W3C

- DRAFT -

Web Content Accessibility Guidelines Working Group Teleconference

13 Jan 2015

See also: IRC log

Attendees

Present
AWK, Mike_Elledge, Marc_Johlic, Michael_Cooper, EricE, Jon_avila, Kathy_Wahlbin, Katie_Haritos-Shea, Loretta, James_Nurthen, Joshue
Regrets
David, Kenny
Chair
AWK
Scribe
jon_avila

Contents


<trackbot> Date: 13 January 2015

<Joshue> <https://www.w3.org/2002/09/wbs/1/TutorialsCIT/>

<Joshue> Scrub of actions https://www.w3.org/WAI/GL/track/actions/open

<AWK> Chair: AWK

<AWK> https://www.w3.org/WAI/GL/wiki/Scribe_List#2015_Scribe_History

<scribe> scribe: jon_avila

Final Tutorials Review/Publication approval. <https://www.w3.org/2002/09/wbs/1/TutorialsForms/>

awk: first item is forms tutorial
... thank you to those who provided comments. Looks like we have variety of comments.

<AWK> Tutorial: http://www.w3.org/WAI/tutorials/forms/

awk: bottom line is that we'd like to publish and we have 1 supporting, 4 suggest changes, 3 say only with changes address, and 1 that says don't support w/ comments.
... Ideal to get a vote on whether we say yes or no -- we want to make decision either contigent or say not yet. We want to get this published as these are great resources.

marc: on fence as to say publish yet or not -- most of comments around getting ARIA examples and techniques. Could be swayed to move if there is more support to add in ARIA example after publishing.

awk: under form concepts - several comments about moving ARIA up in discussion. Made suggestion on different ways to structure language.
... understand we want to use native elements whenever possible. The way it is written makes it seems like ARIA is rare. There should be an easy fix.

Eric: can quickly comment as we go through; EO had been cautious on certain techniques -- want to be more neutral on techniques -- we could add more weight to best practices.

<Ryladog_> Focus should *first* be on HTML semantics - then - on ARIA

eric: tutorials are not set in stone -- will change over time. Will try to add more ARIA example as quick as possible

JO: only half in IRC -- but wanted to say happy to go as they are -- don't want to delay -- Eric says we can update as needed -- ok to put these out for review.

marc: Think tutorials are great -- like to get ARIA in there so we could point people to them.

khs: HTML semantics should be front and center -- ARIA is fine after -- room to add more about ARIA -- but keep ARIA as later option -- not first option.

ME: question on timing on when the updates will occurs.

Eric: Funding has run out on tutorials -- can find time here and there to update and make improvements in the near future

awk: Pull requests into github could be used to get updates in

Eric: we will just do updates -- but it will take some time from other projects that we have allocated. Hard decision to look at what we want to change.

JO: What Eric is saying is practical. Good to get these out and see where they go -- that would be prudent.

awk: key question to answer - does the info in their represent our view of what people should do for accessibliity to say WCAG really likes these -- it's a legitamite question.
... If these are in the current form for 3 or 6 moths -- is that we want to put out there? That's the question.
... tutorials don't need to represent every method of doing something.

JO: moving goal post of user agent support, etc. emphasis on various technologies, etc.

<Ryladog_> +1 to Josh

JO: We are playing catchup on many things -- so we need to look for clangers and address those -- but I didn't really see any. Is there anything you can't live with?

awk: move on to next seciton -- labeling controls. Charles had a few show stoppers. Not sure if any of them have been addressed.
... Charles says it's important to use the for attribute in the first example, awk not sure if it's a show stopper. Eric have you resolved any of these yet?

<Ryladog_> +1 to Charles comment

eric: Switching a few labels here and there are no big problem we can just make that change.

awk: Section on hding labels - do we need to update to make it clear that labels are important to users other than screen readers.

ME: think it is clear that you can understand visually but a screen reader user may need additional information. Ok with it.

awk: We need to recognize a reality of what people do.

JO: Emphasis there are certain use cases where this is the case to show the real world examples.

kw: Wanted to make a note about bugs where labels are hidden but they are not announced by VoiceOver or in Firefox. Could lead people down path where it might not work with AT.

awk: Eric posted note about hiding elements which talks about when labels are hidden with display none and visibility hidden.

kw: Code here would be fine with CSS clip.
... it's the off-screen one that prevents it from being spoken in FF and VO.

jo: Could catch from kathy - a note about anti-partern could be useful.

awk: Charlies search should always have search role -- that's a problem.

jo: disagrees that it isn't needed.

<yatil> ee: We can add a link to this section of the Page Structure tutorial: http://w3c.github.io/wai-tutorials/page-structure/sections/#search-section

marc: had commetns about adding in aria-labelledby
... As for role = search, not show stopper, but perhaps a note that could be added to one of these.

