/irc-logs / w3c / #html-wg / 2007-04-02 / end

Options:

# Session Start: Mon Apr 02 00:00:00 2007
# Session Ident: #html-wg
# [00:03] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [00:04] <Hixie> hsivonen: if you can explain print, let me know
# [00:05] <mjs> MP is Mobile Profile right?
# [00:05] <Hixie> yes
# [00:05] <mjs> that is not only a design mistake, it is also a lie
# [00:05] <Hixie> hsivonen: and btw, the <font> wysiwyg stuff is just a strawman, i will be looking for better solutions in due course
# [00:05] <mjs> since the target devices clearly don't use an XML parser
# [00:06] <mjs> and it's also not for the web, since it is intended for walled garden content
# [00:06] <Hixie> mjs: it's made by the xhtml2 wg...
# [00:06] <Hixie> hsivonen: and we already have optional authoring tool configuration hooks: <meta>
# [00:07] <mjs> I admit I don't know what Print is for
# [00:09] <Lachy> I don't know what XHTML Basic, Print, 1.1 or any other profiles are for
# [00:10] <Hixie> Basic is for mobile phone network walled gardens, to provide checkbox-compliance feature lists
# [00:10] <Hixie> as in, to have a special thing that browser vendors can support that isn't expected to be compatible with the web
# [00:10] <Hixie> (because compatible-with-the-web is much harder)
# [00:21] <hsivonen> Hixie: meta is deeper in the document tree than the root. that's why the root works better for hooks
# [00:22] <hsivonen> Hixie: but then, it seems that oXygen, for one, can't hook configuration to an attribute
# [00:22] <Hixie> well PIs are the other solution
# [00:23] <hsivonen> it can hook configuration to any combo of filename extension, root namespac and root local name
# [00:23] <Hixie> but why wouldn't <meta> work?
# [00:23] <hsivonen> oXygen does PIs with paths
# [00:23] <hsivonen> Hixie: if you have a streaming tool, you have to look past the first element start
# [00:24] <hsivonen> PIs with schema file paths that is
# [00:26] <Hixie> we're talking about an editor here right?
# [00:26] <Hixie> so they have to load the whole file, surely
# [00:26] <Hixie> maybe i'm not understanding the use case
# [00:26] <Hixie> for XML tools, they can use PIs
# [00:26] <Hixie> I'm talking about HTML tools
# [00:26] <Hixie> (magic comments are another solution)
# [00:27] <hsivonen> Hixie: well, for editors, the editors would need more config to match a meta (e.g. arbitrary XPath or specific awareness of XHTML)
# [00:28] <Hixie> i was assuming we were talking about html editors here
# [00:28] <hsivonen> Hixie: for XML editors, special knowledge about XHTML meta is uglier than having a generic configuration for defining attribute patterns for the root element
# [00:28] <hsivonen> I was assuming we're also talking beasts like oXygen
# [00:29] <hsivonen> or Etna
# [00:29] <hsivonen> or nxml-mode
# [00:29] <hsivonen> that is, document-oriented generic XML editors
# [00:29] * Joins: dbaron (dbaron@71.198.189.81)
# [00:29] <hsivonen> I guess locking down features of Dreamweaver would be the HTML-only thing
# [00:30] <hsivonen> Hixie: the argument against PIs was that SOAP loses them
# [00:30] <hsivonen> (that sounds so very wrong)
# [00:38] <Hixie> for XML editors, they can use namespaces and proprietary attributes and markup and PIs, that's fine
# [00:39] <Hixie> it only affects the language if we're talking about HTML editors, imho
# [00:43] * Joins: cying (cying@75.18.219.204)
# [00:48] <hsivonen> Hixie: PIs yes.
# [00:48] <hsivonen> Hixie: proprietary attributes may get flagged as errors
# [00:49] <hsivonen> by other tools
# [00:51] * Quits: heycam (cam@203.214.79.176) (Ping timeout)
# [00:51] <hsivonen> Hixie: I don't *want* the attribute-based stuff. but if Jirka wants it, I don't have good enough objections
