | so the tag is right |
| 01:26 | <GPHemsley> | despite the comment that "sr, hr, bs are preferred for most modern uses" |
| 01:27 | <GPHemsley> | The one that interests me is the one that doesn't use a valid tag at all, for http://en.wikipedia.org/wiki/Emilian-Romagnol_language |
| 01:27 | <GPHemsley> | Wikipedia uses eml |
| 01:27 | <AryehGregor> | #mediawiki-i18n is the place to ask if you're curious here. |
| 01:28 | <AryehGregor> | The overwhelming majority of Wikipedias do use valid, current language codes. |
| 01:28 | <AryehGregor> | The remaining ones are theoretically being renamed over time, but it's a hassle to do the rename, and I don't think any have actually happened in the last year or two. |
| 01:29 | <GPHemsley> | doesn't seem too hard... just change the domain and have the old one redirect, no? |
| 01:30 | <AryehGregor> | I imagine so. I'm not a Wikimedia sysadmin, so I dunno. Anyway, they haven't gotten done, from what I've seen. |
| 01:30 | <AryehGregor> | But I don't follow that too closely. |
| 01:46 | <AryehGregor> | Okay, I finally realized that the correct translation of "decentralized extensibility" is "allowing the use in HTML of nonstandard, proprietary technologies that can possibly only be processed by a single vendor's tools". I will post accordingly from now on. |
| 01:46 | <AryehGregor> | (starting with my last post, in fact) |
| 01:52 | <cardona507> | I am reading http://dev.w3.org/html5/html-author/ and it says "The manifest attribute gives the address of the document’s application cache manifest, if there is one. If the attribute is present, the attribute’s value must be a valid URL." can it be a path like <html manifest="manifest/foo.manifest"> or must if be a URL like <html manifest="http://www.example.com/manifest/foo.manifest">; ? |
| 03:30 | <Hixie> | AryehGregor: hm? |
| 04:15 | <GrantG> | Does anyone know where to report z-index issues when applied to links on IE8? |
| 04:16 | <GrantG> | IE8 seems to have issues with CSS's :hover property when used on links that are positioned in a z-index |
| 04:17 | <GrantG> | hello? |
| 04:18 | GrantG | leaves the address of the site experiencing problems in IE8 and below (and only in IE) as http://help.grantgalitz.com/index.php |
| 04:19 | <GrantG> | The CSS is correct, though the use of proprietary extensions is true. |
| 04:19 | <GrantG> | CSS3 in application/xhtml+xml and text/html mode shows up the same for the site. |
| 04:21 | <GrantG> | leaving |
| 04:21 | <GrantG> | (I guess everyone is asleep or dead)... |
| 04:21 | <GPHemsley> | ...or you need to wait longer than 5 minutes for a response |
| 04:22 | <erlehmann> | that. and well, the WHATWG is certainly not the first place i'd choose to bitch about IE |
| 04:22 | <GPHemsley> | Does anyone know of any reason why Safari wouldn't center an image that has the usual margin/text-center CSS, while Firefox does? |
| 04:22 | <GPHemsley> | err, text-align |
| 04:23 | <erlehmann> | GPHemsley, images are inline replaced content, aren't they? akin to inline-block, so i'd expect the firefox behaviour to be correct, then. |
| 04:23 | <GPHemsley> | erlehmann: No idea. I just want to know how to fix it. ;) |
| 04:24 | <erlehmann> | GPHemsley, provide example |
| 04:24 | <erlehmann> | also, what hinders you to make the image display:block and apply auto margin ? |
| 04:26 | <GPHemsley> | http://gphemsley.pastebin.com/d35994e89 |
| 04:27 | <GPHemsley> | erlehmann: Hmm... hadn't thought of changing the display. Hang on |
| 04:27 | <GPHemsley> | Yup, that works |
| 04:28 | <erlehmann> | GPHemsley, be aware that text-align: center; on the image means nothing |
| 04:29 | <GPHemsley> | erlehmann: Perhaps that's where Firefox and Safari are in conflict? |
| 04:29 | <GPHemsley> | Or it's the display property |
| 04:29 | <GPHemsley> | not sure |
| 04:30 | <erlehmann> | further research is necessary. try a normal image with text-align on the parent element, i believe that is most likely |
| 04:36 | <GPHemsley> | meh, I don't have the patience to right now :/ |
| 04:48 | <erlehmann> | GPHemsley, if you dont pinpoint the error and report it, it will never be fixey |
| 04:48 | <erlehmann> | d |
| 04:48 | <GPHemsley> | I know, but I don't use Safari :P |
| 04:56 | <erlehmann> | use another webkit instead |
| 04:56 | <erlehmann> | or try browsershots.org |
| 05:48 | <Hixie> | Philip`, hsivonen: any idea if we need the <!--/--> magic in <textarea>? |
