Silver Task Force & Community Group

14 Feb 2020


jeanne, Chuck, sajkaj, CharlesHall, bruce_bailey, PeterKorn
Joe_Cronin, Shawn
Shawn, jeanne
Rachael, , sajkaj


Design for How-to (and name)

<Rachael_> scribe?

<Rachael_> scribe: Rachael

<jeanne> How-to design: https://raw.githack.com/w3c/silver/ED-draft=comments-changes-js/guidelines/explainers/template.html

<Rachael_> Jeanne: We've all been working hard to address comments. We woudl like to discuss comments that have come up a number of times.

<Rachael_> ...we had a request to see the template. I created a link to that. Anyone who sees it will be able to understand the structure.

<Rachael_> ...we added normative and informative labels.

<Rachael_> ..they are explained, I hope. If someone who is good on plain language can review it and provide feedback, we'd appreciate it.

<Rachael_> Jeanne: I put it at the end. What do people think about that? We have had suggestions to move the methods first.

<CharlesHall> +1 to at end

<Rachael_> Peter: Curious about color contrast of the green bar. It meets current calculation but have we checked it against the newer calculation.

<Rachael_> Jeanne: Would someone who has the new calculation check that?

<Rachael_> ...there should be links on the test page.

<jeanne> Rachael: THere is an argument for putting it at the beginning to meet the expectations of WCAG users. Not a hill to die on.

<bruce_bailey> my preference would be to keep methods on the end

<Rachael_> Charles: If it fails, it may just need to be bold

<Rachael_> Chuck: I +1 at the end because it is a jumping off point.

<bruce_bailey> Agreed, can't be first. So last is second best spot.

<Rachael_> Charles: I don't think it should be first because it has a lot of technical content and code/pseud code

<Rachael_> Janina: I agree with that rationale.

<Rachael_> Rachael: That is a good argument.

<Rachael_> Jeanne: Vote on it or can we move on?

<Chuck> +1 for last

<Rachael_> ...is there anyone who feels we should move it to the first spot?

<Rachael_> Jeanne: Lets leave it last and see what feedback we get from AGWG.

<Rachael_> ...was there anything else on the design for the how to?

<Rachael_> ...I will take all the guidelines we have written and put them into this format.

<Rachael_> Charles: When you tab forward to the methods page, the link to guidelines in the breadcrumb doesn't work

<Rachael_> Jeanne: If you look at the guideline, I say the sections below contain helpful information. On the method, I want to say all of it contains helpful information but is not required

<bruce_bailey> eye-dropper says background light green is B8E5D0, so SAPC of 134%

<Rachael_> ....that addresses a lot of the design related changes. Does anyone else have comments they've seen that this redesign didn't address?

<Rachael_> Charles: Is it true that nothing in the methods section will be normative?

<bruce_bailey> Black on pink (#FBECEE) is SAPC of %145.

<Rachael_> Jeanne: That is correct. The tests today are not normative.

<Rachael_> Charles: One of the challenges that is clear from the comments is the need for an understanding map between WCAG 2.x and Silver.

<Rachael_> Jeanne: I will paste in a link and you can tell me if I've addressed it.

<jeanne> ED draft https://raw.githack.com/w3c/silver/ED-draft=comments-changes-js/guidelines/index.html#guidelines

<Rachael_> ..link is the new editors draft, guidelines section. I started off with an explanation of what is normative and what is informative. I talked about why. The 3rd paragraph maps things.

<Rachael_> ...then I put in a table that maps.

<Rachael_> Jeanne: If you have suggestions, I welcome the help.

<Rachael_> Peter: My biggest suggestion is to change the name Silver to WCAG 3

<Rachael_> Jeanne: I'm eager to do that but I'm holding off until its formalized. Then I'll create a branch with the name change only. I will do that last.

<Rachael_> ...the survey closed. Anything else on introduction?

<Rachael_> CharlesHall: I think its sufficient. I want to come full circle. Nothing in the tests section including the tests are considered normative.

<Rachael_> Jeanne: I think that is how it is today so that shouldn't be controversial. I expect that the bigger concern will be the lack of an SC.

<Rachael_> CharlesHall: I've always seen tests as required because you can't test the guideline without them.

<Rachael_> Jeanne: We've been having this conversation on and off for a couple of years. We think that's not necessary.

<Rachael_> ...Its something the AGWG decided to do with the 2.0 structure but W3C has changed some of their specs since then. We spoke with several people in government orgs and they didn't feel they needed normative text.

<Rachael_> bruce_bailey: These guidelines seem awfully squishy for being the only normative parts. Clear language and contrast are too open. Sufficient is difficult.

<Rachael_> We ended up with use dark on light with ADA because it's clear. Sufficient is not really testable until we give a test. But if the tests aren't normative what do people do? How do you prevent churn?

