W3C

- DRAFT -

Web Content Accessibility Guidelines Working Group Teleconference

05 Jan 2016

See also: IRC log

Attendees

Present
AWK, Josh, wayne, jon_avila, JF, Sarah_Swierenga, MichaelC, Katie, Haritos-Shea, JamesNurthen, Laura, Katie_Haritos-Shea, Kathy, Mike, Elledge, Jan, Rakesh, Eric, LisaS, David_MacDonald
Regrets
Chair
Joshue
Scribe
AWK

Contents


<Joshue> [1] https://github.com/w3c/wcag/issues

trackbot, start meeting

<trackbot> Meeting: Web Content Accessibility Guidelines Working Group Teleconference

<trackbot> Date: 05 January 2016

+AWK

AWK+

<Lisa_Seeman> i am beeing told the meeting password is oncorrect

<Lisa_Seeman> trying wcag

<Lisa_Seeman> for webex, should the password be wcag?

<Lisa_Seeman> sorry, .my fault

<Mike_Elledge> I'll scribe

<Mike_Elledge> JO: Katie gets Karmic Brownie points for scribing next week.

Quickref Update

<yatil> https://w3c.github.io/wai-wcag-quickref/?currentsidebar=%23col_customize&tags=images

<yatil> https://github.com/w3c/wai-wcag-quickref/issues?q=is%3Aissue+milestone%3A%22Public+Review+Comments%22+is%3Aclosed

<Mike_Elledge> EE: Update on quick ref. Good feedback on functionality and polish. If go to tab there are ticks on buttons to make clearer. Share button on bottom. Other techniques to satisfy success criteria.

<yatil> https://github.com/w3c/wai-wcag-quickref/issues?q=is%3Aissue+milestone%3A%22Public+Review+Comments%22+is%3Aopen

<Mike_Elledge> ee: Remaining issues, close 2 of 3 this week. Fix some missing techniques. Then final review and publish.

<Mike_Elledge> ee: Thx for good feedback during public phase. Went really well and appreciate!

<Mike_Elledge> JO: Final look and feel complete?

<Mike_Elledge> ee: This is final. Looked at it again and simplicity won out.

<Mike_Elledge> JO: One comment: color coding principles and guidelines. Would nicely break them up and make it easier to relate.

<Mike_Elledge> ee: Color coding is hard to translate and make an accessible version. Think it's clear enough.

<Mike_Elledge> ee: Blue spaces and indentation are pretty good for differentiating.

<Mike_Elledge> JO: Find blue blocks a bit too much. Maybe a think blue line to left if I was doing it.

<DavidSurface4> thanks Eric

<Mike_Elledge> JO: Thanks.

New edits for COGA extension

<Ryladog> Note: I liked the big blue blocks

<Mike_Elledge> LS: Have been working on...

<Lisa_Seeman> https://www.w3.org/WAI/PF/cognitive-a11y-tf/wiki/Proposal_for_WCAG#Instructions.2C_navigation_and_non-trivial_information_are_provided_with_a_clear_writing_style

<DavidSurface4> lost you lisa

<Mike_Elledge> LS: We made a draft of COGA (Cognitive) to take to WCAG. Want to put in most useful form. Feedback was that not testable enough for adoption.

<Mike_Elledge> LS: Started looking at rewording it. Haven't finished, but first pass on most content. Want feedback then do a second pass. Have revised suggestion for success criteria for writing style.

<Mike_Elledge> LS: For example, changed: "Clear writing style includes: (then defined points)." Have some more things to define and tweak. Sure WCAG will want to make others. Also have exceptions.

<Mike_Elledge> LS: For example, if "University has a particular style." Or if it's a literary work may be an exception.

<Mike_Elledge> JO: General question: How will police this, how to write testable procedures besides author's subjective opinion?

<Mike_Elledge> LS: Human testable until tools catch up. If you look at list you see testable criteria.

<Mike_Elledge> JO: Have gone into grammatical criteria?

