Accessibility Guidelines Working Group Teleconference

15 Oct 2019


AWK, Janina, Bruce, AlastairC, Chuck, Fazio, jeanne, Lauriat, Rachae, Detlev, Laura, Raf, MichaelC, pkorn, JakeAbma, MarcJohlic, KimD, johnkirkwood, Katie_Haritos-Shea, david-macdonald
Bruce, Laura


<bruce_bailey> Scribe: Bruce

<Zakim> bruce_bailey, you wanted to ask if we are on for CSUN

<bruce_bailey> Bruce asks about CSUN

<bruce_bailey> Silver and AG meeting might be too crowded

<bruce_bailey> still exploring options

<bruce_bailey> still TBD

<bruce_bailey> http://docs.google.com/document/d/17ByXEqqXuqWtDzxq2J6Vch27n4RWMZig7P426aPiIto

<bruce_bailey> Challenges with Conformance for Large, Complex, and/or Dynamic Websites

<bruce_bailey> Peter Korn and Janina to speak to this

<bruce_bailey> draft by PK

<bruce_bailey> PK: we have been thinking a lot about conformance in Silver

<bruce_bailey> ... useful first step to look a challenges for large dynamic sites

<bruce_bailey> ... applications as well, also perhaps also non-web ICT

<bruce_bailey> ... document aims to try and capture issues in this context

<bruce_bailey> ... thanks for many comments and edits, pk has not review yet

<bruce_bailey> PK: questions and comments?

<bruce_bailey> Jeanne: for people 1st looking at document, an overview please?

<AWK> Challenge #1: Specific WCAG Guidelines & Success Criteria Expecting Human Involvement

<bruce_bailey> PK: Good idea, there are four buckets

<bruce_bailey> ... many SC requirements involve human involvement

<bruce_bailey> ... from a time point of view, exponential time burden for dynamic sites

<bruce_bailey> ... anything which requires human review is huge challenge

<AWK> Challenge #2: Large, complex, and dynamic websites are always “under construction”

<bruce_bailey> ... content being added and changed faster than can be reviewed

<bruce_bailey> PK: may be updating a flow

<AWK> Challenge #3: 3rd party content

<bruce_bailey> ... conformance section of WCAG does not make allowance

<bruce_bailey> PK: even smaller sites, e.g., dating, will have lots of editors

<bruce_bailey> ... users not well education about accessibility

<bruce_bailey> ... ex 1.4.1 not only using color, how could that be moderated?

<AWK> Challenge #4: Non-Web Information and Communications Technologies

<bruce_bailey> PK: things that cannot be automated much wider for non-web ICT

<bruce_bailey> PK: document has abstract in each section

<bruce_bailey> ... several SC captured, but by no means complete

<bruce_bailey> ... very much a work in progress

<bruce_bailey> ... please do comment

<bruce_bailey> Jeanne: Work is building on what we learned 3 years ago in Silver research project

<bruce_bailey> ... thanks PK and others contributing.

<Zakim> alastairc, you wanted to ask whether a more declarative approach would be desirable, or should we explicitely allow for that in the conformance?

<bruce_bailey> ... we have not returned to anaylsis in a while, so this is very helpful

<bruce_bailey> Alastair: Recognize all challenges from work with FaceBook and other companies

<bruce_bailey> ... wrt to testing automatically, organziation address by creating processs

<bruce_bailey> ... "how we do things" instructions / guidelines

<bruce_bailey> ... so not fully automatic, but oriented to in-house processess

<bruce_bailey> ... would this kind of approach be supported by Silver?

<bruce_bailey> ... Or would you see tests being more declaritive and testable?

