/irc-logs / w3c / #html-wg / 2008-12-15 / end

Options:

# Session Start: Mon Dec 15 00:00:00 2008
# Session Ident: #html-wg
# [00:07] * Quits: Sander (svl@86.87.68.167) (Quit: And back he spurred like a madman, shrieking a curse to the sky.)
# [00:13] * Quits: heycam (cam@130.194.72.84) (Quit: bye)
# [01:15] * Quits: adele (adele@67.188.228.184) (Client exited)
# [01:25] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: sex break)
# [01:35] * Joins: heycam (cam@130.194.72.84)
# [01:48] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [01:57] * Quits: karl (karlcow@69.70.209.57) (Quit: This computer has gone to sleep)
# [01:58] <MikeSmith> Hixie: valid e-mail address is currently defined as a set of comma-separated tokens
# [01:59] <MikeSmith> I'm wondering if it could be space-separated instead
# [02:01] <MikeSmith> since each token is a "dot-atom "@" dot-atom" e-mail address, and dot-atom can't contain any whitespace
# [02:09] * Joins: karl (karlcow@70.81.84.24)
# [02:15] * Quits: karl (karlcow@70.81.84.24) (Quit: This computer has gone to sleep)
# [02:18] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: sex break)
# [02:23] * Joins: karl (karlcow@70.81.84.24)
# [03:06] * Joins: marcos_ (marcos@87.196.232.155)
# [03:08] * Quits: marcos (marcos@87.196.201.71) (Ping timeout)
# [03:24] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [03:28] <MikeSmith> Hixie: given that "makeapp" is the only value allowed so far for bb@type, would it make sense for the missing value default state for bb@type to just be the "make application" state?
# [04:18] <Hixie> MikeSmith: spaces are allowed in the full syntax for an e-mail address, so that would be incompatible
# [04:19] <Hixie> MikeSmith: it would make sense today, but it wouldn't make sense 20 years from now
# [04:57] * Quits: karl (karlcow@70.81.84.24) (Quit: This computer has gone to sleep)
# [05:03] <MikeSmith> Hixie: I see
# [05:13] * Joins: karl (karlcow@69.70.209.57)
# [05:23] * Quits: Lachy (Lachlan@85.196.122.246) (Connection reset by peer)
# [05:23] * Joins: Lachy (Lachlan@85.196.122.246)
# [06:04] * Joins: shepazu (schepers@128.30.52.30)
# [06:50] * Quits: heycam (cam@130.194.72.84) (Quit: bye)
# [06:57] <MikeSmith> Hixie: about my earlier question on bb, I see after reading more that intent of <bb> without a type value specified is to disable the button
# [06:59] * Joins: anne (annevk@77.163.243.203)
# [07:02] * Parts: anne (annevk@77.163.243.203)
# [07:18] * Quits: gavin_ (gavin@99.226.207.11) (Ping timeout)
# [07:24] * Joins: gavin_ (gavin@99.226.207.11)
# [07:30] * Joins: heycam (cam@203.217.82.242)
# [07:55] * Quits: Lachy (Lachlan@85.196.122.246) (Quit: This computer has gone to sleep)
# [08:55] * Joins: aaronlev (chatzilla@92.227.71.237)
# [09:36] * Quits: gavin_ (gavin@99.226.207.11) (Ping timeout)
# [09:41] * Joins: gavin_ (gavin@99.226.207.11)
# [09:49] * Joins: tlr (tlr@128.30.52.30)
# [09:57] <MikeSmith> http://shortyawards.com/category/douche
# [09:57] <pimpbot> Title: The Shorty Awards - douche leaderboard (at shortyawards.com)
# [09:57] <MikeSmith> Guy Kawasaki ruling the boards
# [09:59] <MikeSmith> hmm, I had actually meant to post that in a different channel. oh well
# [10:00] * Joins: Lachy (Lachlan@213.236.208.22)
# [10:51] * Joins: ROBOd (robod@89.122.216.38)
# [11:06] <MikeSmith> hsivonen: it appears that "autosubmit" is no longer valid on <menu>
# [11:06] <MikeSmith> http://lists.whatwg.org/pipermail/commit-watchers-whatwg.org/2008/001196.html
# [11:06] <pimpbot> Title: [html5] r2019 - [] (0) Remove autosubmit='' from the element. (at lists.whatwg.org)
# [11:06] <MikeSmith> r2019
# [11:49] * Joins: Sander (svl@86.87.68.167)
# [12:18] * Joins: maddiin (mc@87.185.218.121)
# [12:19] * Quits: Julian (chatzilla@217.91.35.233) (Ping timeout)
# [12:21] * Joins: Julian (chatzilla@217.91.35.233)
# [12:49] * Joins: myakura (myakura@122.17.190.200)
# [13:04] * Joins: jgraham_ (james@213.115.244.54)
# [13:04] * Quits: jgraham_ (james@213.115.244.54) (Quit: I'll hit the bottom and escape)
# [13:35] <karl> MikeSmith: thanks very much for everything you did in the WG. Really really valuable.
# [13:35] <karl> And welcome to Sam Ruby
# [13:37] <MikeSmith> karl: I didn't throw any shoes at George Bush on behalf of the group, or call him a filthy dog to his face.