<Mike_Elledge> LS: Yes. "Short clips...etc."

<Mike_Elledge> JO: These could be used as algorthms. Notifying the user via a browser plug in May have a subjunctive clause out of place..etc. Why don't you try x?

<Mike_Elledge> LS: "Use simple rather than complex words." So use 3,000 most commonly used words. Seemed a sensible number, can change. Idea is that tester can identify word not commonly used.

<Mike_Elledge> James: Some more testable than others. Can't differentiate between opinions and facts.

<Mike_Elledge> K: Based on certain assumptions.

<Ryladog_> Based on research

<Mike_Elledge> LS: Would be "can you reword it, here's a contrary POV"

<Mike_Elledge> JN: Muddying the waters if something can't be tested.

<Mike_Elledge> LS: Tried to go with 9 out of 10 approach.

<Mike_Elledge> JF: Have a problem with 9/10 approach. May have client that disagrees.

<JF> +1 to James' point

<JF> s/JF: Have a problem with 9/10 approach. May have client that disagrees./JN: Have a problem with 9/10 approach. May have client that disagrees.

<DavidSurface4> originally it was 8/10

<Mike_Elledge> LS: Two issues. For opinion and facts, 9 out of 10 is quite strong.

<DavidSurface4> then we changed it to a "strong correlation"

<Mike_Elledge> LS: Understand about paying vendors, want to be black and white.

<Ryladog_> why do I keep getting kicked out

<Mike_Elledge> JN: One other concern. Some of this seems English-biased. Would it work in German, too.

<Mike_Elledge> LS: Ranking style can be replaced based on local needs.

<Mike_Elledge> JA: Concerned about testing too. Did you consider Flesch-Kincaid? All feeds into readability.

<DavidSurface4> your pinning out JF, turn down volume or push mike a way a bit

<Mike_Elledge> LS: Looked at it. Written out what the criteria are, can be changed if user testing indicates it. If there's an algorithm can use it, if appropriate.

<Mike_Elledge> JA: Trying to explain will be difficult, like to be mechanical whenever possible.

<Mike_Elledge> LS: Have provided exceptions based on active voice and present tense.

<Mike_Elledge> JA: My concern, how to we measure it. How do we measure conformance?

<Mike_Elledge> LS: Easy to create a tool to identify passive or future tense. Don't want to limit, but easy to build 99% of time.

<laura> Some existing readability tests: Gunning-Fog Index, Flesch Reading Ease, Flesch-Kincaid, grade level, Reading Level Algorithms. Check:

<laura> http://juicystudio.com/services/readability.php

<Mike_Elledge> JA: Have seen tools. My concern is not tools, but what is the cut-off. The line that says this is compliant, this is not. Lots of middle/grey ground.

<Mike_Elledge> LS: Give me an example.

<Mike_Elledge> LS: These are pretty much yes or no.

<Mike_Elledge> JF: General use.

<Ryladog_> Cue please

<Mike_Elledge> LS: Look at rest of phrase. There are lists available.

<Mike_Elledge> JO: Let's not get lost in the weeds--see both of your points. Substantial queue. Good input.

<Mike_Elledge> WD: Couple of things. Answer to different languages. In some cases there's judgement involved. What do you mean by non-trivial.

<Mike_Elledge> LS: Put in a definition. Something that user may search to find.

<Mike_Elledge> LS: If writing a blog unlikely someone would search. If looking for SC, would look up. Will be more obvious in final format.

<Mike_Elledge> WD: Would take me 2 hours to figure out if something was trivial. This will be a hard sell for academics.

<Mike_Elledge> LS: Have given exception. If a particular style or format.

<JF> As an add-on to the resources Laura listed: http://john.foliot.ca/readability-and-its-implications-for-web-content-accessibility/

<Mike_Elledge> JO: Clear writing will be a benefit for everyone.

<Mike_Elledge> WD: Many academics agree.

<Mike_Elledge> LS: There's an exception if you will be penalized.