# [01:03] * Parts: hasather (hasather@81.235.209.174)
# [01:07] * Joins: st (st@62.234.155.214)
# [01:19] * Joins: Lachy_ (chatzilla@220.245.91.147)
# [01:31] * Joins: kingryan (kingryan@64.81.240.149)
# [01:31] * Joins: heycam (cam@130.194.72.84)
# [01:42] * Quits: h3h (bfults@70.95.237.98) (Quit: h3h)
# [01:58] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [01:59] * Joins: marcos (chatzilla@131.181.148.226)
# [02:00] * Joins: karl (karlcow@128.30.52.30)
# [02:09] <marcos> lachy or lachy_, you around?
# [02:10] <Lachy_> yes
# [02:36] * Quits: st (st@62.234.155.214) (Ping timeout)
# [02:37] * Joins: st (st@62.234.155.214)
# [03:07] * Joins: CameronB (cameron.bu@209.153.128.248)
# [03:17] * Quits: CameronB (cameron.bu@209.153.128.248) (Quit: using sirc version 2.211+KSIRC/1.3.12)
# [03:18] * Joins: cbulock (cbulock@209.153.128.248)
# [03:23] * Joins: Zeros (Zeros-Elip@67.154.87.254)
# [03:31] * Joins: olivier (ot@128.30.52.30)
# [03:53] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Ping timeout)
# [03:56] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [03:59] * Joins: Shunsuke (kuruma@58.1.246.135)
# [04:31] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Client exited)
# [04:33] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [04:40] * Quits: chaals (chaals@82.32.5.17) (Ping timeout)
# [06:20] <Hixie> hsivonen: it *should* get flagged as errors
# [06:20] <Hixie> hsivonen: it's proprietary markup.
# [06:39] * Quits: marcos (chatzilla@131.181.148.226) (Client exited)
# [07:05] * Quits: Shunsuke (kuruma@58.1.246.135) (Ping timeout)
# [07:13] * Joins: marcos (chatzilla@131.181.148.226)
# [07:48] <karl> http://www.w3.org/QA/Taxonomy.html
# [07:49] <karl> to remember.
# [07:50] <cying> http://www.digg.com/apple/iPhone_Software_Architecture_Leaked
# [07:50] <cying> (off topic)
# [07:50] <karl> cying: yes definitely. Maybe you should avoid ;)
# [07:50] <cying> oh yes
# [07:51] <cying> i meant this for another room, apologies
# [08:37] * Quits: dbaron (dbaron@71.198.189.81) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [09:00] * Quits: heycam (cam@130.194.72.84) (Ping timeout)
# [09:08] * Quits: kingryan (kingryan@64.81.240.149) (Quit: kingryan)
# [09:21] <Hixie> i wish the archive system linked threads across months
# [09:24] <Lachy_> yes, it would make more sense if the threaded view listed all threads that started in that month and included all messages from the following month(s)
# [09:25] <Hixie> i was more worried about just the "In Reply To" and "Replies" links, but yeah, that too
# [09:26] <Lachy_> oh, don't they do that?
# [09:26] <olivier> cvs -d :pserver:cvs@cvs.hypermail.org:/CVS checkout hypermail
# [09:26] <olivier> get to work guys :)
# [09:26] <sbuluf> wouldn't the advanced search form help?
# [09:26] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
# [09:30] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [09:33] * Quits: Lachy_ (chatzilla@220.245.91.147) (Quit: Chatzilla 0.9.77 [Firefox 2.0.0.3/2007030919])
# [09:34] * Joins: heycam (cam@203.214.79.176)
# [09:38] * Quits: cying (cying@75.18.219.204) (Quit: cying)
# [10:21] * Joins: anne (annevk@83.82.206.111)
# [10:27] * Joins: ROBOd (robod@86.34.246.154)
# [10:28] * Quits: sbuluf (logp@200.49.140.131) (Ping timeout)
# [10:31] * marcos is going to follow Mike's example and from now on EvryThingWillBAbbrtdInSomeAbritrryWay03 :P
# [10:32] <marcos> that way google can find things efficiently :P
# [10:32] <krijnh> :D
# [10:33] <krijnh> April fool is done you know ;)
# [10:34] <marcos> hehe, that's what I should have written!! "Ok Mike, April 1st is finished, ha ha. Enough with your jokes" :D
# [10:34] <krijnh> :p
# [10:34] <anne> heh
# [10:34] <marcos> Funniest thing is that google automatically flags Mike's emails as spam
# [10:34] <anne> if someone posts something silly today I'm going to use that one :)
# [10:34] <krijnh> anne: I'm sure someone will
# [10:35] * anne thought for a while he was a spammer (when he joined the WHATWG list)
# [10:35] <marcos> That must mean that lots of people are marking him as spam on gmail
# [10:35] <krijnh> anne: me too :|
# [10:35] <krijnh> Does he add SPAM-LOW himself?
# [10:36] <krijnh> Or is that an Opera thing?
# [10:36] <marcos> not sure.
# [10:39] <krijnh> I don't get why he likes those strange naming conventions, when he also has other principles; http://wiki.welldesignedurls.org/Principles
# [10:40] <krijnh> "Avoid Magic Numbers", "Hide Irrelevant Information"
# [10:41] <marcos> He subscribes to the "use magical numbers" and "hide all relevant information" and "make everything really hard to find later" principles.
# [10:42] <marcos> oh no! that's his!!!
# [10:42] <krijnh> It is, yeah
# [10:43] <marcos> well there is someone who does not practice what they preach
# [10:43] <Zeros> I thought his response suggested that this was something TAG has been doing for a while.
# [10:43] <marcos> LOL, I can't believe he cited the TAG!
# [10:47] <marcos> next time he replies, I'm going to quote some of his principles to him about URLs and why http://esw.w3.org/topic/HTML/DeclWebFormExpr04 is a crap URL... better still, I'll just ignore him and maybe he will go away.