# [13:37] <MikeSmith> I wish I had been able to do that at least.
# [13:37] <MikeSmith> but who knows, I may still even have a chance for that
# [13:38] <jgraham> interesting
# [13:38] <MikeSmith> would that Muntazer al-Zaidi had had something bigger to throw -- like, say, a refrigerator
# [13:39] <karl> :))) there might be more opportunities for shoes throwing. You should train first on less high-profile peopel
# [13:39] * jgraham decides to practice ducking
# [13:41] * karl wonders what an IRC shoes would look like
# [13:45] * MikeSmith packs up his pony for the ride home to Tokyo
# [13:45] <MikeSmith> bbiab
# [13:46] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: sex break)
# [13:51] <Philip> karl: If you spend a lot of time training by throwing shoes at people, I think it's unlikely that you'll ever be let close enough to George Bush
# [13:53] <Philip> I suppose that's similar to the curse of being an evil mastermind - you can't practice by making a very small prototype doomsday device that just wipes out Greenland or somewhere nobody cares about, because the rest of the world will still notice and will stop you, so you have to start with the giant full-scale untested one and it'll never work properly
# [13:53] <Lachy> is there a reason for the sudden change of WG chair?
# [13:54] <Philip> I thought it was suggested when MikeSmith took over the role that it was probably temporary
# [13:54] <Philip> so change is not unexpected
# [13:55] <Lachy> well, Mike was doing such a good job and prior to now, there was no warning about it
# [13:56] <Lachy> I just hope Sam does the job well
# [13:57] <Philip> When Mike took the job, I don't think anyone even mentioned it until after he'd been there a week or so, whereas now we're being told three weeks in advance
# [13:58] <jgraham> Lachy: Well we already know that DanC had his hours cut and was looking for another job
# [14:00] <jgraham> I guess Sam already has quite strong ideas about how the spec should progress so it might be interesting at least
# [14:00] <Lachy> indeed. That's my concern about appointing him
# [14:03] <Philip> I assume your concern is that he has quite strong ideas *and* they are not the same as your ideas?
# [14:04] <Philip> (i.e. you wouldn't mind someone being appointed if they strongly agreed with you?)
# [14:06] <Lachy> Philip, I would prefer someone who is capable of remaining objective, especially on contraversial issues
# [14:06] <Lachy> and it wouldn't hurt at all if they happened to agree with me :-)
# [14:07] * Joins: MichaelC (Michael@128.30.52.30)
# [14:07] <karl> Lachy: the role of the chair is well defined.
# [14:07] <karl> and when you take over this kind of role, you specifically accept the rules which are going with it.
# [14:07] <Lachy> where are these rules listed?
# [14:07] <karl> translation: A participant and a person fulfilling a precise role do not have the same behaviour.
# [14:08] <karl> Same thing, I have my personal ideas about html, but put that on rest when I was an employee of w3c, because it was not my role to voice them, but to try to help the group
# [14:09] <Philip> Lachy: By 'objective', do you mean it should be someone who looks at all the evidence and makes a rational decision, or do you mean someone who remains uninvolved and leaves the technical issues to other WG members?
# [14:10] <jgraham> karl: So what effect will Sam being chair have on his ability to implement the ideas in http://intertwingly.net/blog/2008/11/20/Half-Full
# [14:10] <pimpbot> Title: Sam Ruby: Half Full (at intertwingly.net)
# [14:10] <jgraham> (he specifically states there that the chair should implement those ideas)
# [14:10] <karl> Lachy: you will find information in this document http://www.w3.org/Guide [Member Only] and I have no access but you do.
# [14:11] <karl> jgraham: we will see :) I'm no Sam Ruby :)
# [14:11] <jgraham> (FWIW I do not belive that Sam being appointed chair is a bad thing)
# [14:11] <jgraham> karl: In that case I don't understand your earlier statement
# [14:12] <Lachy> Philip, both
# [14:12] <karl> jgraham: to not mix up ideas of someone and the rules driving a role, these two are two different contexts.
# [14:13] <karl> :) anyway, I have to move my butt from point A to point B. ;) bbl
# [14:13] * Quits: karl (karlcow@69.70.209.57) (Quit: This computer has gone to sleep)
# [14:22] * Joins: aroben (aroben@69.137.139.251)
# [14:28] * Joins: rubys (rubys@75.182.87.110)
# [14:30] * Joins: karl (karlcow@70.81.84.168)
# [14:32] * Quits: gavin_ (gavin@99.226.207.11) (Ping timeout)
# [14:38] * Joins: gavin_ (gavin@99.226.207.11)
# [14:50] * jgraham likes how MikeSmith linked to rubys's wikipedia page which was recently edited extenstively by one Sideshowbarker ;)
# [14:50] * rubys smiles
# [14:51] <Philip> And now that page links back to MikeSmith's email which links to it :-)
# [14:54] <karl> meta
# [14:54] * Philip wonders what the distribution of cycle lengths in the web looks like - there are presumably long cycles between pretty much everywhere, but maybe very few direct cycles of length 2?
# [14:54] <jgraham> rubys: Congratulations (or, I guess commiserations) on your new role
# [14:54] <rubys> my page on wikipedia links to my weblog which links to my wikipedia page
# [14:55] <rubys> and thanks... herding cats takes a lot of work.
# [14:59] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [15:05] <pimpbot> planet: Co-Chair HTML WG <http://intertwingly.net/blog/2008/12/15/Co-Chair-HTML-WG>
# [15:10] * Quits: Julian (chatzilla@217.91.35.233) (Connection reset by peer)
# [15:13] * karl is trying to imagine sam with a cat woman suit
# [15:14] * Philip is now trying hard to not imagine that :-p
# [15:15] * Joins: Julian (chatzilla@217.91.35.233)
# [15:16] * karl hehe
# [15:17] <Philip> (No offence intended, of course)
# [15:20] * rubys chuckles
# [15:22] <pimpbot> bugmail: "[Bug 6294] Date microsyntax defines incorrect range for month and day fields" ( message in thread) <http://lists.w3.org/Archives/Public/public-html-bugzilla/2008Dec/0021.html>
# [15:31] * Joins: anne (annevk@217.174.106.250)
# [15:48] * Quits: aroben (aroben@69.137.139.251) (Ping timeout)
# [15:50] * Joins: zcorpan (zcorpan@88.131.66.80)
# [15:52] * Joins: aroben (aroben@69.137.139.251)
# [15:55] * Joins: smedero (smedero@72.184.0.219)
# [16:00] * Parts: zcorpan (zcorpan@88.131.66.80)
# [16:06] * tlr is now known as tlr-jogging
# [16:26] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Ping timeout)
# [16:26] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [16:27] * Quits: myakura (myakura@122.17.190.200) (Quit: Leaving...)