<Zakim> laura, you wanted to ask have exitsting readability test been considered?

<Zakim> Ryladog_, you wanted to discuss the 9 out of 10. In WCAG 2 we used 8 out of 10.

<Mike_Elledge> LS: Laura can you send to me in email.

<Mike_Elledge> Katie: Won't be able to rely on only machine test. General consensus was 8 of 10 for WCAG. Will have to stick with that. Understand concern about tools, but this is about human beings.

<Mike_Elledge> Moe: Lot of good dialog. Logistical question: 12 points in clear writing, some seem at higher level than others.

<Mike_Elledge> LS: A and AA something we're considering. AA would be if tested all content. Have to be careful not to lock out people.

<Mike_Elledge> JO: Interesting comments. Hope it's useful.

<Mike_Elledge> LS: Very, especially opinion vs. fact. Will try to make as machine testable as possible, not withstanding Katie's comment.

<Mike_Elledge> JO: Next steps? Review document again?

<Mike_Elledge> LS: We should wait until we've gone through it all again. Comments about format would be helpful now, however.

<Mike_Elledge> LS: Then we'll ask for a complete review.

<Mike_Elledge> WD: Looked up defintion of non-trivial, maybe want a different term. Doesn't seem to match my understanding, maybe urgent instead?

<Mike_Elledge> LS: Trying to get to stage where can hand it off to WCAG. Assume there will be future back and forth.

Github issues in survey https://www.w3.org/2002/09/wbs/35422/20161stSurvey/

<Mike_Elledge> JO: Thrust of it is very good. Work on it and then send it to us for review.

<Joshue> Issue 80

<Joshue> https://github.com/w3c/wcag/issues/80#issuecomment-141807642

https://www.w3.org/2002/09/wbs/35422/20161stSurvey/results

<Mike_Elledge> AK: Ask people about their comments for Issue 80.

<Mike_Elledge> JR: Visual information seems out of scope. If person has turned off CSS then sighted person won't be able to see it.

<Mike_Elledge> JO: If image does provide information...

<Mike_Elledge> JN: All images providing important information, have to differentiate between background images and content images. Need to make it clear that those images don't have same contrast issues.

<jon_avila> +1 we need the ability to indicate what is decorative and what is contextual

<Mike_Elledge> JN: Background images disappear in high contrast mode on Windows, content images don't.

<jon_avila> Not a bug IMO

<Mike_Elledge> JN: Intentional Windows behavior.

<Mike_Elledge> mea culpa, JF. Have misid'd JF for JN.

<jon_avila> +1 to JN

<DavidSurface4> http://www.w3.org/TR/css3-background/#the-background-image

<Mike_Elledge> JA: Check CSS for specs about user agents. Otherwise depending on behavior of particular OS.

<jnurthen> "For accessibility reasons, authors should not use background images as the sole method of conveying important information. See Web Content Accessibility Guideline F3 [WCAG20]. Images are not accessible in non-graphical presentations, and background images specifically might be turned off in high-contrast display modes."

<Mike_Elledge> WD: Would consider using background image for content to be a failure of 1.3.1

<DavidSurface4> James can you drop in a link to that

<laura> F3: Failure of Success Criterion 1.1.1 due to using CSS to include images that convey important information: http://www.w3.org/TR/WCAG20-TECHS/F3.html

<Mike_Elledge> WD: has to be programmatically determined. Can't tell if background image is important.

<Mike_Elledge> JO: Shouldn't be using background images for important content.

<Jan> URL for quoted text: https://drafts.csswg.org/css-backgrounds-3/#the-background-image

<Mike_Elledge> AK: First: fail 1.1.1 if image doesn't appear because of css. Don't think you do, based on wording. Not sure what success criteria this falls under, perhaps 1.3.1, but doesn't fit well, and worry about overloading 1.3.1.

<Mike_Elledge> AK: Maybe a task force issue.

<Ryladog_> Maybe a Low Vision TF issue

<Mike_Elledge> AK: Do people think if image doesn't appear is a failure of 1.1.1?

