IRC Logs

00:00 TabAtkins : Dude, you and tantek were my gods when I started webcoding. You guys were the first blogs I stumbled across that made sense.

00:02 TabAtkins : So, would it be *acceptable* to have <div><nav /><aside .blogroll /><aside><blockquote /></aside></div>? Just trying to ascertain the flexibility that is intended here.

00:02 hober : now that pubdate="" is on <time>, I think it would make more sense if it applied to the nearest ancestor <article> element, or <body> if there is no such ancestor

00:03 TabAtkins : It already applies to the nearest ancestor <article>. Applying it to <body> is definitely needed, though, as <body> implies the semantics of <article>.

00:03 hober : I'm imagining a page that is just a document, where <body> is the sectioning element that represents the entire document. there's no need to add an <article> descendent that contains all descendents of <body>...

00:03 TabAtkins : hober: yeah

00:04 hober : I guess I'll write an email

base : business a long time ago, before he started running into difficulties with content moving between paths.

00:07 Hixie : TabAtkins: if you find yourself needing <div>, then you're not doing markup before style. :-P

00:07 TabAtkins : Hixie: The issue is needing to adjust markup *after* styling fails due to limitations in CSS.

00:07 Hixie : ah

00:07 TabAtkins : When Template Layout is reliably present I'll be happy. ^_^

00:08 TabAtkins : Yeah, so given an initial design where I have a <nav> and two <aside>s, is it okay to just <div>-wrap them to make them display properly, or do I *need* to <aside>-wrap them instead, and peel off the <aside>s wrapping the individual blocks.

00:09 hober : sent (re pubdate="")

00:09 Hixie : TabAtkins: if you're adding elements purely for styling purposes, you should use <div>

00:09 Hixie : that's what <div> is for

00:10 TabAtkins : Yeah, I know, just making sure that, even though there's an element *for* sidebars, it's still cool to collect things into a sidebar without it.

00:10 TabAtkins : I have tiny niggling inconsistencies in my brain that I'm ironing out right now.

00:11 hober : well, you should always use the element with the closest semantic to your need

00:11 ttepasse : The <main>-proposal got shot down, isn't it?

00:11 tantek : but not *over* semantic

00:11 TabAtkins : At least for now, yeah, ttepasse.

00:11 hober : tantek: indeed

00:12 TabAtkins : hober: I think tantek captured the point behind my questions. ^_^

00:12 tantek : TabAtkins - I don't like <main> either, but could have settled for <shead>, <sbody>, <sfoot>

00:12 TabAtkins : ttepasse: main content is anything that's not <header>/<footer>/<nav>/<aside>

00:13 tantek : TabAtkins - I *guarantee* that most web devs will just end up creating a <div class="content"> or something similar to go between <header> and <footer>

00:13 tantek : it such a well established markup pattern

00:13 ttepasse : ?anything that's not <header>, etc.? is not easily adressable by JS oder CSS.

00:13 TabAtkins : I do it about 50/50 depending on whether I have styling ned for it.

00:13 tantek : in fact, I doubt you would find many class="header" and class="footer" on a page without an intervening class="content" or class="main" etc. (s/class/id as apropos)

00:14 TabAtkins : ttepasse: Yeah, I know. But it's sensical to ATs, frex.

00:14 tantek : good point ttepasse

00:14 ttepasse : ATs, frex?

00:14 TabAtkins : AT = accessibility, um, I forget what the T is for. Frex = for example.

00:14 krijn : What's it about?

00:14 TabAtkins : krijn: Pancakes or waffles: which is better?

00:14 tantek : accessibility tools perhaps

00:15 krijn : Pancakes!

00:15 TabAtkins : Good, you and I will flank the infidels.

00:15 krijn : I agree a <main> or <content> thingy would be handy

00:15 TabAtkins : Hahahahahaha Myakura.

