<MichaelC> trackbot, start meeting

<trackbot> Date: 08 October 2013

<scribe> scribe: Kathy

<Joshue108> zaki, take up item 1

Survey for October 8 https://www.w3.org/2002/09/wbs/35422/20131008comments/results

AWK: survey is on the lengthy side, Number 1, 5, 7, 10 are unanimous

RESOLUTION: Accepting 1, 5, 7 and 10 as proposed

2. ARIA Technique: Using ARIA Live Regions or role=alert to Identify Errors

AWK: People like the idea of this but there are a couple of concerns.

<Joshue108> +q

KW: If you add aria-describedby and aria-invalid to the form fields then a screen reader users does not have to remember what they heard through the live region

Sailesh: Concerns about #10 and would like to know what the reasoning is
... On the aria live region, if using the role="alert" do you need to use aria-live region. If you just use role="alert" then you do not need to use aria-live region - it works across browsers and AT

Josh: what is the ways to do discover the error message; role = alert and errors message being read when get to the field

AWK: what is the bug with aria-atomic issue. Was this in the past? What currently supports. We can add this to the support notes

<Joshue108> KW: Well its definitely not required, so if its not done its not a WCAG error - but its a good technique.

<Joshue108> KW: Especially if there is a long list of errors.

AWK: We should not make this too complex but we should let people know what the limitations are. Maybe adding the additional information and have the task force create the technique to show the aria-describedby and aria-invalid

Sailesh: Sent the example through the email

<AWK> http://mars.dequecloud.com/demo/form-alert2.htm

AWK: here is the link from Sailesh

Sailesh: this is for aria role="alert" and aria live region, does not have aria-describedby

AWK: Goal is to get the technique passed through and a sufficient way to describe errors. Adding aria-describedby is not required. Suggest to the taskforce to create an example of this. We can add note about the connection of error message to the field and link to the example when it exists

Sailesh: It should be a separate technique. It depends on the number of error message and how the errors are implemented in the field

Kathy: agree

<Joshue108> +1

AWK: if this is equivalent aria-live=assertive then we should add an example. Josh mentioned that this may be confusing.
... is this necessary?

Josh: the use cases are not always clear. We should indicate which is the preferred method

<Joshue108> KW: There is a reason to do aria-live=assertive - sometimes role="alert" isn't announced in FF and IE -

<Joshue108> KW: So there is a case, we can't cover all versions but it is important to show various use cases.

Josh: Where do we draw the line. Should we include in user agent notes. This one could get messy.

AWK: Another way to handle it, is to add a note about the difference

Josh: The aria live region is assertive when using role="alert"

AWK: Someone needs to spend some time for the suggestions for this. We need to give a note to the taskforce about aria-describedby and aria-live='assertive".

Josh: I will do this.

<AWK> ACTION: Josh to provide suggested edits to http://www.w3.org/WAI/GL/wiki/Using_ARIA_Live_Regions_or_role%3Dalert_to_Identify_Errors [recorded in http://www.w3.org/2013/10/08-wai-wcag-minutes.html#action01]

<trackbot> Created ACTION-219 - Provide suggested edits to http://www.w3.org/wai/gl/wiki/using_aria_live_regions_or_role%3dalert_to_identify_errors [on Joshue O Connor - due 2013-10-15].

<kerstin> I'm trying to join via skype (facing some problems)

RESOLUTION: Leave open.

3. LC-2844: Out of date User Agent Support info

AWK: Comment about the user agent support is not current. This is something that changes. Combine the user agent support into separate document. The information may be redundant. We could have a place where we link to the user agent support
... One of the comments, the ARIA should not be handled in a different way. The proposal is to have all the user agent notes handled in a separate document
... The comments had a few ideas about where these should go
... edit the proposed resolution based on the comments

<kerstin> sorry for delay

AWK: take a look at the change

<AWK> https://www.w3.org/2006/02/lc-comments-tracker/35422/NOTE-WCAG20-TECHS-20130905/2844

<Joshue108> +1

AWK: questions, comments, concerns

<kerstin> +1

+1

RESOLUTION: accept as amended

<Joshue108> +q

<AWK> ACTION: Michael to handle the transition to handling user agent support notes in a single separate document. [recorded in http://www.w3.org/2013/10/08-wai-wcag-minutes.html#action02]

<trackbot> 'Michael' is an ambiguous username. Please try a different identifier, such as family name or username (e.g., cooper, melledge).

<AWK> ACTION: Michaelc to handle the transition to handling user agent support notes in a single separate document. [recorded in http://www.w3.org/2013/10/08-wai-wcag-minutes.html#action03]

<trackbot> Created ACTION-220 - Handle the transition to handling user agent support notes in a single separate document. [on Michael Cooper - due 2013-10-15].

<MichaelC> drop action 2

LC-2853: H45 update for chromevox

<Joshue108> +1

RESOLUTION: Accepted as amended

6. aria-describedby comments LC-2848 and LC-2847

AWK: a few comments. Loretta's we will take under advisement.

James: We removed a number of items from this technique to confuse items with longdesc

<AWK> http://www.w3.org/WAI/GL/wiki/Using_aria-describedby_to_provide_descriptions_of_objects

<Joshue108> +1 to rename

James: maybe we should rename the images example and remove the 3.2.2 reference

<Joshue108> yes, agree with James

James: we don't want to lose the rest of the aria-describedby; we don't want to soften it too much
... we need to include user agent note about aria-describedby not reading in all user agents
... most user agents ask to hear additional information. NVDA does not - it reads automatically.