<Mike_Elledge> JO: If you have an image in the background, can you add a label to provide an alternative.

<Mike_Elledge> JF: Technically you can if you have an aria label on the parent, and put focus on it.

<jon_avila> aria-label is not displayed when images are turned off where as alt on img is

<Mike_Elledge> JO: But won't be in DOM where images are generally kept.

<Mike_Elledge> JF: If it's focusable (with tab) maybe that's the criteria.

<Mike_Elledge> Katie: Has to be visible.

<Mike_Elledge> Talking about losing visual information.

<Ryladog_> By moving focus to an invisible content - myou would create a fail for Visual Focus

<Rakesh> I have to add something

<Rakesh> In addition to what JF says we need to add role img to the div that talks about the image

<Mike_Elledge> DM: Couple of other things. Need a technique with a div with an aria label on it. Sometimes feel it's important to have item as background image. May need tabindex=-1 to get into the ally tree, but focus not needed.

<Mike_Elledge> JO: Should steer away from User Agent issues.

<laura> Low vision Use case #2 under Color and Contrast section: http://www.w3.org/WAI/GL/low-vision-a11y-tf/wiki/User_stories_-_use_cases#Color_and_Contrast

<jon_avila> +1

<Joshue> +1 to Wayne and AWK a la sending this to LVTF

<Mike_Elledge> WD: Access to alt text is a serious problem. Like contrast ratio not sufficient, big technical gap wrt low vision.

<Mike_Elledge> JA: Background image should not include content, AT won't be able to differentiate between content and decorative. Maybe create a new CSS property. Need to be able to address the issue. Until we have something more clear hesitate to remove it.

<Wayne> +1

<laura> +1

<Mike_Elledge> JO: Maybe an opportunity to inform developers about need for sufficient contrast.

<Mike_Elledge> JA: Trend is to invert colors, not same as high contrast. Like in iOS. Have to allow people to change images the way that work for them.

<yatil> Firefox can also change text/background colors, also user stylesheets.

<Mike_Elledge> JO: Need to increase awareness.

<Mike_Elledge> Katie: Don't remove it. Want to bring up focus thing. Shoud not hijack focus to make this functional. Would remove focus for people who need it.

<Mike_Elledge> JO: Katie update on comments.

<Mike_Elledge> AK: Not to remove failure.

<Mike_Elledge> JO: Just modify test procedure.

<Mike_Elledge> AK: Core issue whether having visual information programmatically available should be added. Doesn't mention visual failure now. Somewhat confused about what people are talking about removing.

<Mike_Elledge> AK: If we don't agree with the question, that would be the response. But doesn't feel trivial to figure out all the permutations of the discussion.

<Mike_Elledge> Katie: Based on 1.1, transform content, it could be interpreted that way. Maybe it's an equivalent purpose by having it visual. Visually it still being there is an equivalent purpose.

<shorton> I believe this is an issue with other types of content, not just images

<Zakim> jnurthen, you wanted to ask if we can solve this by just limiting this to background images and to state that if building something on iOS only then this isn't a failure either....

<Mike_Elledge> JO: Use of background images is more frequent, and need to address it. Loop it in with low vision task force.

<Mike_Elledge> JN: Could we solve by modifying failure: Be clear that we're talking about background images and not content.

<shorton> The issue is really the use of CSS for content in general

<Mike_Elledge> JO: Have to make failure more watertight. Is this more relevant to background images?

<Mike_Elledge> JN: Yes.

<Mike_Elledge> JO: Maybe call out in title for failure.

<Joshue> ME: The use of background images for content is to reduce the amount of space.

<Joshue> ME: They are lighter.

<Joshue> JN: Techniques bundle these into one, a la sprites.

<Joshue> ME: Exactly.

<Joshue> ME: Its tough to get devs to change that method.

<Joshue> JN: Using content images with the same techniques is possible.

<Joshue> ME: Thats note been brought to their attention IMO.