| 06:27 | <zcorpan> | "Page start-up performance for this kind of behaviour is important." - http://www.w3.org/mid/104E6B5B6535E849970CDFBB1C5216EB061F0F28⊙Trcmc |
| 06:27 | <zcorpan> | then i guess the perf impact on parsing should be considered |
| 06:44 | <Hixie> | should i just replace the content model flag with tokenizer modes? |
| 06:44 | <Hixie> | that is, when the tree constructor changes the content model flag, just have it change the tokenizer's state? |
| 07:10 | <hsivonen> | Hixie: yes, please. That's what I already do. |
| 07:10 | <hsivonen> | tokenizer states, that is |
| 07:10 | <hsivonen> | not tokenizer "modes" |
| 07:11 | <hsivonen> | Hixie: I don't have any data suggesting that the <!--/--> magic were needed in textarea |
| 07:11 | <hsivonen> | Hixie: I have data suggesting it's harmful in xmp |
| 07:12 | <hsivonen> | well, more like a datum than data |
| 07:28 | <Hixie> | k well i guess i'll just avoid doing it at all outside <script> for now |
| 08:49 | <Hixie> | this sure is a lot of new states |
| 08:49 | <cying> | Hixie: didn't know if you saw this ;) http://www.satine.org/archives/2009/10/16/wuh-working-group/ |
| 08:49 | <cying> | i couldn't resist |
| 08:49 | <Hixie> | i was there :-) |
| 08:49 | <Hixie> | oh you mean saw that you had blogged it |
| 08:50 | <cying> | ya |
| 08:50 | <Hixie> | yes, my alerts mentioned it |
| 08:50 | <cying> | heehee |
| 08:53 | <hsivonen> | Hixie: when a state needs to gather something like UBLIC or YSTEM, I have those as one state plus index into the string |
| 08:54 | <hsivonen> | I'd probably do <script and </script the same way |
| 08:57 | <roc> | Philip`: I think I've found a mistake in your tests |
| 08:58 | <annevk> | hmm, I was at like 400 emails yet to deal with and now it's 670 |
| 09:00 | <annevk> | in other news, there's no implementation of XHR responseXML yet |
| 09:01 | <annevk> | I'm still not quite sure whether responseXML is specified in sufficient detail |
| 09:19 | <Creap> | swedish is good, we just say "v" instead of "double-u" |
| 09:20 | <Creap> | I say what VG |
| 09:20 | <Creap> | and vvv = www |
| 09:20 | annevk | says what-wii-gee |
| 09:21 | <Creap> | I met a guy who called the DOM "DOOM" |
| 09:21 | <Creap> | quite fitting sometimes |
| 09:31 | <Philip`> | Hixie: I'm afraid your question was not of the form "can you run this regexp on your collection of pages?" and therefore I'm unable to answer it :-p |
| 09:32 | <Philip`> | roc: That sounds quite possible - which things seem broken? |
| 09:33 | <roc> | as named in our tests, test_2d_shadow_composite_3() and test_2d_shadow_composite_4() |
| 09:34 | <roc> | basically you do "ctx.shadowColor = '#f00';" and expect a shadow to be drawn |
| 09:34 | <roc> | but the spec currently says that a shadow should be drawn only if at least one of shadowBlur, shadowOffsetX and shadowOffsetY are nonzero |
| 09:34 | <Philip`> | Ah, yes, that's because the spec changed |
| 09:34 | <roc> | yeah, that's what I thought |
| 09:34 | <Philip`> | (and I haven't updated the tests) |
| 09:36 | <Philip`> | At some point I need to go over all the canvas spec changes from the past year or so, and fix the tests |
| 09:37 | <roc> | ok |
| 09:38 | <roc> | thanks |
| 09:40 | <Philip`> | (It hasn't really been a high priority for the past year or so, so I've not got around to it yet :-( ) |
| 09:45 | <zcorpan> | Philip`: to rephrase Hixie's question, can you run the html parser and check for the textarea elements that contain the text "</textarea" case-insensitively |
| 09:45 | <zcorpan> | Philip`: and similarly for all other (r)cdata elements |
| 09:46 | <Philip`> | zcorpan: Using the same (r)cdata parsing that's implemented by validator.nu a few months ago? |
| 09:46 | <zcorpan> | Philip`: yeah |
| 09:47 | <Philip`> | Can do, though it's harder to set up than a regexp search, so I'll try to do it some time later today |
| 09:49 | <zcorpan> | Philip`: ideally with the output url, element name, textContent for each matching element |
| 09:51 | <zcorpan> | Philip`: whether all in one page or separate pages for each element, whatever is easier to set up |
| 09:51 | <zcorpan> | s/element/element type/ |
| 09:56 | <Philip`> | hsivonen: "I choose not to support splitting it out merely in order to show confidence." - I think that sentence is ambiguous, and I think I parsed it wrong when I originally read it |
| 09:57 | <Philip`> | (I thought it meant "Merely in order to show confidence, I choose not to support splitting it out") |
| 09:57 | <hsivonen> | Philip`: English needs parentheses for grouping. |