# [16:36] * Quits: aroben (aroben@69.137.139.251) (Connection reset by peer)
# [16:38] * Joins: aroben (aroben@69.137.139.251)
# [16:46] * MikeSmith wonders where hsivonen might be this week
# [16:46] <gsnedders> ewww… reply from matt
# [16:46] <gsnedders> *oooo…
# [16:46] <gsnedders> That's the sound I want
# [16:47] <Lachy> LOL
# [16:47] <gsnedders> Yeah, I FAIL.
# [16:48] <MikeSmith> hsivonen: wanted to ask again about proposed schema changes around value attribute on input@type=email
# [16:48] <Lachy> gsnedders, I've only seen 2 mails in that thread on www-archive. Were the rest posted elsewhere?
# [16:48] <gsnedders> I wasn't originally public
# [16:48] <Lachy> oh. It makes it difficult to understand what the discussion is about
# [16:49] <gsnedders> the first two emails aren't public; it was sent to Automattic (and they only give a contact form, so there was little chance to make it public) and his reply are both private
# [16:53] <gsnedders> Lachy: You see the things in November archives as well?
# [16:55] <Lachy> gsnedders, possibly. I can't remember if I read them or not
# [16:59] * Quits: anne (annevk@217.174.106.250) (Ping timeout)
# [17:00] <MikeSmith> hsivonen: after thinking about input.email.elem.multiple, that seems less than ideal. seems like there's no other precedent in the current schema for having two different patterns/flavors of a particular sub-typed element that share the same value for their type attribute
# [17:02] <MikeSmith> would seem perhaps better to just keep input.email.elem only but reflect the @value states either just in the attribute model or in the datatype itself
# [17:03] <MikeSmith> e.g.,:
# [17:03] <MikeSmith> ...
# [17:03] <MikeSmith> & input.email.attrs.value.single?
# [17:03] <MikeSmith> oops
# [17:04] <MikeSmith> & (input.email.attrs.value.single|& input.email.attrs.value.multiple)?
# [17:06] <karl> Lachy: I don't know if you had the opportunity to see the (very short) draft I made to explain the way I would see those guidelines.
# [17:06] <Lachy> karl, link?
# [17:07] <Lachy> which guidelines are you referring to?
# [17:07] <karl> Coming, coming, coming, …
# [17:07] <karl> Web Developers Guidelines, the document you were working on
# [17:07] <karl> http://www.la-grange.net/2008/12/11/web-authoring-example
# [17:07] <pimpbot> Title: HTML element for Web Authoring Guidelines Proposal (at www.la-grange.net)
# [17:08] <MikeSmith> hsivonen: or define the form.data.email-list datatype as "either a single e-mail address or a comma-separated list of multiple e-mail addresses"
# [17:09] <Lachy> karl, that's reasonably good stuff.
# [17:10] <karl> it will reasonably need more edits, I bet :)
# [17:12] <Philip> Is it possible to write something so that inexperienced authors will be able to make an informed choice between text/html and application/xhtml+xml?
# [17:12] <Lachy> Philip, yes
# [17:12] * karl realized that syntaxes in application/xhtml+xml and text/html were too different. We will get hit on that
# [17:12] <karl> Philip: yes :)
# [17:12] <karl> Philip: it's why I put
# [17:12] <karl> >There is an important decision for you to make now. Will you serve your document as application/xhtml+xml or as text/html? @@link to explanation about xml and html on the server side and issues on client sides@@.
# [17:13] <karl> in the W3C/DanC culture of editing, "@@" means further edits
# [17:13] <Lachy> karl, I intend to cover that in more depth in the introduction that talks about the syntaxes and provides a gental introduction to writing (X)HTML
# [17:14] <Philip> It kind of reminds me of the Gentoo installation manual saying "now you need to install a cron daemon; you could use dcron, or fcron, or vixie-cron - the choice is yours!" and me thinking "aaargh, I don't understand the difference, please just tell me which one to use for now"
# [17:14] <karl> Philip: exactly
# [17:14] <Philip> (Oh, good, it looks like they've updated the manual to say "If you don't know what to choose, use vixie-cron.")
# [17:15] <karl> I wish people could just put the same markup and decide to serve it with the mime type of their choice
# [17:15] <karl> We could then just give only one syntax in examples
# [17:15] <Lachy> karl, I'm pretty sure I have some good stuff that explains that somewhere, either in an old blog entry or an old unpublished draft of mine. I'll see what I can find
# [17:15] <karl> and that would work for both
# [17:15] * Joins: anne (annevk@217.174.106.250)
# [17:16] <karl> Lachy: ah good !
# [17:16] <Philip> Seems it would be much simpler to just give the examples in text/html, and have a separate section or separate document explaining how to migrate to XHTML
# [17:16] <karl> Philip: or the opposite
# [17:16] <karl> ;)
# [17:16] <Lachy> I've been thinking about ways to present both syntaxes in a simple way. One possibility is to provide both HTML and XHTML for every example, but to present it in a way that shows one at a time and lets the user toggle between them
# [17:17] <karl> to give the examples just in xhtml and how to migrate to html
# [17:17] <Philip> When people are reading the document, I'd guess they'll just copy-and-paste whatever example looks like it's vaguely relevant, and they won't even notice which syntax it's for
# [17:17] <Philip> karl: I'd prefer the approach that is compatible with what 99.9% of authors already do ;-)
# [17:18] <karl> Philip: so it's what I was thinking when starting to write
# [17:18] <karl> who is the readership of this document
# [17:18] <karl> Let's take the 99.9% you are casting. Who are they? as in authors != Web documents
# [17:19] <karl> and if we try to go a bit further, will the readership be Web designers mostly
# [17:19] <karl> then the syntax they use is XHTML 1.0/CSS
# [17:19] <Lachy> karl, the readership includes authors who hand code their HTML from all skill levels ranging from beginner to advanced, and ecluding people who only use WYSIWYG editors
# [17:19] <karl> ooops s/CSS/and CSS 2/
# [17:19] <Lachy> *excluding
# [17:21] <karl> Lachy: agreed. Are you sure that they are many people who does hard core html. Most of the people I know in Web agencies and conferences do xhtml. I guess i have my own bias
# [17:21] <karl> the only people I know doing html… are the persons of the html wg in editing their specs or blogs :)
# [17:21] <Julian> I always edit in XML and then XSLTransform to either HTML or XHTML
# [17:21] <karl> s/they are/there are/ damn french
# [17:22] <karl> Julian: I always edit in xhtml 1.1 (these days) and leave it as is. I know. bye bye IE :)
# [17:22] <karl> ;) but I don't have to sell things
# [17:23] <Philip> karl: If people use an XHTML doctype, and sometimes use trailing slashes in elements, but serve as text/html and mix in un-closed <br> elements and write <a href="foo?a=b&c=d">, would you say they are writing XHTML or HTML?
