HTML Issue Tracking Teleconference

03 Apr 2008

See also: IRC log


MikeSmith, DanC, Julian, Laura, JoshueO, Gregory_Rosmaita, PatrickIon, Robert_Gonia, DougSchepers, ChrisWilson, anneVK


<DanC> action-48?

<trackbot-ng> ACTION-48 -- Dan Connolly to start mailing list on phrase level semantic elements -- due 2008-03-13 -- PENDINGREVIEW

<trackbot-ng> http://www.w3.org/html/wg/tracker/actions/48

<DanC> DanC: anybody mind if I withdraw?

<DanC> close action-48

<trackbot-ng> ACTION-48 Start mailing list on phrase level semantic elements closed

<DanC> next meeting should be 4p PT Thu 10 Apr, ChrisW to chair

Convene, review agenda, plan next meeting

ISSUE-14 aria-role

<Steve_f> hi Dan, i am unable to attend as i have to go and pick up my daughter from nursery at this time, any chance of making it later or earlier next week +/- 1hour?

<DanC> you can ask chrisw, Steve

<DanC> http://www.w3.org/html/wg/tracker/issues/14

<oedipus> GJR: PF convened a special meeting at a special time to accommodate participation by HTML WG members in discussion of aria-role in HTML5 on 25 march 2008 but no one save PF people turned up -- SimonP valiantly tried to join the call, but could not dial in for some technological reason and was there on IRC


<trackbot-ng> ISSUE-14 -- Integration of WAI-ARIA roles into HTML5 -- OPEN

<trackbot-ng> http://www.w3.org/html/wg/tracker/issues/14

<Steve_f> ok no problem

<DanC> pointer to records of that meeting, pls?

<DanC> or meeting call?

oedipus: we convened a special meeting to help others to attend, but nobody other than PF people managed to join
... we have meeting minutes but they are member-only

<DanC> ACTION-8?

<trackbot-ng> ACTION-8 -- Michael Cooper to discuss with PFWG role attribute vs aria attribute -- due 2008-02-21 -- OPEN

<trackbot-ng> http://www.w3.org/html/wg/tracker/actions/8

oedipus: we have not been able to get a critical mass of people from outside PF to join in the discussion

<DanC> close action-8

<trackbot-ng> ACTION-8 Discuss with PFWG role attribute vs aria attribute closed

<DanC> close ACTION-9

<trackbot-ng> ACTION-9 Discuss UML tools with DanC closed

<oedipus> member-confidential record: http://www.w3.org/2008/03/25-pf-minutes.html

<DanC> ACTION-23?

<trackbot-ng> ACTION-23 -- Gregory Rosmaita to coordinate tests using ARIA -- due 2008-02-21 -- OPEN

<trackbot-ng> http://www.w3.org/html/wg/tracker/actions/23

<Steve_f> Dan - can you note that action 54 is in progress still, we are waiting on response from PF WG on use of normative statements.

<DanC> action-54?

<trackbot-ng> ACTION-54 -- Gregory Rosmaita to work with SteveF draft text for HTML 5 spec to require producers/authors to include @alt on img elements -- due 2008-04-10 -- OPEN

<trackbot-ng> http://www.w3.org/html/wg/tracker/actions/54

<DanC> Steve_f, we're not there just now

oedipus: we don't have the level of collaboration with the HTML WG that we would like

<DanC> re tests... al gilman gave me http://lists.w3.org/Archives/Public/www-tag/2008Mar/0069.html

<Steve_f> apologies

<oedipus> Paciello Groups' ARIA tests: http://www.paciellogroup.com/blog/?p=53

oedipus - we will neither add nor [remove] any roles at this point

oedipus - navigation and landmark roles [two types]

<DanC> DanC: do the Paciello Groups' ARIA tests enumerate the roles?

scribe: we doing a comprehensive check to see that all the states are explicitly spelled out

DanC: It would help me to see some tests

<DanC> I read http://hsivonen.iki.fi/aria-html5-bis/

<DanC> Text last updated: 2008-03-31 by Henri Sivonen

<Cathead> Cathead is joshue

<oedipus> Public PF comments list: http://lists.w3.org/Archives/Public/public-pfwg-comments/2008JanMar/

DanC: has anybody looked at Henri's posting?
... oedipus, can you look at Henri's doc now?