| 09:57 | <Philip`> | (but it seems it was intended differently) |
| 09:57 | <hsivonen> | yes, it was intended differently |
| 09:58 | <Philip`> | (Maybe "I don't choose to support splitting it out merely ..." is less ambiguous) |
| 09:58 | <hsivonen> | yeah |
| 10:20 | <annevk> | I don't understand why we need namespaces for something classes handle perfectly |
| 10:20 | <annevk> | i.e. grouping |
| 10:29 | <erlehmann> | so i just made elinks acutely aware of the <video> object |
| 10:30 | <erlehmann> | but i somehow doubt they'll drag libtheora along ;) |
| 10:31 | <roc> | aalib |
| 10:33 | <erlehmann> | roc, aalib wont do anything without decoding capabilities |
| 10:57 | <Hixie> | hsivonen: i've recast just the RCDATA mode as new states: http://www.whatwg.org/specs/web-apps/current-work/#rcdata-state |
| 10:57 | <Hixie> | hsivonen: is it ok? |
| 11:19 | <hsivonen> | Hixie: RCDATA looks sensible at least after a superficial look |
| 11:19 | <mikekell1> | Hi guys - is this Smylers guy heavily involved in the working group? |
| 11:19 | <Hixie> | hsivonen: k |
| 11:19 | <hsivonen> | Hixie: are the script states still being edited? |
| 11:19 | <Hixie> | yes |
| 11:20 | <Hixie> | actively |
| 11:20 | <mikekelly> | ok so he is giving me the 'official party line' ? |
| 11:20 | <mikekelly> | :) |
| 11:20 | <Hixie> | RAWTEXT is done though |
| 11:22 | <mikekelly> | I think the 'interventionism' stuff is pretty disturbing, I have to say |
| 11:23 | <mikekelly> | if you don't like the pain of interop then perhaps writing web architecture related specs are not for you, eh? |
| 11:24 | <mikekelly> | hiding your subjective opinion behind some brittle notion of 'user experience pragmatism' is a little dubious. |
| 11:24 | <mikekelly> | to say the least. |
| 11:25 | <mikekelly> | and more than a little risky given that the entire planet will have to put up with what you produce for the next.. xyz years |
| 11:25 | <Hixie> | hsivonen: ok, the script stuff should now match what it was before, just in states instead of flags |
| 11:26 | <Philip`> | mikekelly: I haven't read this discussion much, but Smylers generally seems to say things that seem sensible and reasonable |
| 11:26 | <hsivonen> | Hixie: do you intend to edit it now like zcorpan requested? |
| 11:26 | <mikekelly> | ok well he's quoting and contesting points that are either clarified or justified underneath |
| 11:27 | <Hixie> | after checking this in, yes |
| 11:27 | <hsivonen> | ok |
| 11:27 | <mikekelly> | Philip`: he's not being particularly constructive. |
| 11:27 | <Philip`> | so if there was such a concept as an official party line, he probably would be roughly aligned with that |
| 11:27 | <Hixie> | first i want to convert it to this method |
| 11:27 | <Hixie> | then i will change it |
| 11:27 | <Philip`> | (which is not true of all HTML WG members) |
| 11:27 | <Hixie> | so that hopefully we can isolate the mistakes in the conversion from the mistakes in the additions |
| 11:28 | <mikekelly> | is 'authors' an official HTML term? |
| 11:28 | <hsivonen> | mikekelly: to the extent we have anything official, yes |
| 11:28 | <Philip`> | mikekelly: It's the term used to refer to people who produce HTML |
| 11:28 | <mikekelly> | well I think that term is crap and shuold be clarified |
| 11:28 | <Philip`> | (vs users, who are just viewing stuff through user agents) |
| 11:29 | <mikekelly> | web content is barely ever 'authored' in html |
| 11:30 | <mikekelly> | users complete forms which render to areas of pre-defined templates |
| 11:30 | <hsivonen> | mikekelly: for the purposes of the spec, "author" is the person who writes the templates |
| 11:30 | <mikekelly> | ok well that is not a very clear term then |
| 11:30 | <mikekelly> | why not use something more sensible like 'provider' |
| 11:31 | <mikekelly> | I don't know, whatever, there's got to be better terms than author |
| 11:31 | <hsivonen> | mikekelly: author is pretty established |
| 11:31 | <mikekelly> | I don't think so |
| 11:31 | <Philip`> | 'provider' means lots of other things in this context, like hosting provider and internet service provider and whatever |
| 11:32 | <mikekelly> | what? |
| 11:32 | <mikekelly> | how are those entities relevant to the HTML spec? |
| 11:33 | <Philip`> | They're relevant to discussions around the spec, e.g. "my provider won't let me set MIME types for my video files" |
| 11:33 | <daedb|> | 'author' is a lot more clear than 'provider' |
| 11:33 | <mikekelly> | no it isn't. lol. |
