Accessibility Guidelines Working Group Teleconference

13 Apr 2017

See also: IRC log


allanj, Laura, AWK, Glenda, Greg_Lowney, ScottM, JF, Katie_Haritos-Shea, MichaelC, KimD, jasonjgw
Ad hoc
jim, Glenda, AWK


<AWK> Chair: Joshue


<LisaSeeman> made it

<Glenda> I can volunteer to scribe.

<allanj> scribe: jim

<AWK> Scribe: Jim

Adapting Text options discussion

<laura> Proposal H (Level AA) and I (Level AAA): An in tandem 2 SC approach

<laura> https://www.w3.org/WAI/GL/wiki/Issue_78_Options#Proposal_H_.28Level_AA.29_and_I_.28Level_AAA.29:_An_in_tandem_2_SC_approach

<AWK> Issue 78: https://github.com/w3c/wcag21/issues/78

<allanj> laura: 2 SC AA and AAA, AA no mechanism, remove "at least", AAA has mechanism

<AWK> Current text of SC: https://rawgit.com/w3c/wcag21/adapting-text_ISSUE-74-78-79/guidelines/sc/21/adapting-text.html

<laura> David: "I think this a good use of wording from 1.4.5 "If technologies being used

<laura> can achieve it,..." It's probably the best of all of them... and I can support it."

<laura> https://lists.w3.org/Archives/Public/w3c-wai-gl/2017AprJun/0168.html

<allanj> laura: call for vote

<Zakim> Greg, you wanted to clarify my concern about the "at least" clause meant removing the entire clause, and my proposal for the split into two SC

<allanj> gl: clarify my past comments. not just remove "at least" but remove entire specified range

<Greg> First, at Level A, "If and when author formatting is overridden by the user agent, the page does not lose any essential content or functionality." Then, in the testing instructions, we'd list font family, foreground color, background color, line spacing, letter spacing, and word spacing. Note that those aren't the only types of formatting to which this should apply: for example, no information sho

<Greg> uld be conveyed by bolding or italics alone.

<Greg> Second, at level AAA, "The content provides a mechanism that allows the user to override the default formatting to, at minimum, the following: … Exceptions: ..."

<allanj> gl: AAA would apply to closed systems

<LisaSeeman> checking on my dog, will be back soon

<allanj> awk: gl wording changes how I think about SC.

<allanj> gl: if user uses the UA to change settings, then page should still work. at the AAA the UA is not available to make changes, so author must do more.

<allanj> awk: why don't you require XX, and it doesn't matter what UA does or does not support. Functionality is still necessary.

<allanj> ... must be able to respond to that Question.

<allanj> gl: suggesting the level is higher.

<allanj> awk: no, we have specified with WCAG2 that some things have to get done. don't care what the mechanism. it just needs to get done. use a widget or don't... but the SC need to be met

<allanj> ... balance between UA requirements that are not yet filled vs author doing lots of work.

<allanj> laura: the hard part is making the first one work.

<allanj> ... davidm adamant about hard metric for testing

<allanj> gl: true for both SC

<Zakim> JF, you wanted to say I generally agree with Greg L

<allanj> laura: wayne wanted all colors and fonts tested.

<allanj> jf: +1 to GL proposed wording.

<allanj> ... user should be able modify authors code (author proposed, user disposes)

<Zakim> shawn, you wanted to reiterate Andrew's point - what user needs regardless of UA capability

<allanj> shawn: does GL wording meet user need. is there an out if the UA doesn't allow change

<allanj> gl: then it falls to AAA and the author makes a mechanism

<allanj> khs: like using 1.4.5 lang. "if the technology doesn't allow..."

<Ryladog_> If technologies being used can achieve it,..."

<allanj> gl: proposal - A for when user agent overrides, because is easy. Author mechanism, then more work. so AAA or AA

<allanj> awk: so content is still funtcional when user with UA makes change ...

<allanj> gl: level A

<allanj> awk: any issues... mobile, etc.

<allanj> laura: testing?

<allanj> gl: questions about PDF

<laura> Testabilty: https://www.w3.org/WAI/GL/wiki/Issue_78_Options#Testabilty

<allanj> laura: it can be done in pdf

<allanj> awk: ... with different viewers

<allanj> awk: if user can't override because UA doesn't allow, then its not a failure. is that correct?

<allanj> gl: yes, that is correct

<laura> The SC 1.4.5 language “If the technologies being used can achieve..." to address scoping in H

<allanj> gl: pdf would pass A, but not AAA

<allanj> gl: if UA allows no changes, then author provides a dropdown to allow user to make changes to formatting.

<allanj> awk: how does pdf fail?