AWK: should we say some

James: have to be careful not to imply how this works with the API
... suggests some proposed language

<AWK> https://www.w3.org/2006/02/lc-comments-tracker/35422/NOTE-WCAG20-TECHS-20130905/2848

AWK: updated the language
... does this capture what James was saying

James: yes, we may want to point out the Voiceover does not support longdesc but does support aria-describedby

<Joshue108> correct

AWK: it goes deeper still but I am not sure we want to do that. We do in the longdesc technique we have user agent information. Either technique is ok
... is the information balanced?

<Joshue108> +q

Josh: I think it is fine.

<Joshue108> +1

AWK: any objections?

RESOLUTION: accepted as amended

8. HTML Technique: Using heading tags to markup row header cells that group rows

AWK: Comment show concern about this
... Saliesh was this from you?

<AWK> http://mars.dequecloud.com/demo/Census_2013.htm

Saliesh: yes, this technique was created with table headers with groups
... under each of the groups there is series of form fields. The group can be defined as a heading. A user can scan the headings and understand the structure of the table. The table navigation keys can then move through the table. The group header can be read by JAWS

<Joshue108> +q

Sailesh: method avoids headers and id markup

AWK: there are two tables

Sailesh: the one on the bottom shows this technique

Josh: I am confused about this. to what degree this is made accessible by user agent heuristics

Sailesh: the headings provide a way to navigate to the section of the table you want to view. In this technique assumes one level of headings

Josh: concerned about adding headings

Sailesh: it is valid code and you can add headings

Josh: extra layer of complexity, HTML has semantics to mark this up

James: what about using both; the header and id could be added with headings

<Joshue108> +q

Sailesh: the second example simplifies the markup for the developer
... tools if used to create the table markup but developers need to remember to add it
... Voiceover support is limited
... is it is another method but not required

Josh: interesting case. Could also have these as 4 separate tables. Not convinced about this and it may be confusing

James: it would simply for some developers; people who are auditing code would state that it does not meet 1.3.1 for relationship between header cells

AWK: Also concerned about this. There is markup for conveying these relationships within tables. It does not look like there is not full support

Sailesh: it is valid code. Josh mentioned about breaking it up into multiple tables but this is not always an option

James: If you are moving down the column. How do I know I am skipping the cells? It is not reading the new headings

Sailesh: the markup will provide the structure of the table

<Joshue108> +q

James: we would not know what all the headings are for the cell. You would need read multiple different items - headings and cell headers

Sailesh: Ins+D in JAWS will read the nearest heading

James: one keystroke vs. two keystrokes

<Joshue108> -q

Sailesh: this is alternative to headers and id; some developers do not code like this. The recommended way is headers/id but this is an alternative approach where it cannot be used

<Joshue108> KW: I wouldn't want to give devs an out from using header/ids.

<Joshue108> +1 from me too.

AWK: agree with Kathy. Users will be navigating using table navigation. There is value to having headings within cells but have a hard time passing a table if there is not markup to associate the headers

James: what about making this an advisory technique

AWK: do you think that the example is sufficient to meet 1.3.1

Josh: navigating by headings could bounce the user into the middle of a table without them knowing it, so I don't like the sound of that.

AWK: what is the advisory?

Sailesh: Advisory is when you don't have assistive technology support

AWK: This could also be a usability technique
... if we say it is not sufficient then another technique would need to used

Sailesh: it is similar to landmarks and headings being used. Users will figure out the structure

AWK: what do we want to do here

<kerstin> what about leaving it open having a second look on it and asking SR-users?

AWK: do we have any heading technique about adding it within a table?

<Joshue108> I'll do that.

AWK: does someone want to look at the techniques for heading to see if we can incorporate this into the technique

Josh: I will do that

<AWK> ACTION: Josh to investigate adding table example to existing table techniques. [recorded in http://www.w3.org/2013/10/08-wai-wcag-minutes.html#action04]

<trackbot> Created ACTION-221 - Investigate adding table example to existing table techniques. [on Joshue O Connor - due 2013-10-15].

RESOLUTION: not accepted

AWK: We will pick up next week at #9
... Sailesh is asking about #10
... there is a technique and it is ready for review for the working group

<kerstin> bye

<AWK> We start next week with #9

bye

This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/I wonder degree this will work on complex tables by user agents/ to what degree this is made accessible by user agent heuristics/
Succeeded: s/with headings a user could get into the middle of heading without knowing that they are in the table/navigating by headings could bounce the user into the middle of a table without them knowing it, so I don't like the sound of that./
Found Scribe: Kathy
Inferring ScribeNick: Kathy
Default Present: +1.978.443.aaaa, +1.650.506.aabb, +1.703.225.aacc, Michael_Cooper, AWK, Kathy_Wahlbin, Peter_Korn, Sailesh_Panchang, Joshue, Marc_Johlic, kerstin, James_Nurthen
Present: +1.978.443.aaaa +1.650.506.aabb +1.703.225.aacc Michael_Cooper AWK Kathy_Wahlbin Peter_Korn Sailesh_Panchang Joshue Marc_Johlic kerstin James_Nurthen
Regrets: Alan David
Agenda: http://lists.w3.org/Archives/Public/w3c-wai-gl/2013OctDec/0009.html
Found Date: 08 Oct 2013
Guessing minutes URL: http://www.w3.org/2013/10/08-wai-wcag-minutes.html
People with action items: josh michael michaelc

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.