00:15 krijn : Now see what you've done :)

00:16 tantek : TabAtkins - is that some sort of subconscious circle vs square preference? or perhaps flat (pancake) vs cellular (waffle)

00:16 TabAtkins : I use a <div .main/.middle/.content/.block> for the non-foo stuff whenever I need it for a CSS or JS hook, yeah.

00:16 krijn : Me too

00:16 TabAtkins : It's about a genital vs oral preference. Freud was all over this.

00:17 krijn : Waffles remind me of tables

00:17 hober : yes, because the whole purpose of them is to deliver maple syrup (non-sarcastic)

00:17 krijn : And tables are evil, it seems

00:17 tantek : krijn HTML tables are simply misunderstood ;)

00:17 krijn : No, they are evil

00:17 hober : krijn: fortunately, maple syrup is tabular data

00:17 krijn : According to authors at least :)

00:17 hober : so marking it up in a waffle is approprate

00:17 krijn : So, concluding, we need a <waffle> tag

00:18 TabAtkins : I support this proposal.

00:18 krijn : Can we overrule Hixie without discussion already?

00:18 krijn : Or are we still not there yet?

00:18 TabAtkins : Let's vote.

00:18 TabAtkins : Aye for <waffle>, nay for <hixie>.

00:18 krijn : Aye

00:19 krijn : For the record, beer and aye saying don't match

00:19 TabAtkins : A <hixie> element ridicules your use of presentational markup. It's like <audio> but demeaning.

00:19 beowulf : someone say waffles?

00:20 TabAtkins : Yes, how do waffles relate to the struggle to evangelize semantic markup, and the difficulties contained in keeping your fingers from getting sticky when using <td>s.

00:23 ttepasse : I hope <hixie> has a textual fallback.

00:23 krijn : Oh look, contaminants!

00:23 krijn : :)

00:25 krijn : Btw, the biggest liquor store in the Netherlands now uses HTML5 for their website \o/

00:25 TabAtkins : Woo! \o/

00:25 krijn : In case anybody was waiting for that information

00:25 krijn : Thanks everybody in here for making that happen :)

00:30 krijn : And before I get to emotional, nn everybody

00:30 TabAtkins : later

00:31 krijn : o.

00:31 krijn : o/

00:31 ttepasse : Why are there such strong restriction for permissible aria roles on some elements?

00:32 ttepasse : By my reading I can't do <table role="grid main"> or am I wrong?

00:37 TabAtkins : Where's that section again? I can never remember the name of it so I can't find it in the TOC.

00:37 ttepasse : 3.2.6

00:40 TabAtkins : Hmm, so you're saying that the table itself is the main content all by itself?

00:42 ttepasse : Yes. Think about a better markuped webserver directory listing. There may be additional stuff like a logo and navigation but the main content is the listing of files and directories.

00:42 Hixie : send mail if you think the restrictions should be relaxed

00:44 tantek : TabAtkins - the obv use case for a <table> being the main content all by itself is a CSV file that's been transcoded into an HTML <table>

00:44 TabAtkins : Yeah, I was thinking something like that.

00:44 Hixie : can a role="" take two values btw? i thought it could only take one.

00:45 TabAtkins : Well, we've got a use-case for such right here.

00:45 ttepasse : /xhtml-role/ has even an example with two, so yes.

00:45 TabAtkins : A combination of a type and landmark role.

00:45 tantek : in fact, I wouldn't mind seeing a <csv> ... </csv> element that acted like a <pre> except 1. parsed for the comma separation and implied columns accordingly, and 2. provided a DOM similar to <table> for accessing all the rows and columns.

00:45 TabAtkins : ...

00:46 TabAtkins : I must code support for this tonight.

00:46 Hixie : i wonder what the UA conformance criteria are for two roles

00:46 tantek : This would rapidly accelerate the "webification" of the numerous CSV files out there (more and more being published due to OpenGov and OpenScience) efforts

00:46 Hixie : i don't understand what it would mean

00:48 hober : ttepasse: keep in mind that aria's role="" isn't the same attribute as the xhtml2wg role attribute spec

00:48 tantek : hober, thank goodness.

