HTML-A11Y telecon -- 02 Dec 2010

<silvia> MikeSmith: the examples from my book are not up at http://html5videoguide.net/ <- might be useful for the video element test suite for html5

<MikeSmith> you mean "now up"?

<silvia> yes :)

<janina> agenda: this

<richardschwerdtfe> scribe: Rich

Action Items

Action 85

<richardschwerdtfe> janina: still not done. give me another week

<MichaelC> action-85 due 9 Dec

<trackbot> ACTION-85 Find someone to create a change proposal for ISSUE-80 http://www.w3.org/html/wg/tracker/issues/80 due date now 9 Dec

<richardschwerdtfe> michael: due 12/9

86

<richardschwerdtfe> change proposal due tomorrow

<MichaelC> action-87 due 5 Jan 2011

<trackbot> ACTION-87 Create change proposal for ISSUE-134 http://www.w3.org/html/wg/tracker/issues/134 due date now 5 Jan 2011

87

<richardschwerdtfe> michael: now dow january 5

Fallback mechanisms

<oedipus> http://lists.w3.org/Archives/Public/public-html-a11y/2010Nov/0278.html

<inserted> scribenick: richardschwerdtfe

Martin: there are a lot of elements that have a lot of fallback mechanism, like alternative text on images
... we have the same for MathML
... we have some that have repair fallback

<oedipus> MK's list of element types:

<oedipus> 1. elements like <img> and <math> that have some kind of @alt on the top element,

<oedipus> 2. elements like <video>, <audio>, <svg>, <source>, and <map> where there is some kind of @alt on a child element, but no @summary on the container,

<oedipus> 3. elements like <object>, <video>, <audio>, and <canvas> that have repair fallback, but no accessible fallback like exposure to AT (yet).

<oedipus> 4. elements like <iframe> and <embed> that do not have any fallback content at all.

Martin: For example the object element can have an image that can be a repair
... we have the canvas element which can have a shadow DOM
... We need to define how the accessibility of fallback content can be handled consistently
... Image has no child elements. So, it has fallback content attached to it. However video does not
... the track element does
... we recommend to normalize and harmonize the repair fallback mechanisms for embedded fallback content

<oedipus> re: iframe, there is a bug filed on behalf of the UAWG on requiring @title for FRAME and IFRAME

<oedipus> http://www.w3.org/Bugs/Public/show_bug.cgi?id=10709

janina: are we sure that for all of these that we need each of those boxes checked off?
... the alternative content for video may be more than one alternative. we may not have them all.

janina; for example, we could have sign language in specific languages

janina: people may have premixed these things and then we have a labelling issue
... here is a video with "baked in" audio descriptions

martin: that is true. We also track an audio and video and describe the types of tracks they are
... it would be nice to know what the topic is of an MP3 file is. We don't know what a track is about.

janina: I agree. Knowing what it is about is an issue for everyone

Michael: I did not look at tracks. there may be a lack of understanding on my part.
... The use case table does not define all the text alternatives an object can have. The situational dependency is outside the scope of this discussion
... I was trying to do an exercise of creating a restricted set of fallbacks and define some patterns

john: I have 2 questions largely related to video although probably pertain to audio too
... there has been some discussion as to whether the poster frame should have a text description

<kliehm> ISSUE-142?

<trackbot> ISSUE-142 does not exist

<kliehm> http://www.w3.org/html/wg/tracker/issues/142

John: should the poster be an alt attribute of video. Am I out in left field?

Michael: the alt attribute was not in the version of spec. I reviewed

<JF> JF note to self to ensure @alt is part of <video>

<kliehm> http://www.w3.org/Bugs/Public/show_bug.cgi?id=10642

Michael: This exercise (HTML 5 mapping) as of the editors draft was only an inventoried snapshot
... The end goal was was to use the existing support table and how we back fill it

John: does the alt attribute for the video also serve as the alt attribute of poster?