<bruce_bailey> PK: This is coming up in conformance meetings

<bruce_bailey> ... one thing that can be effective strategy is abstracting from raw HTML

<bruce_bailey> ... ex, wiki -- where tool does not allow ed user to create large text size distinct from named style

<bruce_bailey> ... helps force structure

<bruce_bailey> ... another example, cannot contain H3 under H1 without there being H2

<bruce_bailey> ... can imagine constrained environment where tools help with conformance

<bruce_bailey> ... does not work everywhere, no use of color is counter example

<bruce_bailey> Alastair: That is helpful, but I am still not clear how Silver conformance model allows for this

<bruce_bailey> PK: Focus is on capturing challenges as opposed to conformance approach

<bruce_bailey> ... an organization might define its own accessiblity standards, and Silver might allow that

<bruce_bailey> ... one example is how 508 permits equivalent facilitation

<bruce_bailey> ... that is outside WCAG 2.0 conformance model

<bruce_bailey> ... WCAG currently very strict about what is conforming or not

<bruce_bailey> ... But might Silver be able to stretch or stray from this approach?

<bruce_bailey> Alastair: Concerned with more of a continuim between how organizations address conformance

<bruce_bailey> ... if Silver varies widely, harder to have conformance with Silver

<bruce_bailey> PK: Might be resolved by looking at Guideline or Principle level rather than stricly looking at tests and methods

<bruce_bailey> Alastair: Organizations might take an approach to force heading with CMS, so easier to efforce accessibility

<bruce_bailey> ... Still not sure how Silver might endorse organization approach.

<bruce_bailey> PK: How to organizational address 1.4.1?

<bruce_bailey> Alastair: Has to be instructional

<Lauriat> +1 to Alastair, that first group seems more of a testability question than a conformance model question (I also commented on this in the doc).

<bruce_bailey> PK: So how to address chat with "meet me at the corner, I will be wearing a red hat"

<bruce_bailey> DavidMcDonald: Have you looked at WCAG EM

<alastairc> https://www.w3.org/TR/WCAG-EM/

<bruce_bailey> PK: not yet

<david-macdonald> https://www.w3.org/TR/WCAG-EM/

<bruce_bailey> DM: WCAG EM has quite a bit of overlap with questions this document asks

<bruce_bailey> ... might be useful as you EM tries to answer many issues raised

<bruce_bailey> ... EM is not a page-by-page approach, uses sampling

<bruce_bailey> AWK: agrees EM is on point

<bruce_bailey> PK: not sure that sampling alone address issues

<bruce_bailey> ... what if sample hits "under construction" or user content?

<bruce_bailey> ... will review

<bruce_bailey> Janina: Exercise is what can we build to evaluate sites?

<bruce_bailey> ... human need side we understand

<bruce_bailey> ... for example, alt and synchronized alternative not going away

<bruce_bailey> ... describing problems of organizations is new to us though

<bruce_bailey> Detlev: Distinction should be made between users and editors and 3rd party users

<bruce_bailey> ... bold and emphasis are not show stoppers

<bruce_bailey> ... dialogs and forms are a bigger problem which WCAG EM does not address

<bruce_bailey> ... EM just oriented to page testing

<bruce_bailey> ... need is for addressing building blocks

<bruce_bailey> ... example focus and triggers can be managed in building blocks, which defeats arguements about dynamic content and size of site

<bruce_bailey> PK: Those are important strategies, and we are using them

<bruce_bailey> ... but with process from software development (includes JavaScript and HTML) and organization can do good job with testing those

<bruce_bailey> ... but they will not be bug free, and bugs effect accessibility

<bruce_bailey> ... these strategies still run into 100% non-errors bar that is in the conformance model

<Detlev> Fully agree that 100% may not be doable at that scale!

<bruce_bailey> ... conformance is page-based, so one error falls conformance

<bruce_bailey> ... current model does not provide for less-than-perfection

<bruce_bailey> ... current model does not credit catching most problems most of the time

<bruce_bailey> Katie: We are having push back from browser developers, and maybe they only care about browsers, so can we focus maybe on building blocks?

<bruce_bailey> ... if we focus on authoring process, that is where we can get most of work done, especially with large websites.

<bruce_bailey> PK: Agree. These are all useful and important strategies and lead to getting arms around problem.

<bruce_bailey> ... still, none lead to zero SC failures

<bruce_bailey> Katie: Focus can be on improvement, and fewer errors.