| 11:33 | Philip` | wouldn't object to changing the term "author" is there was something that seemed significantly better |
| 11:33 | <Philip`> | s/is/if/ |
| 11:33 | <mikekelly> | if you look at web pages like they are flat files - then fine |
| 11:34 | <mikekelly> | that is not how a large majority of modern content providers output html content |
| 11:34 | <mikekelly> | fact. |
| 11:35 | <mikekelly> | author has personified conotations for a start |
| 11:37 | <mikekelly> | what if my HTML is rendered automatically from properties of Model atrributes in an MVC application |
| 11:38 | <annevk> | the author would be the person that wrote the code that generates the HTML |
| 11:38 | <mikekelly> | right... |
| 11:38 | <mikekelly> | that's perfectly clear |
| 11:38 | <mikekelly> | :P |
| 11:38 | <daedb> | I'm not saying 'author' is a perfect term, but it's more obvious than 'provider' at least. |
| 11:38 | <mikekelly> | yeah ok I didnt put all that much thought into that one |
| 11:39 | <annevk> | it is to me |
| 11:39 | <mikekelly> | what about the poor idiot html authors that need patronising and protecting from the evils of 'The Experts' ? |
| 11:41 | <mikekelly> | you can't include something just because it's useful |
| 11:41 | <mikekelly> | it has to be 100% idiot proof |
| 11:41 | <mikekelly> | apparently. |
| 11:41 | <annevk> | you're not making much sense |
| 11:41 | <mikekelly> | I couldn't agree more. |
| 11:42 | <mikekelly> | I quote: |
| 11:43 | <mikekelly> | Content negotiation could succeed if only those who know what they are |
| 11:43 | <mikekelly> | doing touch it, that typical authors aren't somehow tempted to start |
| 11:43 | <mikekelly> | playing with it. That's possible, but not certain. I don't know how |
| 11:43 | <mikekelly> | we'd gather data either way. |
| 11:43 | gsnedders|work | notes mikekelly hasn't replied to Larry's email |
| 11:43 | <annevk> | can you please take your conneg bs elsewhere? |
| 11:44 | <mikekelly> | excuse me? |
| 11:44 | <annevk> | even co-authors of HTTP and URI disagree with you |
| 11:44 | <mikekelly> | ok good for them, then they can explain coherently why what I am suggesting is wrong |
| 11:44 | <mikekelly> | rather than explaining it's not done now |
| 11:44 | <mikekelly> | which I am fully aware of |
| 11:44 | <erlehmann> | =_= |
| 11:46 | <mikekelly> | annevk: I thought you were all against this appeal to authority stuff? |
| 11:46 | <mikekelly> | no? |
| 11:46 | <mikekelly> | I didn't get the memo. |
| 11:47 | <Hixie> | christ, describing the new states zcorpand wants in terms of authoring conformance criteria is going to be very exciting |
| 11:47 | <Hixie> | zcorpan, even |
| 11:47 | <Hixie> | oh, i know, i can just make </script always a parse error |
| 11:47 | gsnedders|work | tries to check the latest commit |
| 11:47 | <Hixie> | when it would be ignored |
| 11:47 | <Hixie> | oh no wait that won't work |
| 11:47 | <Hixie> | gah, this is going to be a nightmare |
| 11:48 | <Hixie> | the rule has to be that the element's text cannot end after a <!--<script> sequence that hasn't been matched by a --> or a </script>--> or something |
| 11:48 | <Hixie> | except the <!-- and <script> are independent... |
| 11:48 | <Hixie> | hsivonen: good luck describing this in your UI |
| 11:49 | <Hixie> | i'll write this up tomorrow |
| 11:49 | <Hixie> | i |
| 11:49 | <Hixie> | am too tired to think straight enough to do this |
| 11:49 | <Hixie> | nn |
| 11:49 | gsnedders|work | reads the commit very quickly, and thinks it doesn't break anything |
| 11:49 | <annevk> | g'night |
| 11:50 | <hsivonen> | nn |
| 12:26 | <annevk> | kind of annoying that the CSS WG never dealt with the CSSOM, nothing makes much sense |
| 12:27 | <annevk> | e.g. [*|ATTR] in a text/html resource ends up as [*|attr] in WebKit and [attr] in Firefox |
| 12:28 | <annevk> | also as [abbr] in Opera |
| 12:30 | <annevk> | in Opera type selectors preserve their case, but when there's a default namespace they serialize to "|type" which is quite the opposite from what you want |
| 12:44 | <mikekelly> | gsnedders|work: I responded to Larry :P |
| 13:04 | <Dashiva> | http://lists.w3.org/Archives/Public/public-html/2009Oct/0608.html |
| 13:04 | <Dashiva> | I'm not sure how that relates to @namespace in CSS... |
| 13:12 | <annevk> | I'm not sure why I let myself get dragged into that discussion |
| 13:13 | <annevk> | Potentially because I hit a small wall in CSSOM work |