<Rachael_> ...it might work but I think we need more crunchy bits.

<Rachael_> Jeanne: Do you think this is something to address in the draft or something that we address going forward?

<Rachael_> Bruce: I think we still move forward but I think we need to plan to address it. Its a big question.

<Rachael_> Jeanne: We haven't had a chance to talk about this in a lot of detail. I think its an important issue to spend more time on.

<Rachael_> Jeanne: Is this a topic for the f2f?

<Rachael_> Bruce: I think we need to address it sometime.

<Rachael_> Jeanne: Now that we can see it laid out, this is what bubbles to the surface. I will put it on the f2f agenda

<Rachael_> Jeanne: Lets move onto the scoring example. This is a big issue but is being worked.

Scoring Example

<jeanne> https://docs.google.com/document/d/1LfzTd_8WgTi0IUOOjUCRfRQ7e7__FRcnZow4w7zLlkY/edit#

<Rachael_> Chuck: I'm supportive of discussing normative at f2f

<Rachael_> Jeanne: We care about the results of the website vs. the website itself. I dug out a spreadsheet of results for a 32 page dynamic web app. Complex, dynamic. I only took the first 2 columns.

<Rachael_> ...I would like people to picture that we have new guidelines comparable to WCAG 2.x. This is how it was scored using WCAG 2.x and how it would be scored using WCAG 3.x

<Rachael_> Jeanne: Are we measuring pages, paragraphs, sites, images? What are we measuring? I wanted to illustrate that. I marked which are instance-based.

<Rachael_> ...most I left as page base since that is the results I have.

<Rachael_> For images they have 83% correct. For 1.3.1 it failed horribly. 6%. Part is instant based. Part site-wide. That will take reworking. Correct reading sequence is page based.

<Rachael_> I went through and scored them. Then I took the score from the clear language rubric and put that in. I added it and divided by # of rows applicable. They got a score of 60%

<Rachael_> It would be a fail under any circumstances. I took those numbers and plugged them into the next section. For each criterion, I plugged it in under each usage. I may have been too generous saying it benefited people with cognitive disabilities. If anyone can review that, let me know.

<CharlesHall> the cognitive requirement would include the quality of the alternative text

<Rachael_> ...I went through all SC and which plugged in where. I came up with a % under each disability category.

<Rachael_> ...When we have enough content, we need to start measuring sites to see which qualify bronze, silver and gold

<Rachael_> ...Is that clear? Lets discuss.

<sajkaj> scribe:

<Rachael_> Charles: I think people will latch onto the number.

<sajkaj> scribe: sajkaj

pk: Concerned what the site is doing, so cannot evaluate whether a particular failure is really, really important or incidental
... Uses example of alt, was it critical to function?

js: Alt guidance not written

pk: Is all of 2.x approach wrong over alt?

js: We need to show how it applies
... We're having so much trouble explaining scoring at all

pk: Don't see how the score is meaningful without context.
... If score measures how accurate to reality, sure. It does that.
... But that doesn't explain whether critical site functionality is accessibly available, or blocked

js: Agrees, but doesn't have an idea of how to demo it

pk: Understand and wish I had something more constructive

ch: Relates to a question I was going to ask about description of page vs. instance based

pk: We've also talked about critical flow in Silver

<Rachael_> +1 to critical flow

js: Oh! Could add a column
... So, how to score that?

sj: Suggests we're turning to statistical analysis and scoring, but that's not our expertise; and we need to get expert help building the model well

js: If we can include it, I'd like to

rm: I've done this with a similar scoring system
... On the critical path we forbid the top score if error exists on the critical path

ch: Similar to non interference?

rm: A combo of path and scoring

js: Will to try to intro this in a new version

rm: Asks how this plays with conformance model

js: It plays with it and we haven't worked it out yet
... Will try to build it in, but may need to lay it over to the next draft
... Moving to scoring a rubric
... Using the same real world (anonymized) experience ...
... Believe i got a good measure on how they would have scored on that
... Short paragraphs was about 73% and shows one could still pass with a low socre somewhere
... Want to keep moving to make sure we get to open comments ...
... Some sections sitll not done ...

<CharlesHall> i will do the Scope section this weekend

js: Abstract and Intro

rm: Accepts the action

js: Available all weekend for you
... Anyone able to do accessibility supported?


js: OK, we're still good

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2020/02/14 20:00:52 $

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/Chrales/Charles/
Present: jeanne Chuck sajkaj CharlesHall bruce_bailey PeterKorn
Regrets: Joe_Cronin Shawn
Found Scribe: Rachael
Found Scribe: 
Found Scribe: sajkaj
Inferring ScribeNick: sajkaj
Scribes: Rachael, , sajkaj

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

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]