# [10:50] <Zeros> hmm, the archive seems to be missing emails that were blocked and then accepted with the archive policy thing
# [11:00] <Lachy> Mike's problem is that he thinks he has the right and responsibility to dictate his own conventions to everyone else on the mailing list, instead of adhering to the widely accepted conventions that everyone else follows
# [11:37] <anne> I did think his topic change made sense though... But I suppose your comment wasn't specific to that...
# [11:43] * Joins: hasather (hasather@81.235.209.174)
# [11:58] <Lachy> my comment was more related to him changing the subject to include the wiki page titles, and somewhat related to him not quite formatting his replies properly by putting a blank line between quotes and responses to improve readability
# [11:59] * Joins: edas (edaspet@88.191.34.123)
# [12:12] * Joins: karl (karlcow@128.30.52.30)
# [12:26] * Parts: hasather (hasather@81.235.209.174)
# [12:31] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
# [12:42] * anne wonders what RelaxNG has to do with the default style discussion
# [12:59] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
# [13:00] * Joins: icaaq (icaaaq@217.13.228.226)
# [13:07] * anne thinks CSS definitions should be normative for a certain class of products...
# [13:08] * anne at least SHOULD, anyway
# [13:08] * gsnedders thinks at least SHOULD for all implementations that implement CSS
# [13:09] <anne> well, that's what "certain class of products" indicates, but yes
# [13:09] <gsnedders> it's unclear whether you included a superset or a subset, or meant the same
# [13:09] * gsnedders shudders at the thought of being off the net this week… so many emails to catch up :\
# [13:11] <anne> just delete the bunch
# [13:11] <anne> read the diff of web-apps and comment on what you don't like
# [13:11] <gsnedders> just skim the subject lines, more likely
# [13:11] <anne> unless you mean outside of HTML stuff
# [13:11] <gsnedders> I'm following the diffs as an implementor regardless
# [13:19] * Quits: edas (edaspet@88.191.34.123) (Ping timeout)
# [13:48] * Joins: marcos_ (chatzilla@203.206.31.102)
# [13:53] * Quits: gsnedders (gsnedders@86.139.123.225) (Quit: Don't touch /dev/null…)
# [14:10] * Joins: edas (edaspet@88.191.34.123)
# [14:11] <anne> yay, more feature requests :)
# [14:12] <krijnh> Super markup!
# [14:13] * Quits: edas (edaspet@88.191.34.123) (Ping timeout)
# [14:37] <Lachy> I wonder if Oliver's search for previous discussions was simply one search for "superstructure", which of course gave no results.
# [14:38] <anne> beyond me
# [14:38] * anne wants Bat markup
# [14:39] <Lachy> what confuses me most is that he's clearly seen the HTML5 spec since he linked to it, yet failed to realise that it has everything he asked for
# [14:39] <Lachy> except for skip links, which are redundant anyway
# [14:39] <anne> "seen"
# [14:39] <Lachy> ah, so he hasn't read it :-)
# [14:39] <anne> speccing XHR is pain
# [14:40] <anne> it's obvious impl have to fix something, but it's unclear what's going to be fixed
# [14:40] <Dashiva> Like how many times to fire readystatechange for 3?
# [14:40] <anne> that's settled
# [14:40] <anne> at 1
# [14:40] * Joins: zcorpan (zcorpan@217.211.77.236)
# [14:42] <anne> things like: http://tc.labs.opera.com/apis/XMLHttpRequest/open/027.htm
# [14:43] * Joins: edas (edaspet@88.191.34.123)
# [14:48] <Dashiva> Well, that one goes badly in gecko because they clear readystatechange handlers on open()
# [14:49] * Dashiva fiddles
# [14:49] <anne> I don't think so
# [14:49] <anne> if they do... well...
# [14:50] * anne was more concerned with IE
# [14:52] <Dashiva> Not signaling readystatechange when going 4->1, you mean?
# [14:53] <anne> when invoking the second .open(), yes
# [14:59] <Dashiva> And opera doesn't signal 0->1, and sometimes skips 3
# [15:02] <Dashiva> Typo in second paragraph of open() method. onreadystate instead of onreadystatechange
# [15:03] <anne> I don't think you're reading the latest draft
# [15:03] <anne> http://dev.w3.org/cvsweb/~checkout~/2006/webapi/XMLHttpRequest/Overview.html?content-type=text/html;%20charset=utf-8
# [15:04] <Dashiva> Indeed I'm not
# [15:24] * Joins: Shunsuke (kuruma@219.110.80.235)
# [15:32] * Joins: ROBOd (robod@86.34.246.154)
# [15:33] * Joins: gsnedders (gsnedders@86.139.123.225)
# [16:07] <gsnedders> yay! 0 unread emails! I can go on holiday and see how many I have on on Friday! Actually, another came in while typing this single line :(
