HTML Accessibility Task Force Teleconference -- 05 May 2011

<trackbot> Date: 05 May 2011

<MikeSmith> Zaki, this will be WAI_PFWG(HTML TF)

<MikeSmith> scribe: Stevef

Media subteam report

MS: stuffed up the passcode, remember to chnage for next time

JF doing mop up on issue 152 multi track API, essentially resolved, discussion about @king attributes, spec shows 5, but wiki page with other values, at least 3 more and maybe some others

JF: require captions and subtitles, maybe commentary still being discussed, bug filed against this item.
... as far as spec, john is unsure

MS: is there anything that chairs are waiting on?

JF: not that i know
... 2-3 bugs resolved
... if we can get @kind values resolved amicaably before last call should be good

Text subteam

MS: any questions, no move on

<MikeSmith> scribe: oedipus

SF: met monday -- discussing validation issues with text alternatives with judy and on list using the [text] subject line tag -- reopening issues -- gathering information on validation issue decisions
... ones most worked on at the moment: meta genarator and @title -- also work on role=presentation (would like to revisit)
... talk about FIGCAPTION -- trying to reason it out rather than reopen at this time -- agree on for past 2 years -- judy and geoff (WBGH) working on FIGCAPTION and longdesc

<JF> +q

SF: moved formal objection for @alt and @title -- waiting on response

<Stevef> oedipus: i can take it back now

JF: also work on meta genarator decision -- fairly verbose discussion currently -- whittle down to 2 parts to present all evidence and in more sumarized fashion

SF: issue of meta genarator tag a generic thing that was used in past in a LOT of published content -- didn't do to placate validator

JF: 10 years of legacy content with meta generator strings -- must be considered
... turning meta genarator string into "magic token" is problem

SF: meta generator exceptions and error rules in HTML5 investigated?
... MikeSmith know of any?
... have meta generator that is exception to use of @alt -- shuts up validator -- any other errors that tell validator to ignore specific syntax
... can imagine if press rules, will be other requests for rules -- why a single exception for @alt

MS: worded in spec not to suppress error, but says alt required unsless there is a meta tag with name="generator" or a @title

<JF> +q

SF: flagging a conformance error

MS: could perceive that way, but are other things in same class -- not unique in that regard
... behavior class performance criteria for others as for generator -

SF: examples?

MS: select element has required attribute and no multiple attr and size of 1 (have to compute, not just value of attr) if all are case, must have option attr that either has a value attr or doesn't have any text content
... other cases: complicated assertions in spec

SF: doesn't have anything to do with setting flag via meta generator
... is there a flag one can set for changing error to non-error?
... happy to discuss offline with MS

MS: happy to discuss offline -- not the only case

JF: interested in question by SF -- what we have is a meta string -- example you gave VERY tightly focused -- generator string in hand of document and effects ENTIRE document -- for CMS systems when in template can negatively affect entire site
... anything in spec that changes performance criteria for entire document

MS: this case is an extreme one, but not the only one
... presence of single element with certain attr in document affects every image in document
... talked with SF about this offline, but think in terms of validator behavior, don't think many users will expect when put in meta generator element in document that that means they are opting out of checking for @alt on img -- something consciously doing by adding generator? -- that to me is strong argument -- people want conformance checker to alert them to things that might otherwise miss

JF: CMS supports 150 content authors who don't see code ever -- just add content to template -- don't know what meta generator string is -- 3rd party conformance tools have to deal with fact that @alt not needed if meta generator inserted -- allowed because of meta generator string -- assert most authors willl have NO idea they are creating a problem because generator string says confromant

<Leonie_Watson> +1 to JF

MS: conformance checker hat on this doesn't seem to be optimal behaviour -- should persue, but not now -- supportive of trying to see whether this is what other people working on ER tools think is acceptable behavior
... talk with henri about it on list good approach

JF: ping RichS from IBM

SF: OpenAjax Alliance

MS: FIGCAPTION -- state of things: had agreement with TF just a few weeks ago --changed -- don't know how or why

i don't remember consensus on FIGCAPTION

<JF> JF wanted to include the wiki page for the @kind values for the multi-track API discussion: http://www.w3.org/WAI/PF/HTML/wiki/Track_Kinds

SF: misunderstanding in regards what WAI-CG agreed to over 2 years ago -- spec changed
... FIGCAPTION supposed to be a caption, not a text alternative -- use cases that are WCAG conformant that a CAPTION is sufficient textual alternative

MS: think we need to not operate under assumption that HTML WG chairs are going to delay LC for Text Subteam to complete work and have reviewed by TF -- not prudent assumption -- chairs plan to follow tight schedule for September 2011 publication -
... need to bring up for discussion in TF

JF: suggest MS discuss this with Judy Brewer

MS: anything else on text subteam?

<Stevef> SF: FIGCAPTION supposed to be a caption, not a text alternative -- use cases that are WCAG conformant that a CAPTION is sufficient textual alternative

SF: if no @alt, but @title, role still presentation for element/attr and will NOT be in A11y API or accessible to ATs
... allowing that to occur means that it will not work for AT users -- role=presentation means content not available

<MikeSmith> scribe: Leonie_Watson

JF: What does title do for us from an accessibility perspective?
... It seems to me that until the browsers do something useful with title...

SF: There is an implementation issue here that needs to be resolved. The conformance rule is contradictory.
... Either the spec has to change, or the implementation has to change.

<oedipus> GJR: @title has NO place in graphical/visual rendering space -- if no @alt, no text inline -- only @title available onMouseOver (device dependent and impermenant [only rendered for a short time and then disappears

JF: We have incomplete support for the title attrib, in that it only produces a tooltip in most browsers. For users with cognitive issues, you could argue this is helpful.

<Zakim> oedipus, you wanted to say working on @summary for table and issue of hidden versus discoverable metadata

JF: It's not available to all users though, only some users.

MS: It sounds as though there's still some work to be done, but that there are people working on it.

Canvas sub team

GJR: We've been working via email. We're waiting for the results of Rich's latest exchange with Ian. We're waiting for the inter group dynamic to work itself out. I think there is movement there.
... We're also aggresively pursuing testing canvas with ATs and browsers. We've gotten very good feedback/participation from AT providers.

MS: I don't think there's anything the TF needs to step in on there.

<MikeSmith> scribe: oedipus

ARIA Mapping/Bug Triage

MS: pertient people not here --

JF: bug triage team needs to take a new pass after last few weeks of action

Review of Open Action Items

<MikeSmith> http://www.w3.org/WAI/PF/HTML/track/actions/open

MS: nothing with high priority -- a bunch due recently for text alternatives sub-team -- doesn't need further review
... don't need to reveiw as group

Scribe for Next Week

MS: volunteer for 12 May 2011?
... Janina will chair next week
... other business?

[ADJOURNED]

- DRAFT -

HTML Accessibility Task Force Teleconference

05 May 2011

Attendees

Contents