introductions Joe Cronin - amazon
https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan
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.)[End of scribe.perl diagnostic output]