TPAC agenda

introductions Joe Cronin - amazon

https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan

WCAG SC conformance observations

PK: We have been thinking about the challenges in COnformance and looking at the ways success criteria are "conformable" in the context of large, complex, and dynamic websites.
... millions of pages, constantly updates
... user content
... looking at the challenges in hope that there will be a new conformance model for Silver or an alternative model for these contexts.
... the collection of challenges of applying WCAG to this subset of websites fall into 3 categories of challenges:
... the requirement for human involvement

,,, human involvement is much harder when the pages are highly dynamic

scribe: 2) the nature of things that are under construction - there are always parts under construction
... 3) 3rd party -- especially user content. For example, uploading video half a billion updates a day.

Makoto: Japan has a national standard of the Japanese translation of WCAG 2. We did a different conformance model
... pick 40 web pages and then test the selected 40 pages. If the 40 pages meet all the success criteria, I can make a conformance claim.
... we also use screenshots to work around the changing nature of the page.
... we have a time range of the conformance claim. For example, if I test 40 pages today, then I can make a conformance claim. But even if the site changes tomorrow, the conformance claim will not expire.

Cyborg: Silver wants to be more pro-active and Peter's experience seems to be more reactive. How could we use your experience to promote early adoption?

Peter: We just joined Silver, and there is a lot of material to catch up on.
... this has been a topic on our minds, and we understood that Silver is where Conformance is being tackled.
... what I have seen in Silver hasn't been addressing a lot of the big challenges that Amazon has seen.
... we thought our first big contribution would be to help identify those big challenges
... we haven't been able to be "born accessible". We have had to work with where we are.

Cybele: We also want to give an incentive to renovate legacy systems. I'm looking forward to what you are going to be talking about on Friday.
... I want to hear about the fires you are dealing with and how to help them

Peter: You suggested that it would be helpful to have incentives to improve legacy systems. What are the tiers: What do I have to do to get a "C", a "B", an "A"?

Cybele: We have talked about process changes, both outcomes and processes that might not have an immediate outcome
... increased marks in a linear way, is part of the solution, but not the whole solution, We want to get away from the checklist mentality.

Peter: We are in the same place in coming with questions and not detailed thoughts of the answers
... Asking Makoto: How do you pick 40 pages. You run the risk of carefully selecting tuned pages that aren't representative.

Makoto: The number 40 came from UWEM 1.0 methodology in the EU. THey adopted random sampling.
... we decided in Japan to use the random sampling in web.
... we have four options to get 40 pages. If it has less than 100 pages, then every page should be tested. If more than 100, then select 40 pages as you like. It can cause a problem if 40 pages are selected that have been fixed.
... you should include random sampling web pages
... we recommend having 25 pages by random sampling to make the conformance claim more reliable.

Peter: any time you bring random sampling in, you could have user content, and therefore have errors introduced by a 3rd party.

Makoto: Then we use the partial conformance claim from WCAG 2.x

Peter: How do you deal with the "under construction" problem

<pkorn> My audio died (microphone only)

Makoto: Even if it is under construction at the particular time, using screenshots. If the page has a problem, it doesn't pass, but if it does pass, it stays conformant.

<pkorn> You lost my voice, NOT my speaker

Cybele: On the issue of expiring, is that something that could be set to a year (for example)? The example being a company that puts a lot of effort into a website to make it accessible, but couldn't afford to do it every year.

Makoto: We decided that it will depend on the webmasters: If they decide to update it, the standard doesn't care. It is up to the webmasters.
... there are many websites: some update every year, others may make it 8 years and never have updated it. When others see their conformance claim, a newer claim is more trustworthy

<Chuck> so it's more market driven, and doens't really expire

<Chuck> +1 to Jean's understanding

Makoto: It doesn't expire

<janina> I heard that it's up to the organization

Makoto: They have to list a period of testing in the conformance claim, so anyone can see how old or new the conformance claim is.

<Chuck> and I read between the lines that it's market driven.

Cyborg: It's more about transparency than expiration. It's still valid, but people can see the date.

Makoto: THe conformance claim will be valid forever.

Chuck: The conformance claim doesn't expire, but the date is public, so that the newer claim has more currency.

Janina: The date is only part of the story. There is also consideration of how much the site has changed. The older date might be valid if the site hasn't changed
... while a newer site with many changes could be less valid.

<Chuck> That sounds like it is entirely up to the customer to decide based on a static/dynamic site and the age of the conformance claim.

Makoto: The Japanese standards cannot insure the changing state of the web site. In general, we don't care how accessible the web site is today, only when the web site was tested.
... we don't have a legal requirement for accessibility.
... I recommend to my clients to update their conformance claim every 2-3 years, because the older claim won't make sense today.

Chuck: Without the legal aspect, then accessibility is entirely market-drive.

Makoto: People will judge.

Cyborg: the relative nature of these questions. THere are so many challenges in comparing different types of sites. Janina said in addition, that we can compare to itself. We could give incentives for comparing to self, to comparable sites, and to a global standard.

<pkorn> Bye

<Chuck> thank YOU Makoto for taking the time to explain Japan standards to us

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/zakim, this meeting spans midnight//

WARNING: Replacing previous Present list. (Old list: AngelaAccessForAll, jeanne, pkorn_, janina, pkorn)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ 

Present: jeanne janina Chuck Jennison pkorn Makoto Cyborg Joe
No ScribeNick specified.  Guessing ScribeNick: jeanne
Inferring Scribes: jeanne
Found Date: 10 Sep 2019
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.)