<bruce_bailey> ... still need a bar. Can we talk about it differently?

<bruce_bailey> Janina: This conversation really helps get us clarity on where the problem lies.

<bruce_bailey> ... with the tool, with the tech specs, with the practice and policy?

<bruce_bailey> AWK: Next steps?

<bruce_bailey> PK: From the larger Silver plans, we would like to publish this document with FPWD of Silver in November.

<bruce_bailey> ... Question for this group: How much of this document (and its concerns) be bundled with Silver?

<bruce_bailey> AWK: It could be part of Explainer Doc

<bruce_bailey> ... It could be a separate note

<bruce_bailey> ... which would make more sense if it is applied broader than Silver

<bruce_bailey> PK: We are just trying to outline issues, it seems like it could be a stand-alone note.

<bruce_bailey> Micheal Cooper: Speaks against notes proliferation

<bruce_bailey> ... could easily be part of Silver Requirements document

<bruce_bailey> Detlev: Seems like this could apply to building blocks requirements, then does matter if millions of pages tested

<bruce_bailey> Katie: agreed

<laura> +1 to Detlev

<bruce_bailey> PK: Building blocks might not speak to flows

<JakeAbma> +1 it's the way most companies start doing tests anyway

<alastairc> +1 to *enabling* the testing of template level building blocks, but would need to have a view on how they are put together.

<JakeAbma> Make the building blocks accessible as a start

<Detlev> you's still have to have confromance req "complete processes" is some other form...

<bruce_bailey> ... can i log in, can i find the things i am searching for, can i complete X process, can I buy Y

<bruce_bailey> ... an analysis of what use is of App is important

<alastairc> Could be a method of 'sampling'.

<bruce_bailey> ... this also mirror how software is currently tested

<Zakim> Lauriat, you wanted to suggest next steps

<bruce_bailey> ... are all of these flows accessible?

<bruce_bailey> Shawn Lauriet: Suggesting for going forward...

<bruce_bailey> SL: like this as Silver requirements document

<bruce_bailey> ... Building block instead of pages makes perfect sense

<bruce_bailey> ... addressing less than 100% conformance, less than 100% testing of all pages sensible

<bruce_bailey> ... from quick review of EM note, that note does not completely address, but is helpful

<bruce_bailey> PK: Do people think this document is helpful in 2x world?

<alastairc> not sure, need to think about that

<bruce_bailey> SL: Yes, which is why concerns could be migrated into EM note.

<bruce_bailey> PK: Who is editor of EM note?

<bruce_bailey> discussion...

<Lauriat> Not concerns into the note, but some of the things we discussed, like building block evaluations and authoring tool level methods.

<bruce_bailey> MC: Evaluation Framework from ACT in formal process

<Lauriat> https://www.w3.org/TR/WCAG-EM/

<bruce_bailey> Evaluation Framework seems to be a little different

<bruce_bailey> AWK: There was an Evaluation Methodology taskforce, which no longer exists

<bruce_bailey> discussion that probably EM lead to AF ACT work

<bruce_bailey> AWK: ACT work is trying to capture a complete process

<pkorn> I need to step away from the audio conference for a few moments. My ability to send audio dies about 45 minutes into every conference... (sigh)

<bruce_bailey> Katie: ACT work much more focussed on automated testing

<pkorn> OK, I should be back now.

<bruce_bailey> Jeanne: Silver is looking at both

<bruce_bailey> ... Silver looked at EM doc at beginning

<bruce_bailey> ... updating EM does not seem likely, but it can inform Silver work

<bruce_bailey> PK: Updating EM does not seem likely in near term, so what is home for this content?

<bruce_bailey> AWK: We have not considered updating EM, might or might not be sensible

<bruce_bailey> ... this document important, but the right fit is still under discussion

<bruce_bailey> ... If this is going to be published in coordination with Silver FPWD, time is tight.

<bruce_bailey> AWK: How can WG help?