| 13:32 | <mikekelly> | is it possible for a document to have a request property with the details of the HTTP request used to access it? |
| 13:35 | <mikekelly> | so javascript can view it and work out what's going on |
| 14:08 | <karlushi> | http://woorkup.com/2009/10/17/most-used-html-tags-in-50-popular-web-sites/ |
| 14:08 | <karlushi> | Most Used HTML Tags in 50 Popular Web Sites |
| 14:36 | <hsivonen> | Hixie: clearly, you need to make it clear that it's ok to use data-* attributes for passing data to JS libraries imported by the page |
| 14:40 | <Lachy> | hsivonen, who doesn't think it's ok to do that? I thought it was already clear |
| 14:41 | <hsivonen> | Lachy: see Shelley's and Leif's emails to public-html |
| 14:43 | <Lachy> | they've sent a lot of mails, and unfortunately, I'm quite far behind on reading mailing lists. |
| 14:44 | <hsivonen> | there seems to be quita bit of filibustering going on |
| 14:44 | <Lachy> | oh, found this one from shelley? http://lists.w3.org/Archives/Public/public-html/2009Oct/0613.html |
| 14:44 | <hsivonen> | Lachy: that one and Leif's follow-up |
| 14:44 | <Lachy> | Not sure what she means by "they're not meant to be used for anything other than by an author for a single author's needs." |
| 14:46 | <hsivonen> | the four people who've sent the most email to public-html this month are Maciej, Hixie, Julian and Shelley. |
| 14:46 | <Dashiva> | Is Leif number five? :) |
| 14:47 | <hsivonen> | I might be. |
| 14:47 | <hsivonen> | let's see |
| 14:48 | <Lachy> | hmm, it seems Lief might be misunderstanding "These attributes are not intended for use by software that is independent of the site that uses the attributes." as including to 3rd party JS libraries used on the page. |
| 14:48 | <Lachy> | though, I'm not sure. It's not entirely clear what Leif means. |
| 14:49 | <karlushi> | http://www.w3.org/TR/html5/dom.html#embedding-custom-non-visible-data |
| 14:49 | <hsivonen> | Dashiva: yeah, Leif seems to be #5 |
| 14:49 | <karlushi> | >Custom data attributes are intended to store custom data private to the page or application, for which there are no more appropriate attributes or elements. |
| 14:50 | <Lachy> | dammit. Somehow I've ended up with nearly 3000 unread messages across various HTMLWG and WHATWG mailing lists over the past 3 weeks :-( |
| 14:50 | <karlushi> | >These attributes are not intended for use by software that is independent of the site that uses the attributes. |
| 14:50 | <hsivonen> | karlushi: right. the wording sucks. |
| 14:50 | <Lachy> | and it's only going to get worse as I'm incredibly busy travelling over the next 3 weeks too |
| 14:50 | <karlushi> | > |
| 14:50 | <karlushi> | User agents must not derive any implementation behavior from these attributes or values. Specifications intended for user agents must not define these attributes to have any meaningful values. |
| 14:51 | <karlushi> | so I would say that Leif and Shelley are right given the current spec wording. |
| 14:51 | <Dashiva> | Eh? |
| 14:51 | <hsivonen> | karlushi: the spec bug is that it can be read to say that Dojo et al. shoudn't use the data-* attributes. But data-* attributes exist for Dojo et al. |
| 14:52 | <karlushi> | If the spec is misleading, it has to be changed. |
| 14:52 | <Dashiva> | If you use a script on your site, that script is not independent of the site |
| 14:52 | <hsivonen> | karlushi: indeed. |
| 14:52 | <karlushi> | Dashiva, understood that ;) |
| 14:53 | <karlushi> | The full thing is completely ambiguous IMHO. Specifically in the age of clown computing. s/clown/cloud/ |
| 15:03 | <zcorpan_> | Hixie: yeah, i realize explaining the escaping thing for authors will suck |
| 15:06 | <hsivonen> | is it too bad to make it non-conforming to have the string "<script" or the string "</script" in a script? |
| 15:09 | <zcorpan_> | there's no problem with <script> document.write('<script><\/script>') </script> |
| 15:09 | <hsivonen> | did Hixie add some red lines to Damowmow portal or is my browser broken? |
| 15:10 | <hsivonen> | zcorpan_: there is in older UAs |
| 15:10 | <zcorpan_> | i see red lines in opera |
| 15:10 | <zcorpan_> | hsivonen: oh? |
| 15:11 | <hsivonen> | oops. sorry |
| 15:11 | <hsivonen> | I misread what you wrote |
| 15:18 | <zcorpan_> | i guess with the current landscape of tags(), the bugs would kind of go away if we limit document.all to quirks mode and tags() to document.all |
| 15:21 | hsivonen | is not too fond of Microdata introducing a new void element now |
| 15:24 | <Lachy> | woah, I missed the addition of the itemref element till now. |