# [17:23] <karl> XHTML
# [17:23] <Julian> Karl: when I have the choice, I use XHTML exclusively as well
# [17:23] <karl> clearly in their mind, they do xhtml.
# [17:23] <Lachy> karl, despite what people claim, most people do HTML and should not be attempting XHTML
# [17:24] <karl> The technical purity or exactitude here makes us discuss a lot… but it is not their concerns. They really do XHTML in their minds. and that's fine by me. They develop a business around it.
# [17:24] <Lachy> karl, my basic philosopy is that XHTML is not for beginners.
# [17:24] <Philip> karl: It seems clear that it's not clear in their mind, and they just mix and match whatever works and whatever they find in tutorials and examples :-)
# [17:25] <rubys> My answer to Philip's question is "HTML".
# [17:25] <Lachy> the difficulty I will have with this is trying to find the right balance between encouraging people to just use HTML for the majority of their needs, but without unfairly discriminating against those who have a legitimate need/use case for XHTML
# [17:25] <karl> Philip: This is a hardcore specialist opinion :) I know what you mean and understand it, but unfortunately it doesn't work in front of wide community
# [17:26] * Lachy agrees with rubys for once :-)
# [17:27] <karl> Lachy: "legitimate need/use case" FAIL ;) not yours to decide. Or if you start like this in the discusssion, you already alienate your opinions to a lot of resistance. Not a zen path.
# [17:27] <Lachy> karl, wtf?
# [17:27] <karl> :) was it a polite one?
# [17:28] <karl> Lachy: let's try with an image
# [17:28] <rubys> I think it might make sense to explore a different question instead. If you wget (for example) my weblog's front page and store it as a file on disk, is that file valid HTML5? Is that file valid XHTML5?
# [17:28] <Lachy> I didn't say it was up to me to decide which needs and use cases were legitimate and which weren't in all cases. But it's a fact that there are both legitimate and illigitimate needs and use cases for XHTML
# [17:29] <karl> you have a bakery, you sell bread, and to all your customers you say "this is the way you have to eat bread", in the end if the big crowd outside think they have their own right way of doing it, and if they hear all the time that you are right… they will not come to your shop.
# [17:29] <karl> rubys: already too geek
# [17:30] <karl> here we are discussing implementation levels, not users using a markup
# [17:30] <karl> s/we are/you are/
# [17:30] <rubys> Philip's question was one of markup. They key phrase that makes it HTML (as opposed to XHTML) in his question was the mime type.
# [17:31] <Lachy> karl, I don't understand the relevance of your analogy
# [17:31] <karl> agreed on implementers side
# [17:31] <karl> Lachy: no issue.
# [17:31] <karl> the mime type is something that most people doing markup don't care about
# [17:32] <rubys> luckily HTML5 accepts and properly parses most of the non-problematic XHTML idioms.
# [17:32] <karl> so for us, most people do "text/html", for them most people do "XHTML" (yes I chose keywords: mime-type/markup on purpose)
# [17:33] <jgraham> I don't see how users come into this untill you accidentially make them use an XML parser on your "XHTML" that is actually not XHTML
# [17:33] <karl> jgraham: they don't :)
# [17:34] <smedero> karl: I really respect your approach to building consensus ... we need it more often around here. I'm just not understanding why we would want to let authors build documents that they believe are XHTML but in practice are not.
# [17:34] <karl> rubys: agreed. though with the issue of validity that Web designers care about
# [17:34] <jgraham> karl: How does the bakery metaphor work then?
# [17:36] <Philip> jgraham: XHTML is the delicious smell of fresh croissants, and HTML is the mouldy bread crumbs on the floor
# [17:36] <karl> smedero: :) to come back to my initial comment, I just want the exact same syntax most of the time and have a valid document either served as text/html or application/xhtml+xml
# [17:36] <rubys> karl: I do that.
# [17:36] <karl> aaaaah Philip here you touch my poetic fiber.
# [17:36] <karl> rubys: what do you do with <meta charset="utf-8"/>
# [17:37] <rubys> that irritates me, I'd like to see that in XHTML5, but for now, I drop that.
# [17:37] <rubys> I have the charset in my Content-Type header.
# [17:37] <Lachy> karl, I don't have a problem with teaching people the techniques to do that. As long as they have a full understanding of the differences and similarities between the 2 syntaxes so that they can manage it properly
# [17:37] <karl> ding
# [17:37] <karl> rubys: "that irritates me" bingo!
# [17:38] <rubys> fair enough, but what irritates me today is *much* *much* smaller than a year ago.
# [17:38] * tlr-jogging is now known as tlr
# [17:38] <Philip> Lachy: Despite the opportunity cost of teaching them the difference between HTML and XHTML, taking time and effort that is not spent on you teaching / them learning something else?
# [17:38] <karl> Lachy: explaining and giving documents to describe the differences. I fully and violently agree with you.
# [17:38] <jgraham> (if we want bakeries, I think we are supplying the flour, authors are making the bread and users are the customers. XHTML is like a special type of flour that is supposed to make super-nice bread but in practice almost no-one can tell the difference. It is also supposed to be used with special yeast but if you happen to mix some ordinary flour in with the special yeast the customers get food poisoning so in practice people just use the ordinary yeast and do
# [17:39] <Philip> karl: Is it not acceptable to simply specify the charset in the HTTP Content-Type, and not inside the document itself?