00:48 ttepasse : Oh it is?

00:48 hober : tantek: :) srsly people, we already got one (class="")

00:49 ttepasse : There should be a "/TR/ for dummies" tutorial.

00:50 ttepasse : Hixie, http://www.w3.org/TR/2009/WD-wai-aria-implementation-20090224/#mapping_role

csv : to the list of (perhaps crazy) proposals he has to write up for HTML5.

00:51 Hixie : ok so basically they allow it and then immediately say it's pointless

00:51 Hixie : good work

00:52 tantek : Hixie, of course, there's nothing like that in the HTML5 spec, <cough>img border="0"</cough>

00:52 tantek : ;)

00:55 ttepasse : ?Because landmarks do not generally have mappings in platform accessibility APIs, they may occur anywhere within the role string without affecting how the first widget role is exposed.? ... sounds like my <table role="grid main"> is not so pointless but makes me wonder what landmarks are for. Custom UI in UAs?

00:55 Hixie : tantek: blame hsivonen, i hate that as much

00:56 tantek : Hixie - is there "blame" markup in the HTML5 spec? Kind of like blame in CVS...

00:57 Hixie : svn blame blames me for everything, but at least one of the checkin comments says "blame: hs" :-)

00:59 Hixie : btw, tantek, any idea when zeldman and co are sending their feedback in?

01:01 Hixie : tantekc: btw, any idea when zeldman and co are sending their feedback in?

01:01 tantekc : Hixie, I believe we are all taking bits and pieces of http://www.zeldman.com/superfriends/guide/ and emailing them in - hence the "we are optimistic that the official channels provided by the working group will offer a sufficient and fair hearing of our concerns"

01:01 Hixie : k

01:02 Hixie : any idea what the time frame on that is?

01:02 tantekc : As you know, I prefer to document things on the web first, and then email in URLs.

01:02 tantekc : timeframe: as soon as possible

01:02 Hixie : if you could mail in text, that would make my life much easier, fwiw

01:02 tantekc : ok I will mention that you prefer plain text email

01:02 Hixie : thanks

01:03 tantekc : Hixie, have you documented your Communication Preferences anywhere online? E.g. http://tr.im/comms

01:03 tantekc : if you have, please give me the URL and I will happily link to it

01:03 Hixie : tantekc: "Comments are very welcome, please send them to whatwg@whatwg.org. Thank you."

01:03 Hixie : (top of the spec)

01:04 tantekc : I meant in general, your personal prefs.

01:04 tantekc : because obviously you eventually look at bugs filed, blog posts, blog post comments etc.

01:04 tantekc : (perhaps not guaranteed, but that seems to be the emergent behavior)

01:04 Hixie : oh you mean in general? no, no preferred communication style.

01:05 Hixie : i have an antipreference for phone communication and anything involving burning gasoline or going through TSA checkpoints

01:05 tantekc : so the only communication preference you have has to do with the HTML5 project

01:05 Hixie : yeah

01:05 Hixie : well, i mean, my preferences are specific to whatever i'm doing

01:05 tantekc : you've written up a scathing deconstruction of telcons for example (most of which I agree with)

01:05 Hixie : (i have a preference for HTML5, simply because there are certain things that are easier to dealing with)

01:05 Hixie : i wrote an e-mail recently... let's see

01:06 tantekc : anyway - it might be useful to see them in an ordered list

01:06 tantekc : with perhaps a threshold for, -- below this point I will avoid these forms of communication --

01:06 Hixie : http://lists.w3.org/Archives/Public/public-html/2009Jun/0616.html

01:06 tantekc : and then you can list phone, f2f, flying etc.

01:07 tantekc : yeah that one - I read it.

01:07 tantekc : a good write up

01:07 Hixie : e-mail, irc, jabber, aim, msn, yahoo messenger; twitter, blogs; forums i have to register for; meetings. I don't have a phone.