| 15:24 | Lachy | wonders what it's for |
| 15:26 | <Lachy> | woah. That seems to complicate microdata significantly |
| 15:34 | <TabAtkins> | annevk: Thanks for the response re XHR's responseXML. |
| 15:42 | <TabAtkins> | Lachy: Just do what I do and spend an hour every morning wading through stuff. |
| 15:43 | TabAtkins | isn't sure it's quite a good use of 1/24th of his time to be doing so. |
| 16:13 | <Koterpillar> | Does HTML5 support sound recording? |
| 16:13 | <annevk> | nope |
| 16:14 | <annevk> | well, in theory a browser could tie it into <input type=file> |
| 16:14 | <annevk> | but then you still wouldn't have support for streaming it to a server and all |
| 16:15 | <Koterpillar> | so it still doesn't match Flash... |
| 16:16 | <annevk> | still doesn't? |
| 16:19 | <Koterpillar> | never mind, saying own thoughts aloud. In fact, I've seen that audio recording capability in Flash, and hoped that there's something standard along that lines |
| 16:22 | <Lachy> | we might get such functionality one day, but there are so many issues that need to be solved before we get there, it's going to take a while |
| 16:22 | <annevk> | Koterpillar, we should have it, yes |
| 16:22 | <annevk> | Koterpillar, nobody has proposed an API so far |
| 16:43 | <mikekelly> | if I have a page /example.html and my server provides the html page and a custom header e.g. X-Something-Interest: foo - can javascript running off the page get at that custom header? |
| 16:44 | <mikekelly> | I don't know how to ask that question in a smart way sorry :( |
| 16:44 | <annevk> | yes, via XMLHttpRequest |
| 16:45 | <mikekelly> | in retrospect of the current page? |
| 16:45 | <mikekelly> | or you mean make a HEAD request back? |
| 16:45 | <annevk> | either way |
| 16:46 | <mikekelly> | so for every page request in a browser there's a default object containing the request used to get it? |
| 16:46 | <annevk> | no |
| 16:47 | <mikekelly> | but the response? |
| 16:47 | <annevk> | no |
| 16:48 | <mikekelly> | .. so I do have to make a HEAD request then..? |
| 16:48 | <annevk> | you could |
| 16:48 | <mikekelly> | how come the browser discards the request/response data for a page and doesn't make it available to js? |
| 16:49 | <annevk> | just not a feature that's been requested often prolly so it was not added |
| 16:50 | <annevk> | there was some discussion on this on the WHATWG list a while back |
| 16:50 | <mikekelly> | another thing - it's not possible to force a location change with custom headers is it? |
| 16:50 | <mikekelly> | all you can provide is a URI and that's it |
| 16:51 | <mikekelly> | custom request headers* |
| 16:51 | <annevk> | I don't understand |
| 16:51 | <mikekelly> | well say I make a XHR I can modify request headers |
| 16:51 | <mikekelly> | but if I want to point the browser to a new location I can only provide a URI |
| 16:52 | <annevk> | is this about conneg again? |
| 16:52 | <annevk> | please go away |
| 16:52 | <mikekelly> | I can't make the browser include custom headers for 'location changes' |
| 16:54 | <annevk> | TabAtkins, me too |
| 16:54 | <mikekelly> | annevk: this is me figuring out whether I can get around it or not |
| 16:54 | <annevk> | just a little afraid of legacy though I guess @namespace is not used much (if at all) and certainly not in weird ways like that |
| 16:55 | <mikekelly> | which is a good thing |
| 16:55 | <mikekelly> | or at least I'd assume it was a good thing given how shitty you are being about it :) |
| 16:55 | <TabAtkins> | annevk: Yeah, I've never seen it myself. Admittedly, my experience with other people's stylesheets is somewhat limited, but still. |
| 16:56 | <TabAtkins> | I suspect we can make the change now, before namespace cruft gets shoveled into HTML, and be fine. |
| 16:57 | <mikekelly> | I'm guessing the answer as to whether I can give the request when a browser's location changes custom headers is no |
| 16:58 | <annevk> | mikekelly, there's no such thing; the reason I'm shitty is that you wasted more than a few hours of everyone's time by repeatedly bringing it up here and have been ignoring what people are saying to you |
| 16:58 | <mikekelly> | or.. I understand what they are saying.. I'm giving them responses they don't understand.. and the opposite is happening |
| 16:58 | <mikekelly> | that's always possible |
| 17:00 | <annevk> | mu |
| 17:00 | <mikekelly> | I'm forced to repeat myself because I keep being provided with retorts that I've already addressed.. |