<oedipus> GJR treat separately

Michael: I would propose that they treated separately

gregory: separately

Michael: The alt attribute for short text alternatives should be a label.

janina: this is more than just accessibility

eric: I am going to agree that I don't believe it is necessary to have a short text description of the video and the poster.

john: very often the image has nothing to do with the video

<oedipus> poster info may have name of actors, director, producer, writer, etc.

<oedipus> video @alt is "name of video"

John: for example, many of the videos we have on campus read say "Stanford university. this is closed captioned"
... I believe we have two unique and discreet assets
... we do not need a short text alternative that describes both if both are being declared by the author
... does anyone disagree that the poster should have alt text

gregory: what you are going to get for the alt for the video but for the poster you would have a lot more text

eric: whether the video has a poster attribute or not it needs to be described.
... I calling it as a description of the poster is a mistake

john: we can use aria-describedby to describe the video
... I am most concerned that the author creates a graphic alternative that the text equivalent be accurate

eric: what should I use when the poster is the first frame of the video?

john: the fallback should be such that it becomes a child of the video. It should be declared as a unique asset

<kliehm> @label: Casablanca. @poster: Black-and-white film screenshot of a man and woman as seen from the shoulders up. The two are close to each other as if about to kiss.

eric: why should the markup be different when it is used as a video vs. a poster?
... I am concerned that we will have something more complicated that it needs to be

john: if the author does not specify an image file that the alternative text is not related to the video itself?

<inserted> scribenick: oedipus

RS: what happens when poster is played and has @alt text -- how does it get passed on?

JF: poster not played -- an image -- alt from video relates to video itself (name of video, most likely)

RS: if wasn't used for fallback, does poster get put into video to start?

JS: there is a way to create the poster -- separate image -- now another way to do poster with first frame of actual video -- all that happens is that first frame is shown as if static image until play initiated

JF: sounds correct

RS: so now it is being rendered as video

JS: first frame of video frozen (paused/no audio)

RS: how does AT interact with video?

JF: when page first loads, recommendation is that video does not auto-start, so have asset in ready-state -- user has to initiate play -- from visual perspective, if don't declare @poster, will be frame of vide, usually first frame, which is usually a black frame -- if author doesn't do anything else, alt directly related to video
... poster like a curtain

GJR: poster is like a movie poster, video is like the "title card"

RS: trying to get engineering mapping picture clear

JF: agree with MCooper that should move on, but would very much like to continue this discussion on list

JS: agree

<inserted> scribenick: richardschwerdtfe

Janina: any other use cases we should discuss with this group

Michael: i have added that track and poster need to be included in the use cases for video
... Let me summarize the use cases: short text alternatives, long text alternatives, a label (this is accessibility and not), caption is different and gets confused with accessibility
... there is direct accessibility where you don't need a fallback
... there is something called a title attribute which is tooltip but not really the accessible alternative
... all embedded content needs a tooltip and no fallback is needed
... long text alternative applied to most things
... we did not apply caption to the audio and video elements

<inserted> scribenick: richardschwerdtfe

<inserted> scribnick: richardschwerdtfe

<oedipus> i/ijanina: still not done. give me another week/scribenick: richardschwerdtfe/

<inserted> scribenick: oedipus

RS: if i am AT would like to know using fallback because original object won't render
... might help solve some of the issues we have -- fallback means something entirely different
... if have poster, that is fallback content image

MC: identified something described elsewhere in document -- fallback for a11y reasons and fallback for non-support of objects have been confounded

JS: 2 values to table -- first wonderful collection mechansims to explain wshy things handled diff in diff circumstances so someone not in the ken in a11y can distinguish between alternative text and fallback content

<richardschwerdtfe> janina: what applies in different circumstances

JS: what applies in certain circumstances, such as video and poster -- any others to examine?