# [16:33] * Quits: gsnedders (gsnedders@86.139.123.225) (Quit: Don't touch /dev/null…)
# [17:01] * Parts: icaaq (icaaaq@217.13.228.226)
# [17:50] * Joins: h3h (bfults@66.162.32.234)
# [17:56] * anne wonders why people are still proposing features...
# [17:56] <anne> it should be pretty obvious that the WG hasn't started yet...
# [17:56] <gavin_> it should?
# [17:57] <gavin_> there's a mailiing list, that's all some people need!
# [17:57] <Lachy> they probably don't realise we've already got a decent spec to start with
# [17:57] <Lachy> they probably think we're starting from scratch
# [17:57] <anne> most of the upload stuff he asked for is addressed iirc
# [17:57] <anne> except for uploading folders...
# [17:57] * anne isn't sure how that would work
# [17:58] <zcorpan> there should be a mandatory lurking period ;)
# [17:58] <beowulf> i plan to lurk for about 10 years
# [17:58] <Dashiva> We might have a WD by then, though. Better watch out
# [17:59] <Lachy> who would determine who has lurked long enough?
# [17:59] <anne> once MS joins I expect HTML5 will be "officially" submitted in some form to the WG as potential deliverable and then we'll see what happens...
# [17:59] <Lachy> if it isn't accepted as-is, this group will no-doubt fail
# [18:00] <anne> depends on what MS does
# [18:01] <anne> lol
# [18:01] <anne> "This sounds like it's beyond HTML WG and into XForms? (Not sure the actual term, but this is more technical than just HTML.)"
# [18:01] <zcorpan> ha
# [18:01] * anne waits for someone from the Forms WG to take this as an argument in favor of XForms Transitional
# [18:02] <Dashiva> Next we'll have XHTML2 transitional (and the xslt way doesn't count)
# [18:02] <anne> oh, maybe it's now time to say that Fools day is over...
# [18:03] <zcorpan> Dashiva: there already is xhtml2 transitional
# [18:03] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [18:03] <zcorpan> Dashiva: see the role module
# [18:03] * Dashiva cries
# [18:03] * Quits: edas (edaspet@88.191.34.123) (Quit: http://eric.daspet.name/ et l'édition 2007 de http://www.paris-web.fr/ )
# [18:03] <hsivonen> I think Raggett's vision for XForms Transitional assumes a product class where even a single product hasn't been proven workable, so I think it is premature to standardize interop within the product class
# [18:04] <anne> nice point
# [18:04] * anne -> food
# [18:04] <hsivonen> 1) Rename HTML5 to XHTML 2.0 Transitional, 2) ???, 3) Profit!
# [18:05] <Dashiva> What happens if XForms transitional is adopted, but it turns out the gap to XForms is too wide, and a new transitional is needed?
# [18:05] <zcorpan> then we'll have xforms transitional transitional
# [18:08] <zcorpan> the gap to xforms will always be too wide. you can't have a transitional format that is backwards compatible and takes you closer to xforms
# [18:09] <hsivonen> as far as I can tell, "Transitional" is marketing, politics or both
# [18:10] <zcorpan> "XForms" in the name too
# [18:10] <Lachy> all the best technologies has an 'X' in their names
# [18:10] <Dashiva> We'll have to hurry up the versions until HTML X then
# [18:11] <zcorpan> HTMLX!
# [18:11] <Lachy> to compete, we should call it "XWeb Forms 2.0"
# [18:11] <Lachy> it's a nice mashup of XForms and Web 2.0 :-)
# [18:11] <Dashiva> But it isn't transitional
# [18:12] <zcorpan> i've heard it being called "Web 2.0 Forms"
# [18:12] <Lachy> Web 2.0 XForms
# [18:12] <Lachy> Web 2.0 XForms + AJAX Transitional
# [18:13] <Dashiva> AJAXForms already exists, I think
# [18:13] <Lachy> http://ajaxforms.sourceforge.net/
# [18:28] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
# [18:33] * Quits: beowulf (carisenda@91.84.50.132) (Ping timeout)
# [18:35] * Joins: beowulf (carisenda@91.84.50.132)
# [18:41] * Quits: marcos_ (chatzilla@203.206.31.102) (Ping timeout)
# [18:43] <anne> So someone said screen readers widely ignore author stuff as they tend to get it wrong and here someone from WAI proposes an addition to HTML that assumes authors get it right...