JN: Are we talking role of search or type = search?

eric: we have another section in tutoral that talks about page structure. We have a dedicated example there.

awk: what are the benfits of search is it just ARIA landmarks?

JN: Seems like adding role of search might dupliate search text.

awk: People on call seem to agree that role of search is not a show stopper.

JN: Did we discuss table layout CSS and what was the conclusion?

awk: we'll come back to CSS Table layout
... we are mentioning things that can allow you conform and we are applying an amount of preference through order but we are not explicitly saying recommended or not recommended. Everything in the tutorial is recommended but some things might be more strongly recommended. Any othter thoughts on that?

marc: Order may need to be tweaked -- e.g. explicit labels might come before implicit labels.

awk: Eric said he would re-order that for id one to be first. Any one else?

ME: somewhere we should say that we put them in priority or preference order. Could put at top so we are not repetitious

awk: What pages do we do that?
... Any suggestions on best way to handle that?

ME: put at beginning of each section, at beginning at tutorial and then each section such as form concepts. Hard call because people may come directly to specific pages.

awk: Same conversations on resources that have thick boiler plate text at the top of all techniques.
... having it be implicit is better than saying it's preferrable and cloud message
... prefer to have people see more useful examples and let people decide.

JO: agree with Andrew and Jon implicit structure without explicitly saying preferrable but might be useful to say that these are designed to support legacy situations.

awk: Does anyone feel like we should delay publication over this issue?
... Hearig no one -- we will keep it queue for later.

<jamesn> https://bugzilla.mozilla.org/show_bug.cgi?id=1007975

<yatil> http://w3c.github.io/wai-tutorials/forms/labels/#associating-labels-explicitly -> CSS table example

JN: problem with Firefox bug where content will be exposed table semantics

eric: CSS in this example doesn't add too much to the example. Would still work if we could rid of the CSS display: none. We wanted to show that they are not in one element so you can't use the implicit label.

awk: should still be clear since each part of the form is in a separate element. Will drop CSS display: table bit
... Question about content that would valid if copied out into a single document.

<jamesn> -1 to that

ME: Not a big thing but something to point out. e.g. some controls re-use ids in pages so examples are not unqie for the page.

JN: if you modify all ids in the doc to be unique then it will make it more difficult to read.

jo: Clearer code is better for this purpose.

ME: Can live with this -- James makes good point.

marc: Question about using ARIA at bottom was confusion, example talks about aria-label but didn't use aria-label. Perhaps the code sample needs to be updated.

JN: get rid bit about using aria-label attribute -- it's already been covered.

<Zakim> EricE, you wanted to say scope and richness of examples

<Ryladog_> My suggested view is HTML (including 5) and then ARIA

marc: Nice to have some things that are new -- but oh by the way it may not work in certain AT and browswer combinations.

<jamesn> example 2 from aria-9

marc: We do have some examples such as pulling labels from multiple sources.

+1 marc on labels form multiple sources.

marc: example such as phone numbers of dates -- would be good to have in tutorial

<Joshue> +1 to multiple label examples also

<AWK> ARIA provides a way to explicitly associate text on a web page with a form control, using the aria-labelledby attribute to reference the id of an element containing text that can serve as a label for the control. Like other uses of ARIA, support for the use of the aria-labelledby attribute is still developing.

awk: how would people feel about adding to end of explicit label section. See not in IRC

<marcjohlic> +1

<Joshue> +1

<yatil> +1

awk: This wouldn't be an example, but we would acknowledge aria-labelledby by and reference ARIA techniqueds

+1

<Mike_Elledge> +1 So long as you point to an example or further explanatino

<Joshue> ARIA provides a way to explicitly associate text on a web page with a form control, using the aria-labelledby attribute to reference the id of an element. This can contain text that serves as a label for the control.

jo: things that is fine -- first sentence is long. Put change in IRC

<AWK> or "ARIA provides a way to explicitly associate text on a web page with a form control, using the aria-labelledby attribute. This attribute references the id of an element containing text that can serve as a label for the control. Like other uses of ARIA, support for the use of the aria-labelledby attribute is still developing."

awk: Eric, are you agreeable to adding wording and techniques to techniques list?

eric: yes

awk: Rather remove things like g162 which we don't want people to do and add ARIA techniques. Not suggesting we get rid of G162.
... is optgroup support and useful to justify highlighting it in here?

JN: Is ti AT supported? Is there any other way to do what it is mean to do?

awk: abseent knowing more let's leave it.

To my knowledge it isn't well supported.

jo: there is some support but I can't remember.