<bruce_bailey> PK: I see many comments, so there is work to resolve and respond to comments

<bruce_bailey> ... document needs revision and expansion

<bruce_bailey> ... time seems too tight to be part of Silver per se

<bruce_bailey> ... but can work on it simultaneously

<bruce_bailey> AWK: Everyone has link. Please read and offer suggestions.

<bruce_bailey> ... we will raise the topic on this call soon

<bruce_bailey> PK: If publishing week of 11/25, need review by 11/5 or 10/29

<bruce_bailey> Jeanne: Still working on scheduling.

<bruce_bailey> ... this doc is part of that. I will moved to Silver GitHub.

<Zakim> laura, you wanted to ask why does it need to be published in November?

<Lauriat> +1 to Laura, that seemed to come out of nowhere.

<bruce_bailey> AWK: We have Silver First Public Working Draft in November

<bruce_bailey> Laura: Understood, but why does this doc need to published then?

<bruce_bailey> AWK and Janina think this document important to context for Silver

<bruce_bailey> AWK: Topic very much of interest. Again, AGlers, please review.

<bruce_bailey> thanks from PK and Janina

<laura> Scribe: Laura

WCAG 2.2 Criteria review – Essential controls. (2nd week) https://www.w3.org/2002/09/wbs/35422/essential-controls/results

<david-macdonald> test

WCAG 2.2 Criteria review – Focus Visible (Enhanced) (2nd week) https://www.w3.org/2002/09/wbs/35422/focus-vis-enh-acceptance/results

awk: I had some qiuestions. So did detlev

detlev: difficult to measure.

awk: I had question on second bullet

AC: needs to be thick or separated from the component.

<bruce_bailey> https://docs.google.com/document/d/1g9_WBgfhViWAaRFIWWt10CP5rBsEVIWm3vT1vWqrHvI

detlev: don't see that in the SC text.

ac: adjacent colors doesn’t work with focus indicators.

<alastairc> https://alastairc.uk/tests/wcag22-examples/focus-visible-enh-examples.html

awk: felt like focus was not as promentent as others in nomensa example.
... not on example page
... red area looks like it gets shorter.

ac: technically passes.

awk: the get in touch example has a similar problem.

ac: that is example 6. I put that down as a fail.
... Nomensa could be an implementation if improved.

awk: stumbled on 3rd bullet.
... maybe something we can do in the normative text.

ac: adjacent to the user interface component.

<david-macdonald> last bullet last word should be "pixels" not pixel (nit)

ac: one of the techniques is to do an outset.

awk: is there a quantity that we are trying to achieve?
... 2 px thick.

ac: assumption that it is a line.
... will come back to thatl

awk: strange that we are not talking about area.
... suggested edit in survey.

ac: contradicts bruce’s suggestion.

awk: think area is a concept people will understand.



<Detlev> http://3needs.org/en/testing/code/focus-visibility.html

<alastairc> Did you see this? https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html

Detlev: example seems to meet the requirements.
... if so maybe set the bar higher.

AC: test page is at: https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html
... 1 px on the longest edge can just mean an underline.
... careful not to set up difficult targets.
... happy to increase. what would you suggest?

detlev: not sure require a gap, stronger contrast...

ac could increase to 4.5 to 1

scribe: could increase the surface area.

<alastairc> Jake, did you see the place lower down with "Working out the surface area of a circular focus indicato"

jake: wondering about SC text. Focused on rectangular shapes. What about irregular shapes?
... and shadows.

<JakeAbma> https://storage.googleapis.com/spec-host-backup/mio-design%2Fassets%2F0B9msDEx00QXmWDkwVGd5dU1mWUE%2Fpurpose-01.png

<alastairc> Example 13-21 are from material design: https://alastairc.uk/tests/wcag22-examples/focus-visible-enh-examples.html

jake: discussing examples.
... need non-retangular examples in understanding

ac: went with “surface area” for non-retangular.
... for shadows compare to background to get surface area.

jake: not always 2 pixels or longest side.
... maybe go with star example.

<alastairc> For shadows / gradients: https://www.w3.org/WAI/WCAG21/Understanding/non-text-contrast.html#figure-blue-circle-i-versions

david: worried about testing.