| 17:01 | <mikekelly> | my last email on the list to Thomas Broyer was yet more repitition of stuff I'd already said |
| 17:01 | <mikekelly> | and he'd plainly ignored |
| 17:02 | <mikekelly> | I may be abrasive but your attitude stinks just as bad |
| 17:03 | <mikekelly> | at least I'm upfront about it. |
| 18:16 | <zcorpan_> | Philip`: any progress? |
| 18:22 | <annevk42> | hsivonen, did you end up filing a bug on the <base> thing? |
| 18:22 | <annevk42> | hsivonen, it does not seem to be in /issues/ so if there ever was an email about it that has not made an impact |
| 18:25 | <annevk42> | I don't see a bug |
| 18:25 | annevk42 | files one |
| 18:27 | <annevk42> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=7971 |
| 18:35 | <zcorpan_> | annevk: s/kronor/krona/ |
| 18:57 | <annevk42> | isn't kronor plural? |
| 18:57 | <daedb> | kronor = plural |
| 18:58 | <Philip`> | zcorpan_: No, because I forgot about it |
| 18:59 | <annevk42> | ah, but I wrote "every kronor"; fail |
| 20:00 | <Rodi01> | is the dd and dt required inside a figure tag? |
| 20:05 | <annevk42> | Rodi01, dt is optional |
| 20:05 | <annevk42> | Did anyone review http://dev.w3.org/2006/webapi/DataCache/ ? |
| 20:09 | <Rodi01> | I'm using a figure tag to render an avatar inside a comment list. Inside the figure tag, there is an image wrapped by an anchor. Should I wrap everything inside a dd? or use an aside instead? |
| 20:14 | <annevk42> | you don't really need either I think |
| 20:55 | TabAtkins | really doesn't understand how the CSS inline model works. |
| 20:58 | <karlushi> | http://geocitieslycostripod.com/ I think it is the best html5 Web site I have seen for a long time. |
| 21:01 | <Philip`> | That site is clearly unrealistic - it doesn't even have an "under construction" sign |
| 21:02 | <annevk42> | ah, from http://piecesofrakesh.blogspot.com/ |
| 21:02 | annevk42 | thought he recognized that name |
| 21:02 | <TabAtkins> | karlushi: Awesome. ^_^ Unfortunately he's using <legend> in his <figure>. |
| 21:03 | <jgraham> | Philip`: http://geocitieslycostripod.com/under_construction.html |
| 21:04 | <annevk42> | TabAtkins, inline box model is hard |
| 21:04 | <annevk42> | let me rephrase that, CSS is hard |
| 21:04 | <TabAtkins> | Nah, CSS is easy. Inline is hard. |
| 21:04 | <Philip`> | jgraham: Oh, I didn't see that |
| 21:04 | TabAtkins | just wants these stupid inline-blocks to align their tops, rather than their bottoms. WHY IS THIS IMPOSSIBLE?!? |
| 21:05 | <karlushi> | TabAtkins, gravity! |
| 21:05 | <Hixie> | holy crap, go art! a chair of a w3c wg actually maintaining the community! impressive |
| 21:06 | <TabAtkins> | karlushi: We should be able to defy gravity by now. I demand it! |
| 21:07 | <Philip`> | On the subject of data-* name clashes: because the scripts using the attributes are not independent of the site, the site author can just edit the scripts to use non-conflicting names |
| 21:07 | <TabAtkins> | Philip`: True, but that may be non-trivial. |
| 21:07 | <Philip`> | which is also a way of clarifying what it means for a script to be independent of the site (or whatever the terminology was) |
| 21:07 | <TabAtkins> | Philip`: In any case, there's a very simple way to go about things that should make that almost never necessary. |
| 21:08 | <TabAtkins> | Hixie: Details? |
| 21:08 | <Hixie> | see member-webapps |
| 21:08 | <jgraham> | Yeah. I'm reasonably sure that most use cases don't need the bondance and discipline implied by URIs everywhere |
| 21:09 | <TabAtkins> | Wow, that's an interesting substitution. |
| 21:09 | <jgraham> | er... |
| 21:09 | <jgraham> | bondage |
| 21:12 | <Philip`> | URIs make it hard for less-than-perfect developers to accidentally create ambiguous names that will hurt them in the future |
| 21:14 | <annevk42> | so use a URL in a class name |
| 21:14 | <annevk42> | you could even make it so that the data- attributes are scoped by that in your design so there's even less chance of clashes |
| 21:14 | <jgraham> | Philip`: Sounds like the sort of argument that java is great because it has all sorts of mechanisms to prevent less than perfect developers shooting themselves in the foot |
| 21:15 | <jgraham> | (at the expense of great pain to everyone who has to use it) |
| 21:15 | <Dashiva> | And everyone seems to ignore the fact that since you don't have to resolve namespace URIs, anyone can use any URI even if they don't own the domain |