01:07 Hixie : where ";" indicates a threshold, so all things between ; are equally ok

01:08 TabAtkins : tantekc: http://www.xanthir.com/etc/csv.html

01:09 tantekc : anyway - g2g, but just wanted say thanks Hixie for already fixing a bunch of stuff in HTML5 that Zeldman et al gave feedback on (e.g. the "HTML5" vs. "HTML 5" inconsistency) - pretty much all that feedback was a consensus result of the "HTML5 Super Friends"

01:09 TabAtkins : Not perfect yet - I need to copy over all attributes of the original <csv> element - but still, working!

01:09 tantekc : TabAtkins - WHOA! that is super cool

01:09 TabAtkins : Super easy. ^_^

01:10 tantekc : I might wrap the <csv> in a <pre> for backward compat display

01:10 TabAtkins : Would be fine - you sorta expect csvs to be fixed-width anyway.

01:10 tantekc : right

01:10 tantekc : or maybe just use a style sheet

01:10 tantekc : csv {whitespace:pre}

01:10 tantekc : that ought to do it

01:11 TabAtkins : Yeah, that'd work.

01:11 tantekc : TabAtkins - want to collaborate on a joint <csv> proposal submission for HTML5?

01:11 tantekc : like I said, g2g now, but I can start drafting something based on the above

01:11 TabAtkins : Note - that would make the original CSV, and presumably a native CSV, look pre, but wouldn't affect the js-shim csv.

01:11 TabAtkins : Sure, tantekc.

01:13 tantekc : scientists will *love* <csv> for easily publishing data sets on the web

01:13 tantekc : this is going to be a big enabler for open science on today's web. (my current thoughts: http://tr.im/openscience )

01:13 TabAtkins : Heh, cool.

01:17 TabAtkins : So, how do I iterate through a DOM Node's attributes?

01:18 TabAtkins : node.namedNodeMap()?

01:21 othermaciej : node.attributes

01:21 othermaciej : (or really, element.attributes)

01:21 TabAtkins : Since I program in jQuery, not js, does that return an array? Or some exotic DOM structure? If the latter, how do I iterate through it?

01:22 Binarytales : i think it's an array

01:22 TabAtkins : It looks like it at least acts like one. $.each() can handle anything with .length

01:23 othermaciej : it's a NamedNodeMap; I don

01:23 othermaciej : 't know about jQuery details

01:26 TabAtkins : All right, got it. Yeah, $.each() handles it fine.

01:28 TabAtkins : Though I should probably translate this to plain DOM...

01:29 TabAtkins : I need to go look at what jQuery actually does for each of these functions so I can back-translate.

01:39 TabAtkins : Hrm, does IE treat linebreaks funny? I'm trying to do a .split("\n"), but it's a no-go in IE. Perhaps auto-translating to \r\n?

01:39 TabAtkins : Context: textContent of an unknown element, hacked into recognition with the document.createElement shim.

01:40 GPHemsley : TabAtkins: That sounds familiar, but I can't say more than that.

01:40 TabAtkins : Hm, doesn't seem to help.

01:40 GPHemsley : :P

01:40 GPHemsley : hey hey!

01:40 TabAtkins : Not you. ^_^ Changing to .split("\r\n") doesn't help.

01:41 GPHemsley : oh, heh

01:52 TabAtkins : Okay, IE does newline normalization immediately, unless it has white-space:pre.

01:52 TabAtkins : So I got the newlines, now I'm just trying to figure out how to split by them...

02:06 dpranke : hixie / anyone : is there a reason why <select>, <input type="range">, and <input type="number"> don't support the "readonly" attribute?

02:06 Hixie : type=number does

02:07 Hixie : <select>, <input type=range>, <input type=radio>, <button>, etc, don't, because they're not text fields.

02:08 dpranke : you're right about number, sorry, I read too fast.

