<Cyborg> just looked at the conformance spreadsheet, but would appreciate a walk-through

Jeanne: Shawn, do you want to start talking about the conformance prototype next steps?

Lauriat: We had the previous prototype and another way that conformance could work. We were thinking looking at each, fleshing them out and trying to build small-scale versions of each.
... and then looking at how each of them could work
... ease of use of seeing how well you conform, how well the ?? is doing, and maintenance
... so two aspects of two different systems, but trying to come up with something concrete to see how it works

Jeanne: another part of this is that Alastair was working on a User Needs overview
... he took some success criteria and mapped them to high level needs
... so how can we get this moving?
... need more eyes on it, more people writing things

Cyborg: What could be an easy first step to help people get started?

I need to step away for a second.

<jeanne> Cyborg: I have put some thought into gaming the system. One approach is to give the people who are gaming the system more of what they need so they put less focus into gaming the system. I have reached out to someone who is expert in

<jeanne> ... social responsibility in conformance.

<jeanne> Shawn: It was something that came to mind when looking at Tim Boland's proposal.

<jeanne> ... we need to handle before we can put it into a serious prototype

<jeanne> Cyborg: Gaming is an issue and probably needs a multi-pronged approach.

<jeanne> JF: One of the things that came out of the TPAC discussion in AG, was testable statements.

<jeanne> ... even with a scaled approach, there needs to be a pass/fail.

<jeanne> Shawn: That was part of our original prototype. From "this is terrible" to "this is awesome", then there must be a "this much be at least this good".

<jeanne> John: ISn't that a strict true/false?

<jeanne> Shawn: No, it is broader than that.

<jeanne> John: We need to draw a line in the sand on the measure of quality.\

<jeanne> Shawn: We need to take an example, like alt text, and move the needle toward improving things for people with disabilities.

<jeanne> ... we will have to draw a line in the sand for each of the Methods for what that will be.

<jeanne> John: I remain concerned about gaming the system.

<jeanne> Jeanne: Gaming the system is like security, you will never have perfect security and perfect protection from gaming. What we need to do is to improve it from today.

<jeanne> Cyborg: We tried an experiment over the summer to validate whether a site that had a good score actually was accessible to people with disabilities.

<jeanne> ... I couldn't figure out how to make it work, and abandoned it. It's important to do if we can figure out it

<jeanne> Jeanne: It was a good effort, and it could work. We just weren't ready to do validation yet. We have a lot of opinions, but we need to start working out the nitty-gritty details and start testing them.

<jeanne> Charles: That was what I tried to do with the email list last week when I was asking about qualitiative vs. quantification approach to Color Alone.

<jeanne> ... we got a lot of opinion, which was mostly "no", but how do we frame that as a test?

<jeanne> Jeanne: Let's set up the tests for Color Alone and write the tests and make up the data.

Jeanne: Is the question "will it work" or "can we prove it?"

<jeanne> Charles: The hard part is setting up the group by disability. You would need 6 people with color blindness, 6 people with low vision, 6 people with cognitive disabilties. Until we do that, we don't know if it will work

JF: Do we also need to make sure we're including people with combinations of PWD in our user pools? Where do we draw the line?

<jeanne> John: We have to also look at multiple disabilities.

JF: The more user testing we can create and the more diverse the pool is, the closer we get to the goal. We need a framework around that as well, but that's secondary to the compliance question.
... for conformance model it almost needs to be based on hard repeatable tests
... and bronze, silver, goal we include the harder to qualify tests

Jeanne: That's what we're doing.

Lauriat: Yeah, if you meet WCAG, you'll automatically meet Bronze
... I think we just need to start writing test for different things and try to explore how many different ways we can test it
... ex. look at color values, heuristics evaluations, butts in seats evaluation
... we need to explore the kind of testing for a few different types of guidance...include ones in WCAG and ones that couldn't get into 2.1
... then we'll have some of the boundaries we can use for the conformance model to accomodate

JF: I have a list that looks at A and AA criteria in two categories "mechanical" and "editorial" for how can we be compliant on social media when I have no impact on the platform
... maybe looking at the existing criteria in a similar way could help

Lauriat: I think that would be helpful
... it would also be interesting to put the lens of mechanical vs editorial on the guidance that couldn't make it into 2.1

<Cyborg> for prototyping: sounds like it would be helpful to identify 4 SC for a conformance prototype devt, including: one that is easy to test via several methods, one that is hard to test and measure, one that is easy to game, and one whose success does not correlate with usability by PWD

<Cyborg> i'm also wondering if there is a "post market surveillance" question about conformance testing at the time of product development vs testing by users with intersectional disabilities (further towards the edges) to identify whether and how conformance works in the real world

<Cyborg> if you write a conformance test question similar to the one you wrote for plain language, it might help recruit people. maybe a conformance subcommittee could help too

Jeanne: We might need to devote meetings to writing up tests
... I'm reaching out to some folks that might be able to help as well

Charles: And we did conduct a conformance survey. The research report from that is available.

<Cyborg> do we know which 4 SC might match the types i mentioned?

<Cyborg> or is that a question to put out to Silver group as well?

<Charles> Conformance Survey Report: https://docs.google.com/document/d/1iOut3_i1JBQu5_16plZ8u7xCd66T4TbW2zPR1KmWWF4/edit?usp=sharing

Jeanne: We probably need to think about it, since we have the research paper with WCAG SC that are easiest to learn, teach, and test(?) I forget the three things
... I don't want to put much more out to the group...at least not this week.

<Cyborg> just if they would stick to answering that question

<Cyborg> ?

<Cyborg> lol re: the question about active/passive voice

Updates from Information Architecture

Jeanne: We got an answer back related to the information architecture prototype

<jeanne> Laura Carlson's prototype: https://docs.google.com/document/d/1efjXi8fw9uxCKygQ3JZD4PAMtSCFMaH83hX3Yae8jd4/edit#heading=h.3jnqsivluuxt

JF: The understanding documents for that SC seems like it's pretty plain language

Cyborg: I used the understanding documents when working on the plain language...we should add that step to the style guide

<JF> https://www.w3.org/WAI/WCAG21/Understanding/text-spacing.html

JF: What about the use of imagery?

Jeanne: Yeah, we want to improve the use of imagery...but let's refocus on the information architecture.

Lauriat: It probably makes sense to combine the text spacing with some other SCs that are very technology specific to an overall rendering customization guidance
... we're trying to get away from overall high level guidance be technology specific...and this is a lot less technology specific, but it's still very specific to text spacing
... and we could include a few of these types of things at the same level

<Charles> quickly back to the conformance testing: one of the next steps I had imagined is collecting a list of similar models and then perhaps organizing them in a comparative way. I had forgotten that the conformance survey provided us with such an example of another model: the EFQM Excellence Model (http://www.efqm.org/)

<JF> trackbot, end meeting

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)

Default Present: Lauriat, KimD, mikeCrabb, Jennison, Makoto, jeanne, Cyborg, kirkwood, LuisG, Charles, JF, johnkirkwood
Present: Lauriat KimD mikeCrabb Jennison Makoto jeanne Cyborg kirkwood LuisG Charles JF johnkirkwood
No ScribeNick specified.  Guessing ScribeNick: LuisG
Inferring Scribes: LuisG

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 16 Nov 2018
People with action items: 

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.)