PF HTML5 Issues Caucus -- 13 Feb 2009

<janina> argh!

<Stevef> me/ will join call soon

<oedipus> minutes from the latest HTML WG teleconference:

<oedipus> http://www.w3.org/2009/02/12-html-wg-minutes.html

<oedipus> unofficial detailed report on @summary, @headers and @alt:

<oedipus> http://lists.w3.org/Archives/Public/public-html/2009Feb/0285.html

HTML WG Telecon Review

<oedipus> minutes from the latest HTML WG teleconference:

<oedipus> http://www.w3.org/2009/02/12-html-wg-minutes.html

GR: Long discussion about summary
... Chris Wilson stated, as far as he's concerned 2 conditions for it to be restored
... First support - Second potential for misuse. Lack of examples in wild
... Pointed out it was added to HTML4 for a specific purpose and well supported by AT
... Gregory asked if that sufficed as support
... Chris said he agreed it was supported
... But there was an anti-summary argument
... Which was if supporting does no harm, keep it
... Explained that those who use summary tend not to misuse it. Thinks it's a fear of invisible meta data
... Those who are against summary should document on Wiki why it shouldn't be on there
... This page has been in existence in Wiki for some time, and always cited. Anyone against summary should put their argument in the Wiki
... A few people, James Graham, who is just against summary being back in. If it's good enough for VI users, it should be good enough for everyone

<Laura> Dan has previously commended that Wiki page.

GR: Pointed out it was added for a specific purpose.
... Waiting a week for people to articulate reasons for not including summary
... Other than summary, which has been proved wrong

SF: Asked Marco - positive feedback, and he uses it
... Has been researching summary
... Found some good uses case where it's used useefully
... The only study done was by Philip, who pointed out it wasn't a thorough study (scraped a few pages from DMOZ)
... Found contrary examples in government sites that shows it being used properly (1 in 10 had a summary)

<oedipus> Scribe: Gez_Lemon

<oedipus> ScribeNick: Gez

SF: Stuff quoted against summary is not that useful, and there's good data showing it is used and used properly
... MArco has some good arguments for summary, and Steve will pass a link later
... Useful to hear opinions from users, rather than advocates

<Laura> Steve's IRC discussion: http://krijnhoetmer.nl/irc-logs/html-wg/20090213#l-92

<Stevef> me/ tahnks laura

JS: Understand another working draft released?
... Discussion of table headers?

GR: At that point, things got confused
... Dan acted as chair
... Inserting new topic lines, forcing us to move on before discussions were finished
... Tried to get some resolution
... We ended up with "that's fine", "That's nice", but an unbalanced argument
... Thought from the review of the editor's draft, we agreed on wording for headers

JS: Agreed
... Need to be more explicit with what we think is good

GR: We have told them on numerous occasions that a th or td can accept a header element
... Not sure what else they want to close the issue
... We got the change effected
... Still considered by the chairs an open issue
... Confused with what precisely the HTMLWG admin point of view is

<Laura> James Graham is doing headers testing along with Chaals

<Laura> http://lists.w3.org/Archives/Public/public-html/2009Feb/0302.html

JS: Alt come up?

GR: No
... Full half hour on summary
... Pretty ugly at times

SF: Doesn't understand how the Wiki is not considered balance

GR: Replied that these pages had been out there for over a year and a half, and cited every time the issue is discussed
... If people take the time to counter why it shouldn't be retained, they're the ones who dropped the ball

JS: Maybe they don't have good reasons

GR: Frustrating because it feels like there's a 2-tier system about who gets listened to

SF: Agrees, doesn't know what can be done about it

GR: Difficult to get things done when the knee-jerk reaction is, "What does James Graham think about it"
... He's one person, not an accessibility guru

JS: Do we want to hold up working draft?

SF: It's out already

GR: One of the first things they said was they already sent the request and whatever we said would not effect the next draft

JS: The one before last call?