02:11 dpranke : what's the difference between disabled and readonly? It looks like disabled is readonly plus the control doesn't receive events? Why have readonly be able to receive events?

02:12 dpranke : or is there some other useful difference between disabled and readonly?

02:12 Hixie : the point of readonly is to allow users to copy the text from the control

02:12 TabAtkins : OMG, finally got IE to recognize a newline and split on it. The <csv> proposal marches on!

02:13 dpranke : ah. okay. does it make sense to add a clarifying note to that effect to the spec, or is this common knowledge?

02:13 Hixie : it's basic UI design :-)

02:15 dpranke : ok, thanks!

02:25 Hixie : oh dear, othermaciej is now being dragged into actually running these meetings

02:25 othermaciej : Hixie: I can live with running 1/3 of them or so

02:25 othermaciej : but yes, that was part of the cost, along with my soul

02:25 Hixie : nothing says we _must_ have telecons

02:26 othermaciej : I'm not going to propose canceling telecons, at least not right now

02:26 Hixie : in fact our charter, rather than requiring one a week, says that we're not allowed to have _more_ than one a week

02:26 Hixie : and says they should only be as needed

02:26 Hixie : it's your time :-)

02:26 Hixie : you spend it as you think best :-)

02:27 othermaciej : I think some of the people who attend get value out of it, and I'm willing to spend an hour of my time to give them that value

02:27 othermaciej : particularly if it helps some of those people also get involved by email

02:40 othermaciej : this is kind of sad, but I'm actually proud of the way I formatted the agenda email

02:42 rubys : http://www.w3.org/html/wg/tracker/actions/103?changelog

02:43 rubys : that action was assigned to Lachlan... but he never responded to emails checking for status.

02:43 rubys : 2009-01-29 17:54:35: Owner changed to 'Lachlan Hunt' [Dan Connolly]

02:43 rubys : 2009-05-07 16:10:37: Owner changed to 'Julian Reschke' [Dan Connolly]

02:44 rubys : http://www.w3.org/html/wg/tracker/issues/75 is already closed

02:56 othermaciej : rubys: I'll just make those adjustments in-flight rather than confusing people by changing the agenda

02:56 othermaciej : rubys: thanks!

06:35 cardona507 : is there a firefox extension similar to codeburner (the firebug plug-in) for html 5?

07:13 hsivonen : looks like we have a new case of adding attributes in JS to work around validators: http://stackoverflow.com/questions/1145880/can-autocapitalize-be-turned-off-with-javascript-in-mobile-safari

08:14 Lachy : dammit, why did apple feel the need to create more proprietary attributes for autocapitalize and autocorrect?!

08:26 Hixie : did printers get slower? i could have sworn that 20 years ago laserjets were printing at least as fast as the ones i'm seeing on HP's site today, and they warmed up faster at that.

08:27 hsivonen : Hixie: the high end probably hasn't gotten slower but a lower end got introduced and people started bying those

08:27 Hixie : where do you find the high end?

08:28 Hixie : i can't even find a single 802.11x printer on HP's site

08:28 hsivonen : Hixie: at Kinko's

08:28 Hixie : (laserjet printer, that is)

08:28 Hixie : heh

08:28 Hixie : i mean to buy :-)

08:29 hsivonen : hmm. I don't know if HP has exited the high end market and left it to Xerox and Canon

08:29 hsivonen : (I don't know where to buy those)

08:30 Hixie : "First color page in less than 16 seconds" $3199.99

08:30 Hixie : 16 seconds?!

08:30 Hixie : (HP Color LaserJet 5550dn Printer )

08:30 annevk2 : is that including boot-time or so?

08:30 annevk2 : and 3200 for a printer is a lot

08:31 Hixie : yes and yes

08:32 Hixie : for $3199.99 i expect my printer to be printing before i ask it to

08:34 Lachy : I gave up on owning a printer. My old, affordable, inkjet printers kept drying out cause I don't print frequently enough, and when I do, I just do it at the office

08:35 Hixie : yeah this isn't for me

08:35 Hixie : i haven't printed in like a decade

08:37 Hixie : what are the quality printer manufacturers these days? is HP still it?