# [18:44] <zcorpan> we should consult joe clark about how to solve accessibility related stuff
# [18:48] * Joins: marcos_ (chatzilla@203.206.31.102)
# [18:53] * Quits: beowulf (carisenda@91.84.50.132) (Ping timeout)
# [19:13] * Joins: kingryan (kingryan@66.92.187.33)
# [19:15] * Quits: ROBOd (robod@86.34.246.154) (Ping timeout)
# [19:17] * Joins: ROBOd (robod@86.34.246.154)
# [19:33] * Joins: mjs (mjs@17.255.99.50)
# [19:36] * Joins: Zeros (Zeros-Elip@67.154.87.254)
# [19:59] * Quits: anne (annevk@83.82.206.111) (Ping timeout)
# [21:03] * Joins: dbaron (dbaron@63.245.220.242)
# [21:14] * Quits: kingryan (kingryan@66.92.187.33) (Ping timeout)
# [21:20] * Joins: kingryan (kingryan@66.92.187.33)
# [21:35] * Quits: Shunsuke (kuruma@219.110.80.235) (Ping timeout)
# [21:37] * Joins: hasather (hasather@81.235.209.174)
# [22:04] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
# [22:12] <Hixie> lol @ mjs' <q> e-mail
# [22:13] <Zeros> :P
# [22:13] <mjs> Hixie: it was intended completely seriously
# [22:13] <Hixie> oh i know
# [22:13] <Hixie> still funny!
# [22:13] <mjs> but I doubt he'll follow my advice
# [22:14] <mjs> ok, Murray Maloney needs to chill with all the agreement messages
# [22:14] <mjs> I think he'd send a +1 for "black is white" at this point
# [22:17] <Hixie> wow, he agreed with my point that the whatwg should continue to contribute to html5
# [22:17] <mjs> I'm just waiting for him to agree with two exactly opposite statements
# [22:22] * Joins: DanC (connolly@128.30.52.30)
# [22:24] * Joins: anne (annevk@81.68.67.12)
# [22:24] <DanC> hmm... the idea of answering a question once and expecting most people to read it doesn't seem to be working all that well. People seem to ignore answers unless they're mailed specifically to them.
# [22:25] <Hixie> that explains the high volume
# [22:25] <Hixie> any news on chris, btw?
# [22:25] <DanC> not since the update of last weds, which called for some milestone today
# [22:26] <DanC> I started a message to him, but it sounded like nagging, so I'm gonna wait 'till tomorrow, I think
# [22:26] * anne stares at his inbox
# [22:26] * DanC reconsiders this request...
# [22:26] <DanC> [[
# [22:26] <DanC> For the benefit of the users who are new to reading subject matter in
# [22:26] <DanC> this format, what would the group suggest should be researched first, in
# [22:26] <DanC> terms of best understanding where the HTML Working Group is heading?
# [22:26] <DanC> ]]
# [22:27] <DanC> -- http://lists.w3.org/Archives/Public/public-html/2007JanMar/0069.html
# [22:27] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
# [22:27] * Joins: Zeros (Zeros-Elip@67.154.87.254)
# [22:27] <DanC> if I put together some materials, that won't be enough; somebody will still have to mail pointers to the materials in response to every newbie question
# [22:27] <Hixie> we could have a FAQ on the wiki
# [22:28] <DanC> I'm willing to work on the materials if a few people are willing to gently welcome the newbies by pointing to it.
# [22:28] <DanC> hmm... maybe I should re-phrase http://www.w3.org/html/wg/il16 in FAQ form
# [22:29] <Hixie> danc: (there's an whatwg faq you're welcome to crib from if you want, i think the license for that text is the mit license)
# [22:29] <Hixie> DanC: (http://blog.whatwg.org/faq/)
# [22:29] <Hixie> DanC: (lachy maintains it)
# [22:30] <DanC> ugh... I followed a pointer to http://www.whatwg.org/specs/web-apps/current-work/#relationship0 and my CPU got pegged
# [22:31] <DanC> I can't be the only person complaining about the size of the HTML5 spec. or is this a firefox bug of some sort?
# [22:31] <Hixie> the spec is too large, yes
# [22:32] <Hixie> there's an open call for volunteers to write a spec splitter, but i'll get around to splitting it into chapters in due course if nobody else does it first
# [22:32] <DanC> ok
# [22:32] <Hixie> it's a harder problem than it otherwise would be because bert's cross-referencing and section numbering script, which i use, doesn't support chapters.
# [22:33] <DanC> I use XSLT for that sort of thing...
# [22:33] <DanC> and lots of XSLT engines have HTML parsers glued on.
# [22:34] <DanC> odd... http://wiki.whatwg.org/wiki/What_you_can_do doesn't mention "read the spec and send comments"
# [22:34] <Hixie> people don't seem to need to be told that :-)
# [22:34] <Hixie> feel free to add it though
# [22:35] <Hixie> right, lunch time
# [22:35] <Hixie> bbl
# [22:35] * Parts: zcorpan (zcorpan@217.211.77.236)
# [22:35] * anne laughs at the <q src> follow-up
# [22:35] <anne> "i wouldn't classify it as breaking compatibility, but as fixing an error in the 4.01 technical recommendation."
# [22:36] * DanC wishes for a test case to back the compatibility/incompatibility claims around <q src>
# [22:36] <anne> Who do you believe? The Safari developer or the accessibility guy?
# [22:36] <anne> well HTMLQuoteElement.cite returns a resolved URI per DOM2HTML
# [22:37] <mjs> DanC: loads fine for me
# [22:37] <anne> not just the attribute's value
# [22:37] <mjs> (and WebKit is I think only 2-3x as fast at raw page load as Firefox, so I'd be surprised if it pegs the CPU for any significant length of time - maybe something specific to the markup)
# [22:38] <DanC> it takes a lot of time to learn who's an expert in what. test cases reduce that sort of thing to black-and-white data.