<inserted> scribenick: richardschwerdtfe

michael: like IFrame, it needs an alt text alternative

<oedipus> GJR logged bug against HTML5 on behalf of UAWG to make @title mandatory for FRAME and IFRAME since that is what ATs currently use

michael: you might use a short text alternative
... the audio and video elements need captions in summary

martin: it's also object, embed does not but it has fallback for object. So, flash itself could be accessible

<oedipus> GJR trying to find his media specific element unified approach wiki page -- seems to have disappeared

michael: they could use an image a flash, etc.

john: I tend to agree with sylvia's observation and that video should have caption or summary.

michael: we have not looked at mandatory and not what needs to be possible
... does anyone object to expanding it?

janina: oK: no objections

michael: this project is going beyond the scope of the bug triage subteam
... I guarantee that various people will not like the engineering due to the added attributes
... I am not sure that the bug triage team should work on this.

<kliehm> http://www.w3.org/Bugs/Public/show_bug.cgi?id=8885

janina: It seems this is a good approach to solving these issues

Michael: I think a gap analysis is needed
... we can expect pushback on the use cases
... we can support the use cases

<MichaelC> ACTION: cooper to do gap analysis between embedded content use cases and existing support [recorded in http://www.w3.org/2010/12/02-html-a11y-minutes.html#action01]

<trackbot> Created ACTION-90 - Do gap analysis between embedded content use cases and existing support [on Michael Cooper - due 2010-12-09].

michael: can we review the fallback use cases on the next meeting

janina: yes, but I am traveling
... we are expecting a change proposal from Everet tomorrow
... hopefully there will not be much to tweak it up

Issue 22

<oedipus> issue-22 lady of shallot change proposal: http://www.w3.org/html/wg/wiki/ChangeProposals/purely_decorative_images (reflects consensus at TPAC 2010 HTML WG F2F)

gregory: there is a change proposal on this
... it says that purely decorative images should be treated with CSS

<JF> bravo

gregory: .. and WCAG 2.0
... we should have one single place how to treat a decorative image

janina: more tweaking needed?

<Leonie_Watson> +1

<oedipus> plus 1 (if i do say so myself!)

janina: I will ask on list regarding concensing on this

<JF> +1 to moving it forward

<oedipus> issue-31 first 2 paragraphs of the definition of img change proposal: http://www.w3.org/html/wg/wiki/ChangeProposals/first_2_paragraphs_of_definition_of_img (wording comes DIRECTLY from TPAC 2010 HTML WG discussions)

gregory: I had a task assigned me at TPAC
... there was not a place to tie in so neatly

<oedipus> http://www.w3.org/2010/11/11-html-a11y-minutes#item02

gregory: I had difficulty making a change proposal

<oedipus> An img element represents an image.

<oedipus> The image given by the src attribute is the embedded content, and the value of the alt attribute is the img element's fallback content.

gregory: these two are ready to move forward as this is in concensus with what we agreed on

<paulc> Thanks everyone. Leaving to get ready for the HTML WG Weekly meeting.

janina: we are stalled on 31

John: I am watching the clock and we have 3 minutes left on the call
... I want to talk about longdesc
... is there any signal from PF as to when this will be approached?

<oedipus> closed action-72 (Draft a change proposal to issue 31 to update the definition of img element (was HTML WG action 194)

janina: We are coming back to it
... there is activity
... we have other priorities such as on fallback content

<oedipus> all reasearch and proposals at: http://www.w3.org/WAI/PF/HTML/wiki/Verbose_desc_reqs (includes laura's research)

janina: we are coming back to it but due to other priorities not right away

john: when?

janina: can't say just now

<kliehm> Currently there are 12 a11yTF bugs associated with ISSUE-31, at least 4 with ISSUE 30.

<oedipus> http://www.w3.org/WAI/PF/HTML/wiki/Verbose_desc_reqs

gregory: I tried to make them attribute agnostic

<JF> Request that the Task Force press the chairs for answrs t omy measurement questions

<JF> how much is "enough"

<oedipus> plus 1 to JF

I posted the limits

- DRAFT -

HTML-A11Y telecon

02 Dec 2010

Attendees

Contents

Action Items

Action 85

86

87

Fallback mechanisms

Issue 22

Summary of Action Items

Scribe.perl diagnostic output