08:37 Lachy : Hixie, is it intended for personal use by whoever you're getting it for, or office use?

08:37 Hixie : what's the difference?

08:38 Lachy : different feature requirements

08:38 annevk2 : Hixie, canon is also ok, not sure about laserjet

08:39 Lachy : like, an office printer would probably need an ethernet connection and be optimised for network setup, whereas personal use might not need that

08:40 annevk2 : for network setup you could always buy one of those apple thingies

08:40 annevk2 : with a usb-in

08:40 Hixie : the requirements are: high quality (laser jet), network printer using 802.11g or n, color, automatic double-sided printing, fast per page and fast from cold.

08:40 Hixie : and reliable.

08:40 Lachy : annevk2, the Airport Extreme or possibly Airport Express

08:41 Hixie : i'd rather it have built in networking

08:42 Hixie : HP doesn't have a printer that meets those requirements.

08:42 annevk2 : i've no experience at all with such beasts

08:42 Hixie : at least not according to their site.

08:42 Lachy : Hixie, so just printing, and no need for scanner/copier/fax functions?

08:42 annevk2 : my printer is in the closet and I take it out every couple of months to print some stupid expense report

08:42 Hixie : right

08:44 Lachy : Hixie, http://www.brother-usa.com/Printer/ModelDetail.aspx?ProductID=HL4070CDW

08:45 Hixie : looks interesting

08:45 Lachy : still 17 seconds for first print though

08:45 Lachy : for colour

08:45 Philip` : annevk2: You should use a commercial printing service for that, and then charge it on expenses

08:46 Hixie : yeah, still slow though

08:46 Hixie : both starting up and printing

08:46 Lachy : you consider 21ppm slow?

08:47 Lachy : that's 1 page every 3 seconds

08:48 Hixie : that's not faster than 20 years ago

08:48 annevk2 : Philip`, I'll ask for your advice next time :)

08:48 Hixie : so yes, i consider that slow

08:48 annevk2 : Philip`, meanwhile, charset stats? :)

08:48 Lachy : so how many pages per minute are you expecting?

08:48 annevk2 : Hixie, is this how you remember 20 years ago?

08:49 Hixie : annevk2: :-P

08:49 Philip` : Does anybody give double-sided printing speeds?

08:50 hsivonen : though when it exists, it can occasionally be used for printing train tickets

08:51 Hixie : Philip`: yeah, agreed

08:52 Lachy : http://www.office.xerox.com/printers/color-printers/phaser-7760/enus.html

08:52 Lachy : 35ppm for colour

08:52 Lachy : 45ppm for black

08:52 Lachy : "First-page-out time as fast as 9 seconds for color"

08:53 Lachy : Is that fast enough?

08:53 annevk2 : no wifi though

08:54 Lachy : yeah, but I was only searching for speed to see if it was possible to find anything at all fast enough for Hixie

08:54 Hixie : Lachy: hmm, that's better than i've seen so far

08:56 Lachy : Hixie, "With print speeds of up to 35 pages per minute for colour and 45 ppm for black and white and a first-page-out time of nine seconds for colour, the Phaser is considered one of the fastest printers in its class. " -- http://www.highbeam.com/doc/1G1-158653373.html

08:56 jgraham : Maybe there is a limit to how fast you can physically pull paper through a printer without bad side effects

08:57 jgraham : Anyway we had some Xerox phaser thing when I was doing my PhD and it seemed OK. It got used a lot and didn't break down often or anything

08:58 Hixie : cool

08:58 Hixie : good to know what the top of the line is, at least

08:58 jgraham : Colour reprodution was pretty poor. Dunno how fast it was or anything

08:58 hsivonen : are there color lasers whose color reproduction doesn't suck?