# [22:38] <DanC> test_case_wishes += "HTMLQuoteElement.cite returns a resolved URI per DOM2HTML"
# [22:38] <mjs> DanC: changing an attribute that is specified to be a URI to one that is free-form text seems like clear compatibility breakage
# [22:39] <mjs> DanC: assuming anyone uses it
# [22:39] <mjs> the only counter-argument to the compat issue would be showing that authors don't use the cite attribute, or that no UA (including non-browser UAs) parses it
# [22:39] <DanC> yes, it seems clear to me too; so let's use a test case to capture the decision.
# [22:39] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
# [22:40] <DanC> or rather: to capture the issue, to make a decision easy to record and preserve.
# [22:40] <mjs> what would the test case show?
# [22:40] <anne> also, dropping <q> would make more sense than making it even more complicated with features nobody will use
# [22:40] <mjs> that if you do <q cite="Encyclopedia">, the DOM result is weird?
# [22:40] <DanC> yes
# [22:40] <mjs> I wouldn't know how to check whether search engines (say) excpect cite to be a URI or do anything with it, or whether they would find a URI in src too
# [22:41] <mjs> in any case, I don't think anyone denies the compat issue, just whether it is more relevant than the old spec allegedly being in error
# [22:41] <DanC> I haven't seen anybody deny the compat issue, but it's easy to _forget_.
# [22:42] <anne> also, why not use title="" for that information...
# [22:43] <DanC> the meaning of <cite> has drifted significantly since I added it to HTML. It was introduced via TeXinfo from, but it goes back to the chicago manual of style. It's for marking up titles of works.
# [22:43] <DanC> that is: it was originally for marking up titles of works.
# [22:43] * Joins: Philip (excors@80.177.163.133)
# [22:44] <DanC> I was asleep at the wheel when the <cite>Romeo</cite>... <cite>Juliet</cite>... examples got added to HTML4
# [22:44] <anne> woho, <cite> is different from <q cite>
# [22:44] <DanC> oops; right; that's a whole other topic.
# [22:44] <hsivonen> the spec can't be too large. I've browsed it on my Nokia 770 with 64 MB of RAM
# [22:44] <anne> DanC, cool, hsivonen proposed it to be used for title of works
# [22:45] * Hixie wonders if paul nelson in www-html doesn't know what application/* is, in MIME
# [22:45] <anne> (as in, redesigning what it's for)
# [22:45] <anne> as you don't really want to italicize names of people anyway
# [22:45] <DanC> The titles of works that stand by themselves, such as books or
# [22:45] <DanC> newspapers: "There was a performance of Beethoven's Ode to Joy."
# [22:45] <DanC> -- http://en.wikipedia.org/wiki/Italics
# [22:46] <DanC> http://uw714doc.sco.com/cgi-bin/info2html?(texinfo)cite〈=en
# [22:47] * anne wonders why Murray Maloney doesn't provide any arguments whatsoever
# [22:47] <hsivonen> If I get a good explanation why my use of <cite> in my thesis draft is wrong, I can do a global replace to <i>
# [22:47] <hsivonen> that would be fun to explain to semanticists
# [22:47] <hsivonen> currently, I am on the fence
# [22:48] <hsivonen> but honestly, I don't see the default style of <cite> fitting semantics other that "title of work"
# [22:48] <hsivonen> since 1993
# [22:48] <gavin_> geez, I though the empty "+1" messages were bad, but now we're getting empty "-1" messages?
# [22:48] <anne> yeah, crazy
# [22:48] <anne> omg
# [22:48] <Hixie> must've missed that one
# [22:49] <anne> this Gregory guy continues
# [22:49] <DanC> the odd thing is, I have never seen Murray use either +1 nor -1 before my message explaining where they come from and asking that people say more than just +1/-1
# [22:49] <anne> replace <hr> with <ls>
# [22:49] <hsivonen> anne: <ls>?
# [22:49] <anne> logical separator
# [22:49] <Hixie> i agree with the proposal to replace <hr> with <ls>, though i would ask, for back compat reasons, that we use the name 'hr' instead of 'ls'.
# [22:49] <anne> see his latest e-mail
# [22:50] <hsivonen> I don't understand Murray Maloney's -1 to me.
# [22:51] <hsivonen> after all, no one but Hixie has come forward with a credible time commitment and track record for the job
# [22:51] <anne> it's pretty hard to get a +1 or -1 without arguments
# [22:51] <anne> therefore they're best ignored
# [22:51] <DanC> yes, they're poor use of reputation capital
# [23:00] * anne doesn't see any of the messages from Gregory at http://lists.w3.org/Archives/Member/w3c-html-cg/2007JanMar/
# [23:07] <mjs> DanC: I do personally think that anyone who disagrees with the basic goals of the group should go away
# [23:07] <mjs> DanC: otherwise they will waste a lot of our time and theirs
# [23:08] <mjs> I would rather just tell them so than pretend otherwise
# [23:08] <Philip> For HTML5 script type defaulting to text/javascript, is that actually in the HTML5 spec anywhere? I just see a red box with "How to handle the type and language attributes should be defined here, probably with reference to the next section."