# [17:39] <smedero> jgraham: wow. i've got coffee everywhere....
# [17:39] <karl> rubys: me too. I like the fact that you can use xhtml syntax of empty elements :) for example
# [17:39] <Philip> jgraham: I think you ran out of line
# [17:39] * smedero goes to get a new t-shirt
# [17:39] <rubys> In fact, IIRC, meta charset is the only remaining issue that affected my pages.
# [17:39] <jgraham> Philip: where?
# [17:39] <Philip> jgraham: "yeast and do"
# [17:39] <Lachy> it's useful to specify the encoding within the document because the HTTP Content-Type header doesn't stay with the document when moved to a different context
# [17:39] <karl> Philip: it is acceptable to specify the charset in the HTTP Content-Type, but it is not "simply" ;)
# [17:40] <jgraham> ordinary yeast and don't get the good
# [17:40] <jgraham> taste)
# [17:40] * rubys believes that meta-charset that differs from the explicit or implied charset that an XML parser would find when parsing XHTML should be considered a validation error.
# [17:40] * karl is giving soap powder for his shirt.
# [17:42] <Philip> Lachy: The problem with meta charset is that it does stay with the document when moved to a different context, even when that's a character context and not a byte context and therefore charset doesn't make sense any more
# [17:42] <karl> Lachy: yes usually recommended. And also for internationalization reasons, when using the document outside of the Web.
# [17:42] <rubys> Philip: that's a problem?
# [17:43] <rubys> I would say that the fact that the content-type does NOT stay with the document is a problem.
# [17:43] <rubys> http://www.jroller.com/robertburrelldonkin/entry/ruby_s_postulate
# [17:43] <pimpbot> Title: Weirdest Undreamt Use Case (at www.jroller.com)
# [17:45] <Philip> rubys: It's a bit of a problem - e.g. if I use html5lib to parse an arbitrary HTML5 document and then serialise it to a UTF-8 file stream, the meta charset will be wrong
# [17:45] <Philip> and there isn't a trivial way to ensure the correct meta charset is in the correct location when serialising
# [17:46] <karl> http://www.w3.org/International/tutorials/tutorial-char-enc/#Slide0260
# [17:46] <karl> First point: Always declare the encoding of your documents
# [17:46] <pimpbot> Title: W3C I18N Tutorial: Character sets & encodings in XHTML, HTML and CSS (at www.w3.org)
# [17:46] <Philip> (It might be less of a problem than the alternatives, but it's still a problem)
# [17:46] <Lachy> Philip, the meta charset is only a problem when transcoding occurs in the serialisation without updating the meta charset to match. But that doesn't happen frequently enough in practice to be a problem
# [17:46] <rubys> Philip: that's a valid issue: but it also applies to the html5 serialization, right?
# [17:46] <karl> http://dev.opera.com/articles/view/mama-document-encodings/#how
# [17:46] <pimpbot> Title: MAMA: Document Encodings - Opera Developer Community (at dev.opera.com)
# [17:47] <Philip> (<?xml encoding="..."?> seems much more sensible than meta charset, since a serialiser can trivially make sure it serialises an encoding value that is consistent with the encoding it's actually using)
# [17:47] <Julian> Lachy: it happens all the time with XML, so how is HTML5 different?
# [17:47] * Quits: smedero (smedero@72.184.0.219) (Quit: smedero)
# [17:47] <rubys> I agree with Julian's observation. I see it all the time with RSS.
# [17:47] <karl> :)
# [17:48] <Lachy> Julian, explain?
# [17:48] <Julian> I've answered questions about it hundrerds of times in XML forums...
# [17:48] <Julian> If the language has an element specifying the charset, but the serializers are not aware of it, you end up with wrond declarations soon.
# [17:49] <Lachy> rubys, <meta charset> isn't included within RSS, so that's not a problem
# [17:49] <Philip> rubys: It seems to apply equally to all HTML serialisations
# [17:49] <karl> Julian, a link to a summary information somewhere?
# [17:49] <karl> that would be cool
# [17:49] <Julian> That's why the XHML output mode of XSLT 1.0 by default adds it's own meta tag, and that is good.
# [17:50] * karl thinks that it sounds like a requirement for serializers class of products
# [17:50] * Julian agrees with Karl
# [17:50] <rubys> Lachy: http://www.intertwingly.net/slides/2005/xmlconf/
# [17:50] <pimpbot> Title: "Just" use XML (at www.intertwingly.net)
# [17:50] <Lachy> Julian, if the XHTML output mode of XSLT is outputting a meta charset declaration, then that's a bug because meta charsets are meaningless is XHTML
# [17:51] <Julian> The question is: we already have one way for XML (XML decl), and one for HTML, so do we need a third one serializers will have to learn=
# [17:51] <Julian> ?
# [17:51] <Julian> Sorry, I meant "HTML".
# [17:51] <Lachy> ok
# [17:51] <Lachy> Julian, why would we need a 3rd?
# [17:52] <Philip> <meta charset=...> vs <meta http-equiv=content-type content=text/html;charset=...> ?
# [17:52] <Julian> I think we don't.
# [17:52] <Philip> (Is that the "third one"?)
# [17:52] <Julian> I thought "meta charset=" is new?
# [17:53] <rubys> enough people botch http-equiv (dropping quotes) that looking for a chaset attribute on the meta tag is worthwhile.
# [17:53] <rubys> where worthwhile means "actives better results"
# [17:54] <Philip> Julian: It's new to HTML, though it's always been supported by browsers and used occasionally by pages
# [17:55] <karl> Julian: meta charset is new in the spec, but http-equiv seems to be complicated enough that it is not very well used by people hand typing it. On the other side, <meta charset=""/> will require to update a *lot* of authoring tools :/ which is also not perfect.
# [17:55] <rubys> I translate Philip's statement this way: it is new to being documented, but it always has been a part of HTML.
# [17:56] <karl> I think all the authoring tools I have used on the mac always produced a meta http-equiv, so making it invalid seems also counter effective.