<Zakim> AWK, you wanted to ask about using a circumscribed rectangle to determine the longest side length

awk: for detlev- background does need to be considered.
... line would need to be 2 pixels thick. still on the edge.
... for an odd shape use a circumscribed rectangle to determine the longest side length.

ac: would need to be 2 pixels or separated.
... rather increase surface area compared to increasing contrast.

detlev: updated my examples but still hard to see.

awk: seems similar to 27.

detlev: looks thicker.

<alastairc> https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html

ac: 15 comparing sizes of indicator

detlev: could be an either/or situation.

ac: 4 was easier to see than 9

detlev: gap makes all the difference.

awk: seems like we got some things to figure out.
... we are close to moving to editors draft but some small loose ends.
... Jake had a comment: G149 needs the update of course, so not ready, and also add 'might' => "there's another (AA) SC for which it might not be sufficient." (test it!)

jake: need to find this text.

awk: It looks like there is a different edit than described for G149. We need clarity on which way.

jake: there are sites where default browser focus is ok. need to test it.

ac: need to change google doc text.

awk: jake comment. “The SC is on the right track but needs more work.”

jake: Not rectangular shapes are a concern.
... need an example.

AC: if I do an example would that help?

<Detlev> what is the alternative??

AC: will do.

<Detlev> +1 to rectangle around object

awk: longest edge is what we are targeting.

ac: need a mathematical term for it.
... any suggestions for text?

awk: will give it some thought.

<Detlev> "the longest side of the rectangle enclosing an irregular object"

Open issues are related to contrast and thickness, irregular shapes, and techiques updates.

RESOLUTION: leave open

awk: close but important questions.

s/promentent /prominent /

s/thatl /that /

s/jake comment /jake's comment /

trackbot, end meeting

Summary of Action Items

Summary of Resolutions

  1. leave open
[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/10/15 17:11:47 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
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/doid/did/
Succeeded: s/ out tha/ put tha/
Succeeded: s/simialr/similar/
Succeeded: s/adjecent/adjacent/
Succeeded: s/contridicts/contradicts/
Succeeded: s/undersrtand/understand/
Succeeded: s/thin area/think area/
Succeeded: s/incfrease/increase/
Succeeded: s/What a bout/What about/
Succeeded: s/compareed/compared/
Succeeded: s/ had a comment/Jake had a comment/
Succeeded: s/dont /don't /
Succeeded: s/adacent /adjacent /
FAILED: s/promentent  /prominent  /
Succeeded: s/nomensa /Nomensa /
Succeeded: s/implementartion /implementation /
FAILED: s/thatl /that /
Succeeded: s/tagets/targets/
Succeeded: s/rectangula /rectangular /
FAILED: s/jake comment /jake's comment /
Succeeded: s/open issues related to contrast and thickness , irregular shapes, and techiques updates./Open issues are related to contrast and thickness, irregular shapes, and techiques updates./
Default Present: AWK, Janina, Bruce, AlastairC, Chuck, Fazio, jeanne, Lauriat, Rachae, Detlev, Laura, Raf, MichaelC, pkorn, JakeAbma, MarcJohlic, KimD, johnkirkwood, Katie_Haritos-Shea, david-macdonald

WARNING: Replacing previous Present list. (Old list: Rachael, alastairc, Raf, Detlev, Fazio, Justine, Laura, Jennie, Brooks, Chuck, MarcJohlic, stevelee, Katie_Haritos-Shea, mbgower, david-macdonald, janina, bruce_bailey)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, Janina, Bruce, AlastairC

Present: AWK Janina Bruce AlastairC Chuck Fazio jeanne Lauriat Rachae Detlev Laura Raf MichaelC pkorn JakeAbma MarcJohlic KimD johnkirkwood Katie_Haritos-Shea david-macdonald
Regrets: JustineE
Found Scribe: Bruce
Found Scribe: Laura
Inferring ScribeNick: laura
Scribes: Bruce, Laura
Found Date: 15 Oct 2019
People with action items: 

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

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)

[End of scribe.perl diagnostic output]