| 21:16 | <jgraham> | Dashiva: And also the fact that htp URIs are temporarly non-unique so you either have to use tag uris or uuids or something |
| 21:44 | annevk42 | approved some more blog comments and deleted some Ajaxian-copiers that seemed like spam and marked others that were clearly spam as spam |
| 21:44 | <annevk42> | most were made on http://blog.whatwg.org/spelling-html5 |
| 21:45 | <annevk42> | oh, I also fixed a link in one of markp's older posts on contentEditable since I got a bunch of errors logged on my server |
| 21:45 | <annevk42> | top 5 error this month according to the silly statistics software that came with DreamHost |
| 21:55 | annevk42 | changes topic to 'WHATWG: http://www.whatwg.org/ -- logs: http://krijnhoetmer.nl/irc-logs/ -- stats: http://gavinsharp.com/irc/whatwg.html -- Please leave your sense of logic at the door, thanks!' |
| 22:03 | <jgraham> | gah, Hixie is disconnected |
| 22:03 | <annevk42> | anarchy |
| 22:03 | <annevk42> | oh wait, that's nothing new :p |
| 22:04 | <jgraham> | Anarchy in the WHATWG. Sounds like a fun punk album |
| 22:05 | annevk42 | feels another boring last week post coming up |
| 22:07 | <jgraham> | Is there another kind? |
| 22:07 | <annevk42> | not lately anyway |
| 22:09 | <jgraham> | Well if he/she/it wants something else to write about, the end of neutral milk hotel's "On Avery Island" is rather nice |
| 22:11 | <jgraham> | Seem like a more interesting thing to tell the world than what goes on in here anyway |
| 22:13 | <annevk42> | shouldn't we pretend that what goes on here is the most awesome? seems to be better fit the image the narrator tries to convey |
| 22:14 | <jgraham> | Dunno if I can pull off pretense at that level |
| 22:15 | <TabAtkins> | We also probably shouldn't cater to the delusions of the mentally ill. |
| 22:34 | <TabAtkins> | sicking: Microformats are not a specification of metadata vocabularies to be implemented by RDFa/Microdata/etc. They're a competing format based mostly on use of @class. |
| 22:35 | <sicking> | TabAtkins: depends on your definition of microformats i guess |
| 22:35 | <TabAtkins> | I'm going with the microformats.org. ^_^ |
| 22:36 | <TabAtkins> | The only difference is that microformats have to carry around their parsing rules as well, while RDFa and Microdata have general parsing rules that are used by all vocabs. |
| 22:36 | <annevk42> | microformats.org is the namespace used in some of the splitted out Microdata vocabularies though |
| 22:36 | <annevk42> | so maybe you wanna read up on that |
| 22:36 | <sicking> | TabAtkins: from front page "microformats are a set of simple, open data formats" |
| 22:37 | <Philip`> | Maybe microdata should be marketed as JSONa (since it's basically embedding trees of strings into HTML attributes) |
| 22:37 | <TabAtkins> | Hmm, ok, point to you guys. The Microformats community *does* define a number of vocabularies. Microformats is *also* a competing syntax for embedding said vocabularies. |
| 22:38 | <sicking> | i'd say HTML4 class+rel+"a few more" are a competing syntax |
| 22:38 | <TabAtkins> | sicking: And that combination is precisely what is used by Microformats. It's not an arbitrary collection of stuff that people just happen to sometimes use to encode microformat vocabs. |
| 22:42 | <TabAtkins> | Is there any way to copy text out of a putty shell without doing the stupid "Copy All to Clipboard" option? |
| 22:42 | <TabAtkins> | I keep accidentally Ctrl-C-ing and aborting whatever program I'm running. >_< |
| 22:42 | <Philip`> | TabAtkins: I thought it copied automatically when you select text |
| 22:43 | <hober> | TabAtkins: microformats use the existing HTML extension mechanisms; microdata is an addition to HTML's extension mechanisms. When microdata ships, microformats can naturally use microdata's attributes as they currently use class |
| 22:43 | <hober> | which is to say, microdata & microformats are on different layers. |
| 22:43 | <TabAtkins> | hober: Okay, makes sense. That would probably be a very nice net win, actually, so that microformats don't have to define their own parsing rules. |
| 22:44 | <TabAtkins> | Philip`: ... ... ... WHY DID I NEVER KNOW THIS? |
| 22:44 | <Philip`> | TabAtkins: Because you never tried to select some text and then right-click to open a context menu to copy it |
| 22:45 | <TabAtkins> | No. No I did not. |
| 22:45 | <Philip`> | which is how I discovered the copy feature and also the paste feature |
| 22:45 | <TabAtkins> | I know about pasting, somehow. I just can't believe I never knew about copying. Wow. |