# [17:56] <Julian> ok
# [17:56] <Philip> (When I say "used occasionally", I mean it's used on about 0.13% of pages from dmoz.org)
# [17:57] <Julian> is it being made invalid?
# [17:58] <Philip> http://www.whatwg.org/specs/web-apps/current-work/multipage/semantics.html#attr-meta-http-equiv-content-type
# [17:58] <pimpbot> Title: 4 The elements of HTML HTML 5 (at www.whatwg.org)
# [17:58] <Philip> It's valid now
# [17:58] <Philip> (where "It" is <meta http-equiv=content-type content=text/html;charset=...>)
# [17:59] <karl> henri has not updated its enginethen
# [17:59] <karl> just tested it
# [17:59] <karl> <!DOCTYPE html>
# [17:59] <karl> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
# [17:59] <karl> <head>
# [17:59] <karl> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
# [17:59] <karl> <title>untitled</title>
# [17:59] <karl>
# [17:59] <karl> </head>
# [17:59] <karl> <body>
# [17:59] <karl> </body>
# [17:59] <karl> </html>
# [17:59] <karl> and this is invalid
# [17:59] <karl> Line 5, Column 68: Bad value Content-Type for attribute http-equiv on XHTML element meta..
# [17:59] <karl> …Content-Type" content="text/html; charset=utf-8"/>
# [18:00] <rubys> I believe it is treated as not valid in XHTML
# [18:00] <rubys> for your test, what was the content-type?
# [18:00] * karl is checking something
# [18:00] <karl> rubys: direct input
# [18:00] <karl> ah no my bad
# [18:00] <karl> working
# [18:01] <karl> when you leave the namespace in direct input, the validator switch to xhtml
# [18:01] <karl> another syntax difference
# [18:01] <rubys> that switch is unfortunate. the namespace is valid in html5, and will be common among the blue beenie types.
# [18:02] * Quits: Lachy (Lachlan@213.236.208.22) (Quit: This computer has gone to sleep)
# [18:02] <karl> rubys: yes
# [18:03] <Philip> You can explicitly set whether to validate it as HTML5 or as XHTML5
# [18:03] <Philip> If using direct input, there's no content-type, so what would you suggest it should do by default?
# [18:03] * Quits: aaronlev (chatzilla@92.227.71.237) (Ping timeout)
# [18:04] <rubys> html5
# [18:04] <rubys> also, a checkbox wouldn't hurt (default being html5)
# [18:04] <Philip> Oh, validator.nu seems to default to HTML5 anyway
# [18:04] * Joins: smedero (smedero@192.223.6.251)
# [18:05] <Philip> ...and it can't do anything else because it's streaming and starts parsing the document before it's even got to the xmlns
# [18:05] <Philip> karl: Are you using validator.w3.org instead?
# [18:05] <karl> Philip: yes using validator.w3.org
# [18:05] <Philip> karl: Blame the W3C people, then ;-)
# [18:07] <karl> I can't do that.
# [18:07] <karl> ;)
# [18:07] <karl> You never say bad things about your (best) friends! in this case olivier.
# [18:08] <Dashiva> "more olivy"
# [18:08] <karl> Olivier and henri worked together to make it easier to integrate html 5 validation to the validator, still experimental :) but there will be room for improvements.
# [18:10] <Philip> There are already some bug reports like http://www.w3.org/Bugs/Public/show_bug.cgi?id=6296 about how it chooses which language to validate against
# [18:10] <pimpbot> 6296: simonp@opera.com, P2, NEW, text/html with XHTML 1.0 doctype and HTML5 override validates as XHTML5
# [18:11] <karl> W3C is trying to set up an environment for the validators to be more independent financially and have more resources for developing them. see http://www.w3.org/QA/2008/12/validator_donation_program
# [18:11] <pimpbot> Title: Validator Donation Program: day 2 - W3C Q Weblog (at www.w3.org)
# [18:14] <karl> Philip: seen the bug report from simon, though not sure I agree with it. thin line as always between user experience, contexts and technical purity
# [18:14] <Philip> Put AdSense on the validator, and all the problems will be solved
# [18:14] * Parts: rubys (rubys@75.182.87.110)
# [18:16] <karl> Philip: it will not work. The validator has 1) no content on the home page. 2) The MIT policy forbids to have commercial activities with Web sites hosted on their infrastructure.
# [18:18] <Philip> Host it in The Cloud, and all the problems will be solved
# [18:18] * Philip doesn't have any particularly constructive suggestions :-(
# [18:20] <jgraham> karl: How is sponsorship not a commercial activity? Presumably the sponsor gets a kickback in the form of advertising?
# [18:20] <karl> jgraham: they don't :)
# [18:21] <jgraham> karl: So the names of the sponsors will not be disclosed?
# [18:21] <karl> yep
# [18:22] * Quits: MichaelC (Michael@128.30.52.30) (Ping timeout)
# [18:22] <Philip> How is asking people to give you money not a commercial activity?
# [18:22] <Philip> It seems basically like selling stuff, only without the stuff
# [18:23] <karl> huh? that's call a donation… not selling.
# [18:23] <jgraham> http://www.w3.org/QA/Tools/Donate#donate_sponsors suggests otherwise
# [18:23] <pimpbot> Title: W3C Validator Donation Program (at www.w3.org)
# [18:23] <karl> Philip: I guess in the american system of law there is a clear distinction.
# [18:24] <karl> Wikipedia helps us :)
# [18:24] <karl> http://en.wikipedia.org/wiki/Donation
# [18:24] <pimpbot> Title: Donation - Wikipedia, the free encyclopedia (at en.wikipedia.org)
# [18:25] * Quits: maddiin (mc@87.185.218.121) (Quit: maddiin)
# [18:27] * Joins: rubys (rubys@75.182.87.110)
# [18:27] * Joins: MichaelC (Michael@128.30.52.30)
# [18:32] <karl> oh interesting http://versionfive.jp/sample/constructing_webpage_with_html5/ That's better publicity for html 5 than any spec
# [18:32] <pimpbot> Title: 老舗茶屋富士の園 with 福禄屋 (at versionfive.jp)
# [18:33] <karl> hmm
# [18:33] * karl wonders if html 5 needs its own "html5 zen garden" :) to show beautifully html5/css designed Web sites
# [18:34] <rubys> the empty set? :-)
# [18:34] <karl> hehe
# [18:35] * Quits: marcos_ (marcos@87.196.232.155) (Ping timeout)
# [18:36] <gsnedders> karl: What is that? I fail to understand :)
# [18:38] <rubys> http://www.csszengarden.com/
# [18:38] <pimpbot> Title: css Zen Garden: The Beauty in CSS Design (at www.csszengarden.com)
# [18:39] <rubys> excellent advocacy site. Primarily for CSS, but it had the side effect of convincing a lot of people that serving XHTML1 as text/html was a good idea.