<jamesn> Wayne’s comment: Really nice, especially labels.

<jamesn> … An ARIA alternative to <fieldset>: Many developers do not use the form element or fieldset element anymore, but grouping is important. What WAI-ARIA structure should we use to group for a collection of controls that make up virtual field set. We can use the landmark, form, for the entire collection. Do we just use a "region" role instead of a <fieldset>.

<jamesn> … I think we should make suggestions, because the programming trend of abandoning the HTML form is widespread.

jn: Wanye says allow aria grouping mechanisms

awk: ARIA grouping roles -- used once in images tutorial. Need to compare two situations -- using it for form controls is probably more important concept to have.
... people want to use aria grouping because of the issues around styling fieldset and legend.
... worreid that this will take more time. One of us could take an action to do a pull request to add a note that this is availalbe via ARIA. What do people think?

eric: it's not such a complicated example that we could add it with limited work. I will do it myself unless someone else wants to.

awk: sounds like we agree that it's a good thing to have.

mc: Tutorial should go into why and when to use fieldset because I think there is confusion about it. Radio buttons are only the beginning.
... would not block publication.

<AWK> ISSUE: We need someone to think about how to beef up the fieldset section of grouping section in the forms tutorial. "Should fieldset be used around controls other than checkbox and radio button groups? Why or why not? How is the user experience different from when a fieldset is not used?"

<trackbot> Created ISSUE-44 - We need someone to think about how to beef up the fieldset section of grouping section in the forms tutorial. "should fieldset be used around controls other than checkbox and radio button groups? why or why not? how is the user experience different from when a fieldset is not used?". Please complete additional details at <http://www.w3.org/WAI/GL/track/issues/44/edit>.

<jamesn> .... that sounds like an AT bug if that happens

<Joshue> yes

<jamesn> +(describedby for me)

<Joshue> Its an option

awk: we need to have it read expirary date and date format.

sorry my IRC keeps kicking me out

eric: sounds good to me.

awk: add ARIA 9 to techniques because we are using that.
... we are not going to make it through whole one. Other two tutorials were easier to review compared to this one.
... Are there any things in the remaining comments, section 6, 7,8,9, 10, and 11 in the tutorial. Are there any real show stopper.
... Léonie was only one who said don't publish it.
... correciton: Leonie

<Ryladog_> I am concerned that we are providing techniques and examples - not in this document as much - but rather in the WCAG documents - that are designed to address gaps in AT implementations, and this bothers me. We are having to re-do new techniques based on lesson learned out in the wild - which means we are not doing enough research/testing ourselves. I see it as a staffing/volunteer breadth issue. Putting out and then pulling back techniques is hurting u[CUT]

jo: should we say that this may not be strictly best practice but works in terms of WCAG compliance and best current accessibility supportit's what we recommend -- what is our position on this. The term best practice is taken the wrong way. There is a disconnect.
... Leonie says that some things are mis-explained or inappropriate.
... not sure if this will get out of the gate as quick as we think.

awk: perhaps we need to ask people to review existing comments and see what needs to be done. I feel we are quite close taking into account some things we agreed to today .
... certainly talking about these being published in short number of weeks not months

eric: plan to bring it to EO this week including the comments -- will get more complete picture

awk: Leonie is on EO
... will need people to review survey response sections 6 through 11.
... need aria grouping example.
... that's it, thank you.

<Mike_Elledge> Thx! Bye!

<AWK> Trackbot, end meeting

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/01/13 17:35:01 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.140  of Date: 2014-11-06 18:16:30  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/only have/only halve/
Succeeded: s/only halve/only half/
Succeeded: s/Really nice,/Wayne’s comment: Really nice,/
Succeeded: s/An ARIA alternative/… An ARIA alternative/
Succeeded: s/I think we should/… I think we should/
Succeeded: s/Leon/Léonie/
Succeeded: s/seciotn/section/
Succeeded: s/be best practice but /be strictly best practice but works in terms of WCAG compliance and best current accessibility support/
Found Scribe: jon_avila
Inferring ScribeNick: jon_avila
Default Present: AWK, Mike_Elledge, Marc_Johlic, Michael_Cooper, EricE, Jon_avila, Kathy_Wahlbin, Katie_Haritos-Shea, Loretta, James_Nurthen, Joshue
Present: AWK Mike_Elledge Marc_Johlic Michael_Cooper EricE Jon_avila Kathy_Wahlbin Katie_Haritos-Shea Loretta James_Nurthen Joshue
Regrets: David Kenny
Found Date: 13 Jan 2015
Guessing minutes URL: http://www.w3.org/2015/01/13-wai-wcag-minutes.html
People with action items: 

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


[End of scribe.perl diagnostic output]