# [23:08] <anne> The type attribute is optional.
# [23:08] <Hixie> it's not yet defined iirc
# [23:08] <anne> I'm not sure if script execution is defined yet...
# [23:09] <DanC> the basic goals of the group aren't yet established (except inasmuch as they're stated in the charter)
# [23:09] <anne> prolly also involves stuff like onload=""
# [23:09] <Hixie> script execution is defined, but the type= and language= attribute processing isn't yet
# [23:09] <Hixie> onload= is mostly defined
# [23:09] <anne> (I meant the type of script it takes, sorry)
# [23:10] * anne should probably have said onfoo=""
# [23:10] <Hixie> the language and processing of event handler attributes is defined.
# [23:11] <anne> didn't you want to allow different languages?
# [23:12] <Hixie> we could, i haven't thought about that yet
# [23:12] <Hixie> if we do, we'll update that part of the spec i guess
# [23:16] <mjs> DanC: I think the group is unlikely to succeed if the basic goals of the group end up welcoming compatibility breakage, but we can make an official decision when our co-chair arrives (or lets us know that he won't be showing up)
# [23:16] * Joins: beowulf (carisenda@91.84.50.132)
# [23:17] <hsivonen> it would help a lot if the HTML5 spec was required reading for newbies
# [23:18] <Hixie> good luck with that
# [23:19] <DanC> mjs, your reply about Q didn't cite your proposed principles... it's good to connect the detailed discussion to the principles when they apply...
# [23:19] <Hixie> i'm sure most people on the wg haven't even read the html_4_ spec cover to cover
# [23:19] <anne> i haven't
# [23:19] <Hixie> case in point :-)
# [23:19] <DanC> I know how to actually enforse "required reading" for meetings; I'm not sure how to do it for email discussion
# [23:19] <DanC> enforce
# [23:20] <Hixie> how do you enforce it for meetings?
# [23:20] <Hixie> we could use that in csswg meetings
# [23:20] <DanC> for ftf meetings, I make a list at T-2 weeks...
# [23:21] <DanC> ... the list can't be too long, and stuff can only get on it for good reasons.
# [23:21] <DanC> ... but once stuff is on it, and I get the WG to agree to the list (in a telcon, usually), I hold people to that commitment.
# [23:21] <Hixie> what do you do to hold them to it?
# [23:21] <DanC> e.g. if they show that they haven't read it, I cut them off and go to the next speaker, perhaps.
# [23:22] <DanC> or if somebody asks a question that's clearly answered in the draft, I don't take meeting time for the answer.
# [23:22] <Hixie> hm.
# [23:22] <Hixie> i wish bert did that.
# [23:22] <Hixie> not sure the other people on the group would stand for it though
# [23:23] <DanC> that's why I get them to agree up fron
# [23:23] <DanC> after all, "the needs of the many outweight the needs of the few or the one"
# [23:23] <DanC> (modulo spelling errors)
# [23:23] <Hixie> you should give chairing courses
# [23:23] <Hixie> and have other w3c wg chairs take your course
# [23:23] <Hixie> (seriously)
# [23:23] <DanC> I do.
# [23:24] <DanC> in coordination group meetings.
# [23:24] <DanC> I once did it for all chairs, but the group of chairs has become too large.
# [23:24] <Hixie> hm
# [23:24] <Hixie> well it hasn't had much effect on the groups i'm in :-(
# [23:25] <DanC> yeah; I haven't been to hypertext CG meetings for a while. I resumed a couple weeks ago.
# [23:25] * anne thinks DanC is doing a nice job there
# [23:26] <DanC> :)
# [23:26] * DanC makes a mental note to document the readinglist pattern
# [23:31] <Hixie> so we have http://lists.w3.org/Archives/Member/w3c-html-cg/2007JanMar/0304.html http://lists.w3.org/Archives/Member/w3c-html-cg/2007JanMar/0305.html
# [23:31] <Hixie> and we have people telling us that xml's error recovery is a good thing.
# [23:31] <Hixie> gah
# [23:31] <DanC> yay. mjs sent a test case.
# [23:31] <anne> Hixie, yeah, that was funny
# [23:32] <anne> Hixie, also all the people advocating to use a browser with better error recovery...
# [23:32] <Hixie> anne: that's what i meant, yeah
# [23:32] <anne> I believe they referred to using Firefox as opposed to Internet Explorer, but Firefox 3 will show a blank page there as well...