<Zakim> Greg, you wanted to say that the reason I lean towards AAA over AA for the second

<allanj> gl: author has to provide some mechanism to be AAA, to cover the "just in case" the UA has no mechansim

<JF> +1 to AWK regarding "providing a widget"

<shawn> +1 for not requiring widget if common UAs provide the functionality

<allanj> awk: not excited about having to require providing a widget

<allanj> jf: opposed to requiring a widget

<allanj> khs: at AAA no problem with widget

<allanj> awk: non disruption model as A, mechanism is available as AAA

<Greg> If you want to remove my proposed AAA SC that's fine. My preference is to say that AAA is defined as maximally accessible, and that means going the extra effort to be accessible even when there are limitations in the user agent--as there are in a lot of mobile browsers today, which don't yet support user style sheets.

<LisaSeeman> also i gave up a fmily day for the personlization bit

<LisaSeeman> so i would like to get to it

<Greg> So what did others think of my Level A proposal J: "If and when author formatting is overridden by the user agent, the page does not lose any essential content or functionality."?

<allanj> jf: can we make some resolution to move forward

<Joshue108> has Laura not done that already?

<laura> Greg’s: "If and when author formatting is overridden by the user agent, the page does not lose any essential content or functionality."

<allanj> awk: this call is not about action, only discussion. Laura send to list the options.

<Greg> Laura, my apologies if we took up too much time on the first topic.

<allanj> laura: did that

<Joshue108> +1 to AWK

<allanj> awk: then call for vote on list,

RESOLUTION: Group to review options for split SC on the list

<AWK> acxk la

COGA: Support Personalisation SC

<allanj> awk: many challenges from Tuesday call

<AWK> https://www.w3.org/2002/09/wbs/35422/SCs_April_11/results#xq4

<Joshue108> https://www.w3.org/2017/04/11-ag-minutes.html

<allanj> ls: perhaps move discussion to coordination call.

<allanj> ... asked for feed back off list... no response.

<AWK> Current language: https://rawgit.com/w3c/wcag21/support-personalization_ISSUE-6/guidelines/#support-personalization-minimum

<allanj> ls: survey responses. need more discussion .

<allanj> ls: we have changed wording to address comments. that moved us further away from consensus

<Ryladog_> Apologies I have another call.....ciao to all!

<allanj> ls: we want access to metadata. questions about ... is technology ready?

<allanj> ls: simplified version. 5 controls (including "more controls"), and controls must have icons and words.

<allanj> ... pushback on metadata. don't know what it is or what to do with it.

<allanj> ls: thousands of sites use metadata for personalization.

<allanj> ls: people are not comfortable. so we made changes. but it is watered down.

<LisaSeeman> no js. we gave an alternitve

<allanj> awk: 2 bullets - authors use metadate, and if not metadata do this.

<allanj> awk: concerns about test ability due to language. "important" controls, important to whom?

<allanj> ... concerns about 5 controls

<allanj> awk: others have thoughts?

<JF> +1 to Jim "5 controls"

<allanj> ls: did people look at definitions for terms

<allanj> ... if the definitions are not testable we can work on that. COGA is using language found elsewhere in WCAG

<allanj> ls: which definitions need work?

<allanj> ls: testibility of words?

<allanj> note: scribe missed a bullet from lisa

<allanj> jf: 5 items seems arbitrary. so have a fly out menu with 5 items, if I hover over an item then it flys out with 5 more times, does that pass

<allanj> ls: yes. the object ... is to get to a simplified interface.

<allanj> ... use definitions in semantics document for personalization. good examples

<allanj> jf: concern about specifying 5. the user need is "simplified " interface. agree with requirement. the 5 requirement is too prescriptive.

<LisaSeeman> https://w3c.github.io/personalization-semantics/

<allanj> ls: we want critical features in a simplified interface.

<allanj> ls: we want the second bullet in SC add semantics. is that possible in native apps. tried to make it easy by being prescriptive

<allanj> ... if we use language in semantics document ... reads from document

<allanj> ls: if you like the def in semantic doc, we can use that. but it seems vague.

<Zakim> AWK, you wanted to speak to critical features, channeling a comment about google docs

<LisaSeeman> This setting should be used on:

<LisaSeeman> Elements that are essential for the key function (from the user perspective) of the page (Example: The send button for an email draft) or

<LisaSeeman> Elements that are sometimes critical to a user being able to use the site, such as "save my work" or "emergency instructions" or

<LisaSeeman> Elements that are used by over 90% of a user group (such as parents or teachers) use this element most times they use the content

<LisaSeeman> Typically 3 links or buttons would be considered "simplest"

<allanj> awk: critical feature, important information. example google docs.

<allanj> ... do we expect that google docs would have an only 5 controls.

<allanj> ls: perfect examples. some users only need print, save, and x.

<allanj> ls: lets mark it up and see what people think. If SC is too web centric. then provide a simplified version. happy to change wording. not sure what to change

<allanj> jf; push back. in google docs. 4 icons save, undo, etc. then another section on zoom, then ... but these are all critical function.

<allanj> ls: there are some folks who only need 3 -5 "commonthings

<allanj> ... that are used 90% of the time. need an alternative interface. more button with all features, and a less button that only shows 5 things. thats the user need.