GR: No
... Big difference is this draft will swallow web forms 2.0

<Joshue> I think everyone was a little surprised about the draft, everyone who hadn't seen the W3C homepage for a bit anyway

<Laura> HTML 5, Differences from HTML 4 Published Yesterday

<Laura> http://lists.w3.org/Archives/Public/public-html-wg-announce/2009JanMar/0028.html

GR: There is an action item for chairs to coordinate with XHTML WG on syntax and vocabularies
... Good ig Janina gets involved

JS: Okay, send me a pointer

GR: Okay

JS: Do you need an action?

GR: No

<Joshue> +q

JS: Anything else to be said by way of reporting?

<Laura> The whole point of Josh's action 66 was to update the Wiki to be unbiased. That action was completed, approved and closed.

<Laura> http://www.w3.org/html/wg/tracker/actions/66?changelog

JOC: The meeting was quite productive. One of the most fluid in a while
... Had a sense of some forward motion
... If someone has a reason why something should be dropped, they need to make a good case for it

<oedipus> the problem is the "process" is bass-akwards

SF: From reading the IRC, it seemed it kept swinging from side to side
... People need to make objects, then Sam said people need to make their case on Wiki

GR: Dan kept jumping from topic to topic
... Frustrating, as never got to resolution
... The support is there to meet the requirement, but we're still waiting for the contract

<Joshue> +q

<Joshue> -q

JS: Need to send a note. Should we go to formal vote inside formal Wednesday meeting?

JOC: Don't think so. Shouldn't have everyone throwing their 2 cents. That's my 2 cents

JS: Reiterating with additional arguments is one way to go about it. Take advantage of Steve's research

SF: Will blog about it
... Critical assessment of previous data, and my findings to give a more thorough and balanced view
... That with PF support and Marco will have a positive outcome

JS: Shall I wait until that post?

SF: That would be a good idea

<oedipus> suggest JS use http://lists.w3.org/Archives/Public/public-html/2009Feb/0285.html as model

SF: Should have that done by Wednesday
... We can discuss it then
... If there are counter-arguments, it will help to push it over the edge

<Joshue> +q

<oedipus> me oe

GR: Given an action item, and the one topic reply from Simon Pieters was, "You think that's enough to stop it being released?" Next sentence, "It's already published"
... Resulted in hot debate
... Sam Ruby said it would be handled better in future

JS: Also surprised by Simon's note

JOC: Agree

JS: Is it important to worry about this draft? Maybe just reiterate our case

<oedipus> Simon Pieters: "Wait, are you saying that the PFWG formally objects to the lack of summary="" in HTML5 as to prevent publishing another Working Draft?"

<oedipus> http://lists.w3.org/Archives/Public/public-html/2009Feb/0298.html

JS: position has always been keep @summary in

<oedipus> GJR worried about being painted into a corner as knee-jerk reactionaries whose only recourse is to formally object

<Joshue> -q

JOC: Has article already written
... Steve can use if he wants

SF: Article on using the summary attribute. Support, use, and data about usage

JS: What do we do about table headers?
... We should say what we like, to ensure it doesn't get changed
... What still needs to be tested
... Rather have that before sending note to chair

SF: Two different things. The current AT support is inbuilt in AT. Moving to AT, won't be any different
... They have this algorithm that will allow the browser to compute table headers
... The algorithm has problems.