PIon: appears to have a list of roles that are not supported

<anne> Henri's ARIA post made sense to me. I haven't been deeply involved in ARIA semantics though.

<anne> I was mostly concerned with ARIA syntax and that seems to be a resolved thingie now.

<DanC> stuff from cooper to the TAG http://lists.w3.org/Archives/Public/www-tag/2008Apr/0006.html -> http://www.w3.org/2008/03/aria-implementation

DanC: MichaelCooper sent that yesterday
... Noah Mendelsohn replied
... I'm trying to find out how close we are to closing this issue
... my goal it to determine what test cases we have

oedipus: our original proposal was that we preferred native features in HTML over adding them using ARIA
... I think there is some basic understanding of what we're trying to do with ARIA
... [mention of example of math]
... ARIA 1.0 is about repairing what we have today
... but the eventual migration is that we hope that in future, many of these ARIA roles and landmarks will become a native part of the HTML markup language

shepazu: I think we can cover background on ARIA outside of this telcon
... we can have a separate call about this perhaps

oedipus: [mentions posting from jgraham as indicating that ARIA roadmap/plans might not be well understood by all]
... there are tests on the ESW wiki

shepazu: this stuff is actually working in Firefox..

oedipus: we can't provide tests for HTML5 until we have agreement with the HTML WG about how to integrate it

<PIon> Is there an example of the use of a "math" role in the ARIA specification? Does it specify the informative additional markup to be used?

oedipus: nobody has come to the PF working group with their tests

shepazu: the tests can be proposal-level stuff
... the group can then take a look at those
... you don't have to spec out everything first

DanC: that would help me out a lot

oedipus: that would not be satisfactory

<zcorpan> i'm planning to write some aria-in-html5 tests/demos soonish, btw

DanC: a test is a file written by somebody who as a notion of what the expected results are

<anne> Actually, I often start out by writing a test where I'm not quite sure what the expected results are...

oedipus: we want to take what we've learned and bring it to the HTML5 WG and have it integrated into HTML5

<zcorpan> a test without a pass condition is a demo

<zcorpan> demos are still good

<DanC> right, I like to have pass conditions too

oedipus: ARIA is on LC-track

<anne> (Well, eventually I know and they'll have a pass condition, but not initially.)

oedipus: we are looking for more input from zcorpan, hsivonen, others .. need to them to come to the PF group to speak with us
... meeting in the public space
... we did something similar with math
... we have e-mail discussions but communication breaks down
... [having a telcon discussion would help with that issue]
... we want things like ARIA overlays to not be necessary in HTML5 [UAs]
... targets for current ARIA is legacy browsers
... for 1.0 we are restricted by current implementations

<DanC> (is the "no UA changes required" constraint on ARIA 1.0 documented?)

<oedipus> GJR: current and legacy browsers

shepazu: so what you want to discuss is ARIA 1.1 + HTML5

<Cathead> The way current AT works is ARIA not more dependent on how the browser deals with ARIA code rather than the AT itself?