08:59 hsivonen : and ad bureaus and the like used slower and more expensive non-laser stuff

08:59 jgraham : In this case they were for printing science papers where relative colour tends to be more important than absolute colour

08:59 jgraham : s/they were/it was/

09:00 Hixie : yeah if you want to do something where colour really matters, you do ink jets, i think

09:00 Hixie : but then if i'm printing photos, i'm gonna be using a photo printing shop

09:00 hsivonen : are wax printers still on the market?

09:01 Hixie : is that "solid ink"?

09:01 jgraham : 9we were very licky to have a good colour printer at all; it was only because my group had seperate funding from the rest of the department. The department colour laserjet _really_ sucked)

09:02 hsivonen : Hixie: could be. I don't know.

09:02 jgraham : s/9/(/ and s/licky/lucky/

09:37 Mrmil : I hate IE

09:38 krijnh : High five!

09:39 Mrmil : Highway to Hell

10:08 hsivonen : I'm puzzled. With my latest HTML5 parser refactoring, jquery on http://webtrendmap.com/craigmod/161/ dies due to document.body being null

10:08 hsivonen : I wonder how that can happen

10:08 gsnedders : Because there's a bug.

10:08 Dashiva : hsivonen: Script in <head> before body is created?

10:09 hsivonen : Dashiva: jquery is in head, yes

10:09 hsivonen : Dashiva: but it works without the HTML5 parser refactoring

10:09 hsivonen : I wonder if I've accidentally caused onload or DOMContentLoaded to fire prematurely or something like that

10:12 gsnedders : I think you should never fire them.

10:13 jgraham : gsnedders: You miss Philip`?

10:14 gsnedders : Huh?

10:14 gsnedders : jgraham: No, it's my Swedish Mummy, Daddy, and Nanny I miss :P

10:14 jgraham : Well you have started filling in the "useless suggestions" role

10:14 gsnedders : I've always done that quite a bit, though.

10:16 hsivonen : sigh. I have a crash, too.

10:16 hsivonen : how hard can it be to move some code around?

10:17 gsnedders : Very.

10:17 hsivonen : refcounting is a house of cards

10:19 hsivonen : apparently moving code around is fairly easy, but splitting one refcounted object into many is a can of worms

10:19 annevk2 : xmlns issue still unadressed

10:20 annevk2 : are these people just incompetent at XML namespaces while happily using them?

10:20 gsnedders : annevk2: Then I guess we can free it

10:20 hsivonen : annevk2: in the RDFa draft?

10:20 annevk2 : yes

10:20 Philip` : gsnedders: You don't miss me? :-(

10:20 gsnedders : Philip`: Sorry, but jgraham just means so much more to me.

10:20 annevk2 : it has been raised like a bazillion times

10:21 annevk2 : my last email on this subject never had a reply I think

10:21 annevk2 : I think they just don't get it

10:21 Philip` : Their response to http://rdfa.info/wiki/Rdfa-in-html-issues#Script-based_modification_of_DOM (further down that page) seems to miss the issue with that example

10:21 annevk2 : I guess if you don't get XML namespaces they look all nice and dandy :)

10:21 hsivonen : annevk2: usually, if one "gets" namespaces, one wants to avoid them

10:21 annevk2 : right :)

10:26 annevk2 : ah, othermaciej raised the issue as well and manu didn't get it

10:27 annevk2 : he in fact simply snipped that comment from othermaciej's review

10:27 othermaciej : did he?

10:27 othermaciej : I thought he replied to that point (and I replied back)

10:28 othermaciej : it also looks like he pasted my comments about xmlns in text/html into the wiki

10:28 annevk2 : http://lists.w3.org/Archives/Public/public-html/2009Sep/0082.html in 4.3 your point about the DOM being different is no longer there

10:29 othermaciej : it is here though: http://rdfa.info/wiki/Html5-rdfa-wd-issues#4.2_Invalid_XMLLiteral_values

10:29 othermaciej : er

10:29 othermaciej : not in the 4.2 section

10:29 othermaciej : but you know what I mean

10:29 othermaciej : he only directly replied to a few of my points and recorded the rest on the wiki

10:29 annevk2 : i see

10:30 othermaciej : if you look at my further email, what XHTML+RDFa says about prefix binding is broken even for XML

10:30 annevk2 : his reply still shows misunderstanding of the issue but I guess they're working on that then...

