Summary & Minutes of SIlver Conformance meeting of 3 December 2019

== Summary ==

CSUN Face to Face (F2F) meeting survey: We are trying to get a rough 
headcount, what days people want to meet, and any accommodations we need 
to plan for in order to meet at CSUN in Anaheim, CA the week of 9-13 
March 2020.  We did some troubleshooting of the form. If you can't type 
in your email address, reload the page. Otherwise, email Jeanne.

We are looking for sponsors to fund the F2F meeting.

We worked on the sampling proposal.  One undecided question is what we 
mean and intend by Conformance?  Many of us think that it is a claim, 
but others pressed that it should just be how people know they did 
accessibility correctly.  Most groups make conformance claims with the 
VPAT forms (now international), so what should Silver do?

We are in agreement of the sampling requirements for small sites, but 
had considerable discussion for large sites.  First, that there be an 
option to have conformance at a component level.  One recommendation is  
to have primary task flows that must be accessible.  There could be 
thorough testing of the task flow. Should it be pre-release or 
post-release testing?  They serve different purposes, and should Silver 
allow both?

We started discussing the point system and how some ideas from the NIST 
paper could be used to set up a point system that was more fair to 
different types of disabilities. The NIST paper proposed that the 
weighting of guideline points be by the reliability of the evidence of 
conformance. The formulas are normalized, so that no one guideline has a 
higher weight than another, and no disability has more weight than 
another. We could build the percentage of tolerance of failure into it.  
The final score can be expressed as a percentage.  There is a lot of 
math, but it could be hidden from the user of Silver, who would only 
know the points it is worth, not how the points were arrived at. There 
was a discussion of critical errors, but this has been an area that has 
been thoroughly discussed in the last six months and there is 
substantial consensus that we want to move away from that, because 
ultimately, it presupposes that a failure of a component is a greater 
barrier than a component for which workarounds exist.  The problem is 
that when the workarounds are all taken in totality of a website, they 
constitute a critical barrier.   The result is a unintentional 
structural bias toward disabilities with critical component failures.  
There is plenty more to discuss.

Janina gave an update on the Challenges document proposal.  They have 
addressed all but 4 of the Issues filed in Github and are hoping to 
discuss it again in AGWG next week, with the potential of publishing 
before the end of December.

== Minutes ==

https://www.w3.org/2019/12/03-silver-conf-minutes.html

Received on Wednesday, 4 December 2019 02:15:56 UTC