# [18:42] * tlr is now known as tlr-bbl
# [18:43] <jgraham> I guess gsnedders didn't understand the japanese site
# [18:44] <rubys> google translate can help, but the fact that it is (a) pretty, and (b) html5 are likely the points that Karl was making.
# [18:44] <rubys> http://translate.google.com/translate?u=http://versionfive.jp/sample/constructing_webpage_with_html5/
# [18:44] <pimpbot> Title: Translated version of http://versionfive.jp/sample/constructing_webpage_with_html5/ (at translate.google.com)
# [18:45] <karl> rubys: indeed
# [18:47] <rubys> http://translate.google.com/translate?u=http://www.la-grange.net/2008/12/14/longue-marche
# [18:47] <pimpbot> Title: Translated version of http://www.la-grange.net/2008/12/14/longue-marche (at translate.google.com)
# [18:49] * Joins: hober (ted@206.212.254.2)
# [18:50] <gsnedders> yeah, jgraham is right
# [18:52] * karl is in lunch time mode… おなか　すいた
# [18:53] * Joins: tH (Rob@87.102.93.106)
# [18:55] * Quits: gavin_ (gavin@99.226.207.11) (Ping timeout)
# [19:00] * Joins: gavin_ (gavin@99.226.207.11)
# [19:10] * Quits: timelyx (timeless@88.115.12.126) (Ping timeout)
# [19:10] * Quits: Hixie (ianh@129.241.93.37) (Ping timeout)
# [19:10] * Joins: Hixie (ianh@129.241.93.37)
# [19:15] * Joins: Lachy (Lachlan@85.196.122.246)
# [19:20] * Joins: maddiin (mc@87.185.237.210)
# [19:30] * Joins: marcos (marcos@87.196.232.155)
# [19:34] * Joins: adele (adele@17.203.15.210)
# [19:39] * marcos wonders if anyone knows if FTP labels resources with a MIME type? (i.e., using something equiv. to HTTP's content-type)?
# [19:40] <Philip> All you need is a web service that translates any page into HTML5, and now all the pretty web pages on the web can be HTML5
# [19:40] <Philip> marcos: I'm almost entirely certain that it doesn't
# [19:40] <marcos> Philip: thanks.
# [19:40] <Philip> It's up to the client to even decide between ASCII and binary before downloading a file - the server just has a stream of bytes
# [19:42] <marcos> Will I get laughed at if I use the following definition in the widget spec?: In this specification, the network is any data transfer protocol that labels resources with a MIME type. For example, HTTP.
# [19:45] <Philip> (HTTP 0.9 didn't have MIME types)
# [19:45] <gsnedders> HTTP/0.9 didn't have anything.
# [19:45] <gsnedders> just GET /foobar
# [19:46] <Philip> marcos: If you have e.g. an MHTML file where you bundle a load of resources together with MIME types, would you count that as a "network"?
# [19:47] * marcos reads "each of us personally would like to see the XHTML2 and HTML groups brought together or at least the overlaps removed." I thought the role of the chair was to "chair", not to bring in their own agenda.... anyway.
# [19:47] <marcos> Philip: yes.
# [19:48] <marcos> Maybe "network" is not the right word there.
# [19:50] <marcos> I'm trying to make a distinction between things labelled with a MIME type and things not labelled. I currently have "network" and "Local storage or other media". Where "Local storage or other media is defined as any means of data acquisition that does not label resources with a MIME type. For example, from a hard disk or memory card, or over the air from BlueTooth or infrared."
# [19:56] <karl> marcos: ftp works only with two modes to the best of my knowledge: text or binary for transfert
# [19:58] <karl> Maybe it is one of the issues of the Web. Files have not been (are still not) created on the Web for a while.
# [20:05] <marcos> I'll run with what I have so far and see what happens at LC :)
# [20:06] <karl> hmm interesting, there is a mention of FTP in RFC 2046 - http://www.ietf.org/rfc/rfc2046.txt
# [20:06] <karl> 5.2.3.2. The 'ftp' and 'tftp' Access-Types
# [20:07] <karl> "The valid values for access-type "FTP" are
# [20:07] <karl> "ASCII", "EBCDIC", "IMAGE", and "LOCALn" where "n" is a
# [20:07] <karl> decimal integer, typically 8."
# [20:16] * Quits: tH (Rob@87.102.93.106) (Quit: ChatZilla 0.9.84-rdmsoft [XULRunner 1.9.0.1/2008072406])
# [20:31] <Philip> marcos: Can you call them "resources labelled with a MIME type" and "resources not labelled with a MIME type"?
# [20:31] <marcos> Philip: ya, I can do that.
# [20:32] <Philip> marcos: Sounds like the easiest way to make that distinction :-)
# [20:32] <marcos> Philip: true. I'll rewrite the section now.
# [20:33] <Julian> karl: man, that really has to predate the concept of URIs :-)
# [20:37] <karl> indeed
# [20:40] * Quits: smedero (smedero@192.223.6.251) (Quit: smedero)
# [20:40] <karl> I guess there are at least two ways: a server/client dialog where the server says "Yo man, I'm sending you this stream with this format/encoding" Or the protocol says "all your communications buddies must be done on 8bits, if you want more information about the content, inspect the first lines of the content"