10:31 annevk2 : i love it that such crap comes from the same people who claim XML is easy

10:33 hsivonen : othermaciej: I suspect RDFa is layered on top of XML 1.0--not on top of XML 1.0 + Namespaces

10:34 othermaciej : hsivonen: does that negate the error I pointed out in its rules about prefix bindings?

10:34 hsivonen : othermaciej: which is a pretty fundamental layering violation but gets an applause, because it looks like it's layered on top of Namespaces

10:34 othermaciej : (I don't think it does)

10:35 annevk2 : hsivonen, that cannot be for real

10:36 annevk2 : hsivonen, RDFa also references XMLNS

10:36 hsivonen : othermaciej: I think what you said stands

10:36 othermaciej : RDFa cites XMLNS, I am not sure what aspect makes it layered on top of XML 1.0

10:36 hsivonen : annevk2: I guess I need to look up an email from Mark Birbeck

10:37 othermaciej : the text I cited does seem to be treating XML namespace declarations as just ordinary attributes, but that seems to be just a symptom of its generally wrong use of terminology

10:37 Philip` : hsivonen: I thought the idea is it's layered on XHTML (1.1 or Modularization or something), which is layered on Namespaces to some extent

10:38 Philip` : (though not to a particularly well defined extent)

10:38 Philip` : http://www.w3.org/mid/4A291657.9020805@cam.ac.uk

10:38 hsivonen : othermaciej: I think it's a symptom of not layering the spec over Namespace but using xmlns as a trompe-lâÅil

10:39 othermaciej : it would probably make more sense for it to cite Namespaces in XML for namespace scoping instead of having its own rule for how to do it

10:39 Philip` : It might partly come from viewing XML documents as strings which follow a certain grammar, instead of as a DOM/Infoset-like tree

10:39 annevk2 : hsivonen, ah yeah, it does do that

10:40 othermaciej : although that would mean more work to define how it works in text/html

10:40 Philip` : (The XML spec itself seems to be designed from the string viewpoint)

10:41 annevk2 : Philip`, which is weird at least for <?xml encoding="..."

10:41 hsivonen : annevk2: http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2009Feb/0109.html

10:41 othermaciej : anyway, I'm glad Manu put forth a technical proposal this time instead of a procedural one, even if it seems like it needs work

10:42 hsivonen : gotta go to lunch

10:43 hsivonen : just got Snow Leopard in the mail. I wonder if I should wait until 10.6.1 before I install...

10:44 othermaciej : it seems like they could just use the "Namespaces in XML" definition of namespace prefixes in scope, even if they use the prefixes for non-namespace reasons

10:44 othermaciej : 10.6.0 is better than the average 10.n.0

10:45 othermaciej : it seems like HTML+RDFa needs to make @rev and various other attributes conforming

10:46 Philip` : It probably needs to do something similar to what's in http://philip.html5.org/docs/rdfa/

10:47 Philip` : (except probably without copying-and-pasting the RDFa-in-XHTML processing model, because duplication is bad)

10:48 Philip` : (and the issues with e.g. precisely defined attribute parsing apply to XHTML as much as to HTML)

10:52 othermaciej : Philip`: that draft looks much more precise, despite its unfinished state

10:52 othermaciej : Philip`: might be good to point out that draft on the list

10:53 Philip` : I wonder if I should try rewriting it to remove the duplication of the processing model, since that seemed to be the main objection to it

10:59 othermaciej : Philip`: if you have the time to pursue it, then I would encourage you to pursue that and post the draft

10:59 alkarin : greetings everybody