<yatil> I think that this technique might also superseded with SVG <use>

<Joshue> JN: Image fonts also are good.

<Joshue> ME: Maybe we need another technique.

<Joshue> JOC: Eric is making the same point about SVG.

<jon_avila> so that's why SC 1.3.1 may apply

<shorton> agreed

<Mike_Elledge> WD: Using a background image for content is not programmatically discernible.

<Mike_Elledge> WD: Can't detect which O

<Mike_Elledge> S is being used.

<Joshue> JR: We can't say that because we are just on windows.

<scribe> Scribe: AWK

Upcoming Public review of new WCAG Understanding and Techniques docs.

<Jan> JR: Agree with Wayne that can't assume that background images are never hidden when not on Windows

<Joshue> Undestanding Doc: http://www.w3.org/WAI/GL/2016/WD-UNDERSTANDING-WCAG20-20160105/

JO: We will be sending out a CFC on the list today for the public release of the Techniques and Understanding docs

<Joshue> Techniques: http://www.w3.org/WAI/GL/2016/WD-WCAG20-TECHS-20160105/

JO: Please take a look at these

<jon_avila> When are we meeting at CSUN?

AWK: Just a simple approval for public review

JO: No meeting planned for CSUN

KHS: Deque offered to host a F2F

<MoeKraft> Happy New Year!

<Sarah_Swierenga> Happy new year to everyone!

trackbot, end meeting

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/01/05 17:38:08 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/wayne present+//
Succeeded: s/Lisa, the Password is w3c//
Succeeded: s/Present+ Katie Haritos-Shea/Present+ Katie_Haritos-Shea/
Succeeded: s/ee; R/ee: R/
Succeeded: s/May have a subjunctive clause out of place./Notifying the user via a browser plug in May have a subjunctive clause out of place..etc. Why don't you try x?/
Succeeded: s/Present+ Katie Haritos-Shea/Present+ Katie_Haritos-Shea/
Succeeded: s/JF: Muddying the waters if something can't be tested./JN: Muddying the waters if something can't be tested./
WARNING: Bad s/// command: s/JF: Have a problem with 9/10 approach. May have client that disagrees./JN: Have a problem with 9/10 approach. May have client that disagrees.
Succeeded: s/JA:/JF:/
Succeeded: s/JF: All images providing/JN: All images providing/
Succeeded: s/JF: Background images disappear/JN: Background images disappear/
Succeeded: s/JF:  Intential Windows behavior./JN:  Intentional Windows behavior./
Succeeded: s/JA: Technically/JF: Technically/
Succeeded: s/JA: If it's focusable/JF: If it's focusable/
Found Scribe: AWK
Inferring ScribeNick: AWK
Default Present: AWK, Josh, wayne, jon_avila, JF, Sarah_Swierenga, MichaelC, Katie, Haritos-Shea, JamesNurthen, Laura, Mike, Elledge, Jan, Rakesh, Eric, LisaS, Kathy, David_MacDonald

WARNING: Replacing previous Present list. (Old list: AWK, Srini, JF, Laura, EricE, Jan, Joshue108, Kenny, marcjohlic, DavidMacDonald, MichaelC, Wayne, JamesNurthen, Kurt, waume, sayne)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ 


WARNING: Replacing previous Present list. (Old list: AWK, Josh, wayne, jon_avila, JF, Sarah_Swierenga, MichaelC, Katie_Haritos-Shea, JamesNurthen)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, Josh, wayne, jon_avila, JF, Sarah_Swierenga, MichaelC, Katie, Haritos-Shea, JamesNurthen

Present: AWK Josh wayne jon_avila JF Sarah_Swierenga MichaelC Katie Haritos-Shea JamesNurthen Laura Katie_Haritos-Shea Kathy Mike Elledge Jan Rakesh Eric LisaS David_MacDonald
Found Date: 05 Jan 2016
Guessing minutes URL: http://www.w3.org/2016/01/05-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]