# [20:46] <marcos> Apart from HTTP, what are some other protocols that label resources with MIME types?
# [20:48] <Julian> maros: email?
# [20:48] <marcos> Julian: yep, anything else?
# [20:48] <Julian> hmm.
# [20:49] <Julian> I guess protocols modeled after HTTP, such as SIP?
# [20:49] <Julian> What about XMPP?
# [20:49] * marcos takes a look
# [20:50] <Julian> looks like a job for a script that goes through all RFCs and looks for references to the MIME RFCs
# [20:51] * Joins: aaronlev (chatzilla@92.227.71.237)
# [20:52] <marcos> Julian: I'm just trying to add some examples to the widget spec of protocols that use/don't use MIME.
# [20:52] <marcos> nothing too serious
# [20:53] <Julian> ack
# [20:59] <Hixie> rubys: congratulations on your appointment
# [20:59] <karl> something like http://www.ask.com/web?q=%22RFC%202045%22%20%2Bsite%3Awww.ietf.org/rfc
# [20:59] <pimpbot> Title: "RFC 2045" +site:www.ietf.org/rfc - Ask.com Search (at www.ask.com)
# [21:00] * marcos impressed by Karl's fancy-pants usage of the Web :)
# [21:00] <karl> :D
# [21:01] <karl> make it pink and shiny. Disco man!
# [21:02] <Julian> Wow, even without Google's help :-)
# [21:03] <karl> what is Google? :p
# [21:03] <gsnedders> Not evil.
# [21:03] <gsnedders> :P
# [21:04] * Julian can't wait until Google will be topic of a Simpsons episode...
# [21:07] * Quits: gavin_ (gavin@99.226.207.11) (Ping timeout)
# [21:13] * Joins: gavin_ (gavin@99.226.207.11)
# [21:13] * marcos googles Google Simpsons and find interesting results
# [21:18] <rubys> Hixie: thanks!
# [21:23] * Quits: inimino (weechat@67.207.138.202) (Ping timeout)
# [21:27] * Joins: inimino (weechat@67.207.138.202)
# [21:29] <Hixie> rubys: did mike and chris fill you in on the current state of affairs or do you want a kind of "status update" briefing?
# [21:35] <rubys> not really; I've been loosely following the lists for over a year now... I guess they presumed that I knew what I needed to know.
# [21:37] <rubys> do you think a status update would be needed/helpful?
# [21:37] <Hixie> i dunno
# [21:37] <Hixie> i can say what i'm aware of if you like and you can decide for yourself :-)
# [21:38] <rubys> btw, I am planning to come to Google on the 28th and 29th (for ECMAScript), and could come a day or so early -- if that made any sense.
# [21:38] <rubys> 28th and 29th of january
# [21:38] <Hixie> well we should definitely meet for dinenr or something
# [21:39] <Hixie> if you're going to be in town
# [21:39] <rubys> i'd like that
# [21:39] <rubys> when my plans firm up, I'll let you know
# [21:40] * Dashiva already sees how this will end up on lastweek
# [21:46] <karl> Dashiva: you mean the cat woman suit of Sam or the romantic dinner between Ian and Sam ;)
# [21:46] * Julian doesn't see a conspiracy *yet*
# [21:46] <karl> hehe
# [21:47] <gsnedders> Julian: But if Hixie pays, then…
# [21:47] <gsnedders> Julian: It's all a ploy to control the chairs!
# [21:47] <karl> sending photos of separate bills on www-archive@w3.org would do ?
# [21:47] * Julian can be bribed with expsnive Sushi.
# [21:49] * gsnedders makes note in case he ever has reason to bribe Julian
# [21:49] * gsnedders then realizes that would mean going to a Sushi restaurant which he would despise
# [21:49] <Julian> no, we'll need video footage of both of them paying their individual bills
# [21:50] <Dashiva> Preferably a live feed?
# [21:52] <Hixie> rubys: http://www.whatwg.org/status-2008-12 is all i can think of of relevance right now
# [21:52] <Hixie> from the whatwg perspective
# [21:54] <rubys> re "today we have 924 people on whatwg and 383 on htmlwg"... this may double count some people, in particular, I'm in both groups, right?
# [21:54] <Hixie> yeah those numbers likely have huge overlap
# [21:58] <rubys> OK, I have a meeting I have to go to now, but that status is useful, and I'll try to collect my thoughts and present them similarly
# [21:59] <Hixie> cool
# [22:01] <rubys> ... my meeting just got pushed back 15 minutes
# [22:03] <Julian> So, while we're at it... Will there be a telco on Thursday?
# [22:04] <Julian> That would be a good opportunity for Sam to introduce himself...
# [22:05] <gsnedders> "Hi, I'm Sam Ruby, I'm the new divine ruler of the group."
# [22:05] <Hixie> i think e-mail would be more effective at reaching the community than a telecon, personally
# [22:05] <gsnedders> There probably aren't enough people on the telecon for it to be much use
# [22:06] <Julian> Email will be necessary as well, of course.
# [22:06] <Julian> But my assumption was that the new chair would attend the weekly telco, so the question really was: will there be one this week?
# [22:07] <rubys> I'm not co-chair until next month.
# [22:08] <rubys> i.e., ask the current chairs. :-)
# [22:08] <Julian> yes, I realized that after i hit \n
# [22:10] <gsnedders> Remember! Always remember!
# [22:10] <gsnedders> We shall not bow down before him yet!
# [22:12] <Julian> yet
# [22:33] * Quits: ROBOd (robod@89.122.216.38) (Quit: http://www.robodesign.ro )
# [22:39] * Quits: heycam (cam@203.217.82.242) (Quit: bye)
# [22:49] * Quits: tlr-bbl (tlr@128.30.52.30) (Quit: tlr-bbl)
# [22:58] * Quits: aroben (aroben@69.137.139.251) (Quit: Leaving)
# [23:01] * Joins: heycam (cam@130.194.72.84)
# [23:22] * Quits: sryo (sryo@190.245.204.198) (Connection reset by peer)
# [23:22] * Joins: sryo (sryo@190.245.204.198)
# [23:39] * Quits: karl (karlcow@70.81.84.168) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [23:39] * Quits: aaronlev (chatzilla@92.227.71.237) (Quit: ChatZilla 0.9.84-rdmsoft [XULRunner 1.9.0.1/2008072406])
# [23:43] * Quits: shepazu (schepers@128.30.52.30) (Quit: shepazu)
# [23:44] * Joins: shepazu (schepers@128.30.52.30)
# [23:57] * Joins: karl (karlcow@70.81.84.168)
# Session Close: Tue Dec 16 00:00:00 2008

The end :)