# [23:33] <Hixie> especially funny is that the error recovery they want is the one that introduces a security flaw
# [23:33] <Hixie> so they are advocating error recovery over security, even
# [23:33] <Hixie> but when it comes to writing specs, suddenly they're against it!
# [23:34] * anne recalled the security flaw but wasn't sure how it worked again
# [23:34] <anne> <style><script> youdontwantthistoberun() </script>
# [23:34] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Ping timeout)
# [23:34] <Hixie> yeah
# [23:34] <Hixie> then DOS the browser before the </style> tag
# [23:34] <Hixie> or the connection, rather
# [23:35] <anne> but couldn't you just invoke the function using UserJS or Greasemonkey?
# [23:35] <Hixie> e.g. send an out-of-sequence packet to close the connection
# [23:35] <Hixie> ?
# [23:35] <Hixie> you don't have access to the machine
# [23:35] <mjs> xml has error recovery?
# [23:35] <Hixie> this is a man-in-the-middle attack
# [23:35] <Hixie> mjs: no
# [23:35] <anne> ah ok
# [23:35] <Hixie> mjs: that was a text/html document
# [23:36] <Hixie> mislabelled as such, but that's a separate story of ironic hypocrisy
# [23:37] * Quits: kingryan (kingryan@66.92.187.33) (Quit: kingryan)
# [23:41] * Joins: kingryan (kingryan@66.92.187.33)
# [23:44] <DanC> yay. dbaron sent a (pointer to) a test case.
# [23:44] * Quits: st (st@62.234.155.214) (Quit: st)
# [23:45] <Hixie> that actually was a testcase, yeah
# [23:46] <DanC> hmm... nxml-mode doesn't like <script> inside <body>
# [23:46] <anne> we could also drop the quotation marks from <q>
# [23:46] <anne> IE doesn't do it...
# [23:46] * Joins: marcos__ (chatzilla@203.206.31.102)
# [23:47] <DanC> I like the <q> element. Unfortunately, I find dbaron's argument pretty compelling.
# [23:47] <DanC> ah... nxml-mode just wants a "type" on script...
# [23:48] <anne> yeah, HTML4 requires that...
# [23:48] * Joins: _cupid_ (on@82.69.25.35)
# [23:48] * _cupid_ is now known as chrisheilmann
# [23:48] <anne> (there was actually an e-mail about that on public-html today :) )
# [23:48] <chrisheilmann> evening
# [23:48] * anne just reads that Steven actually said "validity error" (re: the missing </style> tag)
# [23:49] <DanC> btw... we were talking about "no versioning" as a principle, but the existing principles start "New versions of HTML must not ..."
# [23:49] <DanC> http://esw.w3.org/topic/HTML/ProposedDesignPrinciples
# [23:50] <anne> Maybe it should be renamed to "no document versioning"
# [23:50] <anne> or something
# [23:50] * marcos__ waves hello
# [23:51] <anne> or we could call it "HTML Level 5" ...
# [23:51] <gavin_> HTML Profile 5
# [23:52] <DanC> well, the HTML 5 spec includes all/much/most of HTML 2, 3, 3.2, and 4, no?
# [23:52] <dbaron> "HTML 5"
# [23:52] * Quits: heycam (cam@203.214.79.176) (Ping timeout)
# [23:53] <DanC> I wonder if "new HTML features..." works better than "new HTML versions..."
# [23:53] <anne> DanC, yes
# [23:53] <kingryan> DanC: how about "no schema versioning"?
# [23:53] <chrisheilmann> HTML XP
# [23:53] <Hixie> DanC: "no versioning" refers to no syntactic features to indicate version, not to the lack of an actual version of the spec
# [23:53] * anne saw someone suggesting HTMLX today
# [23:53] * Joins: heycam (cam@203.214.79.176)
# [23:53] <DanC> "schema" is a dirty word in this context, I think. at lest: it introduces unnecessary complexity
# [23:53] <Hixie> "No versioning syntax" would be a better title
# [23:53] <DanC> or "no version markers"
# [23:54] <DanC> no, syntax is better
# [23:54] <anne> DanC, you attached a videogame checklist...
# [23:54] <DanC> oh my.
# [23:54] <DanC> LOL
# [23:54] <anne> don't worry, it's funny
# [23:54] <DanC> I hope it's 1/2 as funny for the rest of you as it is for me
# [23:56] * anne thinks most people refer to it as HTML5 btw... without the space between HTML and 5
# [23:56] <chrisheilmann> Ok, I now need to find the company expert to fill out my form so I can join
# [23:56] <chrisheilmann> this is complex :)
# [23:56] <DanC> do you know where to look that up, chrisheilmann ? want/need help?
# [23:57] <chrisheilmann> I need to check in the office tomorrow
# [23:57] <chrisheilmann> I talked to Douglas Crockford at the F2E conference in Sunnyvale last month, we are interested
# [23:57] <anne> you're at Y!?
# [23:58] <chrisheilmann> yes Yahoo UK
# [23:58] * anne believes that might get tricky
# [23:58] * DanC hopes for the best
# [23:58] <chrisheilmann> yeah, Doug said the same :)
# [23:58] <chrisheilmann> licensing issues or whatever
# [23:58] <anne> patent policy...
# [23:58] <chrisheilmann> I can always apply for "invited expert"
# [23:58] <DanC> let's put it this way, chrisheilmann: if you join, represeting yahoo, I'm gonna ask for a raise ;-)
# [23:59] <anne> mnot was always "observer" in the Web API WG...
# [23:59] <DanC> the invited expert route is not significantly less tricky
# Session Close: Tue Apr 03 00:00:00 2007

The end :)