<allanj> ls: it was orignal proposal. got lots of push back. and things got worse.

<allanj> mp: critical features depend on what person is trying to do. reading vs writing. 90% of what tasks.

<allanj> ls: 90% of a user groups.

<allanj> mp: different interface for each group?

<allanj> ls: trying to meet needs of folks who can't deal with more than 5 options.

<Glenda> I wonder if a tool like Read & Write for Google Docs is a better solution…in other words…is this an AT issue. See https://chrome.google.com/webstore/detail/readwrite-for-google-chro/inoeonmfapjbbkmdafoankkfajkcphgd

<allanj> ls: want guidance.

<allanj> awk: people have concerns about technology or implementations or ...

<allanj> awk: concerns about altrnatives - semantics vs alternate interface.

<allanj> awk: seems most seem to think that this is not fully baked.

<allanj> ls: can we have a COGA call to work on wording? should we have just one bullet point - semantics

<Zakim> JF, you wanted to ask about "not based on facts"

<allanj> jf: ls said that concerns "not based on facts" what don't we know. clarify...

<allanj> ls: not enough time. we may disagree on that point. I need to build concensus. added bullet about 5. should that be eliminated of make it more flexible.

<allanj> ... what are people happy with.

<allanj> awk: talk on list some more.

<allanj> awk: need to hear back on survey comments...

<Greg> I'm curious about to what extent people are opposed to the approach of the SC, or just the details. (I'm in the latter camp; I think it's a good goal, and a decent approach, but there are details that would need to be fixed.)

<allanj> awk: still more work to do on this.

<allanj> ls: will it ever be enough. or will it be "its too much work" or is it wording?

<KimD> if this is to address intellectual disabilities - this doesn't apply to all sites/content

<allanj> awk: coga attributes only would be a problem.

<allanj> awk: to progress, need to have options that are not already covered by wcag and are not too onerous

<JF> "The Magical Number Seven, Plus or Minus Two"

<allanj> awk: don't know what number I would pick, 10 or 15 or more. not sure

<allanj> glenda: circular conversations are all around many SC. part of process. frustrating.

<allanj> glenda: concerned that we can't name a number. "important" is a concern.

<allanj> glenda: read/write - perhaps this is an AT solution.

<allanj> ls: perhaps 5 or definition of "important"

<allanj> glenda: can live with essential. and that is hard enough to define. don't want more definitions.

<Glenda> scribe: Glenda

awk: WCAG 2.0 use of essential is a binary.
... If we say 5 controls (or what is essential)…then it would be easy for developers to say “everything is essential”.

<KimD> +1 to AWK - people at my company would say everything is essential

Lisa: Do you have a suggestion on how to get around that issue?

JF: I think the longer range goal for using meta data to support personalization is the right strategy. We have to be very pragmatic about what is ready for today.

Lisa: That is why we are giving the alternative.

JF: But just because we have a standard does not mean it will be picked up.

Lisa: adding the metadata is easy.

JF: I believe adding the metadata for coga is not easy in practice.

+1 Glenda also believe adding the metadata for coga is too hard to actually do well. It is a huge undertaking.

Lisa: This is a chicken and egg problem. We need to require it. The metadata standard already exists.

Scribe is leaving…who would like to take over scribing?

<AWK> SCribe:AWK

<scribe> scribe: AWK

Kim: wants to take a step back
... concerned about very complex tools for professionals such as attorneys
... worried that this SC can't apply to every situation
... have hundreds of web sites and they won't be able to take on the addition of all these attributes
... to support this there would need to be exceptions

Lisa: there are ppl on COGA TF that are affected, people in professional settings who need to use complex tools
... huge cost to society
... people may be retiring 3-5 years early because of these challenges

Kim: reality is that not everyone can do the work of an attorney
... practice of law demands a convoluted and complex web site in many cases.

<KimD> +1 to MC

MC: highlighting the difference between issues related to implementability (Kim's ideas) and issues related to user needs (Lisa). We need to not regard these as the same issue.

trackbot, end meeting

Summary of Action Items

Summary of Resolutions

  1. Group to review options for split SC on the list
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/04/13 17:02:41 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/AAA author/AAA and the author/
Succeeded: s/is that correct/is that correct?/
Default Present: allanj, Laura, AWK, Glenda, Greg_Lowney, ScottM, JF, Katie_Haritos-Shea, MichaelC, KimD, jasonjgw
Present: allanj Laura AWK Glenda Greg_Lowney ScottM JF Katie_Haritos-Shea MichaelC KimD jasonjgw
Found Scribe: jim
Found Scribe: Jim
Found Scribe: Glenda
Inferring ScribeNick: Glenda
Found Scribe: AWK
Inferring ScribeNick: AWK
Found Scribe: AWK
Inferring ScribeNick: AWK
Scribes: jim, Glenda, AWK
ScribeNicks: Glenda, AWK
Found Date: 13 Apr 2017
Guessing minutes URL: http://www.w3.org/2017/04/13-ag-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]