JOC: Basically, it's a heuristic

<oedipus> GJS: table model is what to grok

JS: Need a URI to make sure we state what we're happy with

GR: Working on it now
... We should concentrate on the table model
... We have to use single page version if we want W3C branding

<oedipus> http://dev.w3.org/html5/spec/Overview.html#tabular-data

<oedipus> 4.9.13 "Table Processing Model" - http://dev.w3.org/html5/spec/Overview.html#processing-model-0

GR: Table processing model we need to study
... It's not the algorithm, but how we parse and form relationships between data

<oedipus> http://dev.w3.org/html5/spec/Overview.html#header-and-data-cell-semantics - 4.9.13.2

GR: Header and data flow semantics
... Comfortable leaving the analysis of the algorithm to someone like Chaals
... Most effective use of our time

JS: Will draft message referencing that section
... We'll say great job - algorithm another topic

@alt in HTML5 caucus

JS: Apologies for missing meeting, and apologies for lack of warning

<oedipus> http://www.w3.org/2009/02/12-cg-minutes.html

SF: Latest working draft is multipage version

GR: XHTML2 printed is 300 pages. HTML5 is 500 pages printed (approx)

JS: Quirkiness in call. It's back to Wednesday
... But at 3pm US Eastern

GR: IF we get advance notice, I can add to caucus page

JS: Commit to getting notice out in advance

<oedipus> http://esw.w3.org/topic/PF/XTech/HTML5/Caucus

JS: Will go through straw polls we've had in the last few calls, and put them in a single message for clarity
... Will be useful to commit and go to resolution status
... Having read minutes from yesterday, didn't find anything I would object to
... Yet.
... Anything else about alt?

<Joshue> JOC: the only thing I was a little unsure of was the resolution on @longdesc

SF: aria-describedby to replace longdesc?

JOC: Yeah

JS: Strawpoll or proposed

JOC: no semantic hooks for abbreviations

JS: What will that be in terms of backwards compatibility?

JOC: No idea

SF: Guess by the time it's implemented it, if AT is supporting it, it won't be problematic about backwards compatibility

<oedipus> http://esw.w3.org/topic/PF/XTech/HTML5/MultimediaElements

SF: Not saying it should be got rid of, but good evidence that it's not used that well

GR: Has a proposed roadmap
... Alt would map to legend
... longdesc would map to content in open/close figure

<Joshue> JOC: I think the implementation of @longdesc is poor. I wonder if the model is even correct in the first place and maybe using aria-describedby will be an improvement.

We need a child element for figure

GR: Need new element called desc

<oedipus> rssagent, make minutes

JS: Wrapping up - backward compatibility is elephant in room. What else is?

<Joshue> JOC: Its good that Jan does bring up backwards compatibility

<Joshue> Gez: Just to mention the @longdesc. people are confused because if used correctly it is on a different page. Aria-describedby needs to be on the same page.

<oedipus> GL: quick mention of the longdesc attribute which might be superceded; longdesc used correctly - replace by aria-describedby, description needs to be on same page as what it describes, not usually the case in real world; like GJR's suggeston]

<Joshue> Gez: I am therefore not convinced that aria-describedby is a suitable replacement for @longdesc but I like GJRs idea.

<Zakim> oedipus, you wanted to ask if support for my contention that CAPTION is a terse descriptor akin to a header or legend, and @summary is a more verbose descriptor, in the vein of

SF: we need further discussion

GR: I was jumped on for explaining difference between caption and summary
... Caption concise, summary more verbose
... Fortunately, Steve brought up WCAG technique to illustrate how it works
... The caption is the header for the table

JS: Should we be looking at web forms now it's part of HTML5?

GR: Yes
... In XHTML WG, they're considering which version of web forms (sorry, got lost there)

JS: Put on agenda for Friday?

GR: Take care of what we're doing now before tackling this
... Will keep group informed of what XHTML intend to do regarding forms
... Still have issue that we asked Chris Wilson at face to face whether forms would be hard and fast, and it looks like executive decision has been made that forms will be described in web forms 2.0

JS: Useful conversation
... Anything else?

JOC: One quick thing. Coordination meeting on Wednesday at 8 UTC?

JS: Yes, and I'll have agenda out 24 hours before

- DRAFT -

PF HTML5 Issues Caucus

13 Feb 2009

Attendees

Contents

HTML WG Telecon Review

@alt in HTML5 caucus

Summary of Action Items

Scribe.perl diagnostic output