DanC: so, OK, I'm hearing that the communication is not going well

<DanC> close ACTION-11

<DanC> close ACTION-11

<trackbot-ng> ACTION-11 Show how <div aria="something"> works with URI based extensibility closed

ChrisWilson: I'll be on vacation next week and the following week

ISSUE-37 html-svg-mathml

<DanC> . http://www.w3.org/html/wg/tracker/issues/open?sort=product

DanC: issue-37.. SVG/MathML should be a requirements issue, not a spec issue


<trackbot-ng> ACTION-29 -- Dan Connolly to follow up on the idea of a free-software-compatible license for a note on HTML authoring -- due 2008-05-01 -- OPEN

<trackbot-ng> http://www.w3.org/html/wg/tracker/actions/29

ChrisWilson: I think integrating other schema into HTML is a good thing, if for no other reason that it allows us to keep our own schema more targeted
... we have had a mechanism in IE for doing something like this for some time now

shepazu: curious about IE's xml element

<DanC> (pointer to ext element proposal, please)

shepazu: my proposal is for a "point of extensibility"

<shepazu> http://wiki.whatwg.org/wiki/Extensions#Proposal_2:_Extensibility_Element

shepazu: I think the HTML WG should not attempt to change the semantics or syntax of other languages

PIon: I'm here because I think the matter is a spec issue
... math should be integrated for sure
... the Math WG is very willing to discuss this
... and the last week or show has shown some very useful discussion about this
... I think you have a good chance by working with some of the people who've been trying to get this done now for years
... and we are certainly willing to discuss this

Julian: I agree the extensibility question is general is one of the must interesting things that HTML5 should try to solve

<Zakim> DanC, you wanted to note a possible broader issue that svg/mathml and to ask patrick about how long it's likely to take to settle on a good design, and explore more refined

DanC: I sense a lack of consensus about whether there is a need for a general extensibility in HTML

shepazu: my proposal is not for a general
... completely "generic" extensibility mechanism

<PIon> q If there's a more general solution then it should encompass SVG and MathML as candidates.

DanC: the design space around this looks big
... PIon, how long should this discussion take?

PIon: you can't expect consensus to emerge in 3 weeks, but perhaps in 3 months

<ChrisWilson> My point is that interoperability is important 1

<ChrisWilson> whoops

PIon: from our side, we are working on MathML 3.0, and we are still willing to make changes
... so it desirable to try to get this nailed down pretty early, while both groups are still in a position to make changes

<Zakim> ChrisWilson, you wanted to respond to shepazu's xml element comment

<anne> Do we have 90min telcons? If so, I'll dial in

ChrisWilson: shepazu, you were asking about the idea of how the extensibility point idea would [work with/map to] the IE xml element

<anne> (the agenda doesn't say)

anne - we will be on for another 30 minutes, yeah

ChrisWilson: it ends up in _a_ tree, but not _the_ tree
... there are challenges about the programming model
... stylesheets don't work across the boundary
... for example

<Cathead> For the record issue 32 (missing alt) is currently being worked on my myself, Laura C, Steve F and Gez Lemon.

ChrisWilson: you need to remember that the parser does some specific things today
... e.g. our parser fires off "speculative image downloads"
... perception of performance would suffer

<DanC> ACTION: ChrisWilson respond to extensibility discussion [recorded in http://www.w3.org/2008/04/03-html-wg-minutes.html#action01]

<trackbot-ng> Created ACTION-57 - Respond to extensibility discussion [on Chris Wilson - due 2008-04-10].

<DanC> DS: among those participating in the thread, only hickson seems opposed to using one of <ext-like> or <svg>/<mathml>.

<DanC> Anne: how many browser builders have participated?

<DanC> DS: mjs

anne: [notes that some important stakeholders have not contributed to the discussion thus far, so perhaps it is premature to draw conclusions from any perceived current consensus]

<anne> I can't find Maciej's e-mail in the list

<DanC> shepazu, can you help anne find it?

<anne> http://lists.w3.org/Archives/Public/public-html/2008Apr/

Julian: lots of people said they don't want the HTML parser to get further complicated by special-casing for large numbers of new elements

<shepazu> http://krijnhoetmer.nl/irc-logs/whatwg/20080401#l-441

<shepazu> it wasn't on the list, it was on IRC

<Laura> Cathead, Missing alt is action 54 http://www.w3.org/html/wg/tracker/actions/54, issue 31 http://www.w3.org/html/wg/tracker/issues/31.

DanC: meeting adjourned