<scribe> scribe: sajkaj

Amend Representative Sampling proposal with language for transparency

<pkorn_> Is the zoom meeting up? I appear to be alone...

<pkorn_> never mind. wrong xoom room

kd: Concerned to be clear about transparency and how this fits

sl: Didn't want to enforce documenting how tests are done because this can be nearly impossible to actually document

kd: More worried about scoping what's actually tests and what conformance one is claiming

sl: Trying to map that to representative testing ...

<JF> Present÷

sl: We could say "we did representative testing"

kd: Don't think we should super detailed, but thinking how this would work at my org
... we have all kinds of case law ...
... we might say we tested a Federal, state, case ...

sl: a statement of how representative sampling was done as opposed to how testing was done

js: makes sense to me
... how testing is done is often proprietary
... saying how testing is selected is perhaps the transparency we want

kd: it's as far as I got

<JF> Non-proprietary testing = ACT Rules format

sl: I'm thinking of products with dependencies built on different frameworks ...
... It would be one thing if we all switched to a specific kind of conformance claim, but without that ...

js: Any language to propose?

kd: Asks for the link of where this goes ...

<jeanne> February doc: https://w3c.github.io/silver/guidelines/

<jeanne> https://w3c.github.io/silver/guidelines/#sampling

kd: Belongs somewhere here?

js: Believe it replaces the entire section
... What we currently have there probably moves to another doc, best practice, soimething like that

sl: something like wcag-em
... or an update to wcag-em

js: Looking for volunteers to help with this ??
... I volunteer!
... Anything else on this topic?

[crickets]

Survey results of Conformance Scope

<jeanne> https://www.w3.org/2002/09/wbs/94845/2020-06_Conformance_Scope/results

js: Should be public
... 16 responses so far; 7 yes, 5 yes with changes, and 4 no
... Goes through comments

sl: Notes he copied current definition but used "path" instead of "web page"

<Zakim> sajkaj, you wanted to discuss Judy's comment about scope

<jeanne> +1 sajkaj that we are focusing on doing accessibility right

<Lauriat> +1

<CharlesHall> side note: I don’t see my name in the non-responders list, and sorry, I missed the second email with the link. :/

<JF> -1 to anything "fuzzy"

sj: Concerned about Judy's comment that it is possible to claim so little that it isn't "substantial" or meaningful

mc: Suggest we move to recording how we dispose of this kind of issue

[discussion of comments and how to handle disposition]

mc: suggest we discuss further later

[now looking at comments in order]

<jeanne> JSa: There are three atomic steps to a path. You have to access the path, interact with the path, and then exit the path

sj: suggests there are 3 minimal steps to anyu path, getting to the object/page, interacting appropriately, leaving gracefully (as opposed to crashing out)

sl: do have a question for jf on that ...
... "only" there only because I copied from 2.x

<JF> I cannot unmute

sl: want to know whether our consideration that this supports single page still ?

jf: we have stand alone reqs now in 2.2
... e.g. strobing gif on one screen of several that isn't part of your path but interferes with it

sl: have been thinking about that ... we'd need to define that when we come up with tests
... if all those examples are included in scope definitions? does that satisfy "only" concern?

jf: would want to see lang
... need to acknowledge out of path impediments that affect accomplishing the path

<CharlesHall> that sounds like on same page as path has non-interference issues to be in same scope

sl: but I'm suggesting we acknowledge in the scoping of path

jf: how?

sl: same way noninterference works today
... flashing today --- however it occurs when the user accesses content, it counts

<Rachael> +1 to needing a defintion of path.

sl: saying anything that affects th user is included, even when it's out of the path on its own

<CharlesHall> in addition to a definition of path, we may need more than one. again, in ux, there is a happy path and an actual path.

sl: the "how" will per guideline, per context level
... it should be clear that it's included in scoping req

jf: so object to "only"

<Rachael> I don't think "only" is needed

<CharlesHall> depends on the definition of path.

sj: thought we were surveying the concept, not as a full definition

sl: agrees

<bruce_bailey> Here is wording w/o "only":

js: notes comments on IRC that more definition is needed, examples, etc

<bruce_bailey> Conformance is defined for Paths. However, a conformance claim may be made to cover one path, a series of paths, or multiple related paths.

[no objection to dropping "only"]

bb: it's still an affirmative statement -- dropping "only" is fair

<AngelaAccessForAll> +1

js: asks any objections?

sl: agree we need to wordsmith, but is the concept OK?
... Do want to hear more about scaling, because I'm very concerned we scale well

jf: speaks of happy path vs multiple paths
... you can do it one accessible way so never mind all the remaining inaccessible paths?
... that's my scale problem

<CharlesHall> +1 to JF. if 1 path, then all paths.

jf: points to search and sitemap

sl: not sure where this specificity happens, but agree that it needs specification
... multiple ways of getting to a page, accomplishing things, will be challenging but we need to cover

francis: agree with sl -- multiple ways to get at anything

<Lauriat> +1 to JF on making sure we don't break multiple ways requirement

francis: agreeing we don't worry about enforcement -- we're not the polic/court/regulator
... we can't polic, but we need to define well

js: continues with comments ... many more of the same ...
... or similar

sl: would be good to followup with him, I don't quite grok the comments

<KimD> +1 to defining "path" and BB's comments

js: looks at bb's comment asking for definition -- and cap vs lower case

djs: takes up df's comment ...

df: concerned it's not a good fit for many sites/content

<JF> +1

df: concerned the path model doesn't well accomodate lots of possible subpaths

<CharlesHall> I think ‘states’ can be included in the definition(s) of ‘paths’

sl: definitely agree with the one way vs the many problem

<CharlesHall> for reference and fun: https://en.wikipedia.org/wiki/Happy_path

<Lauriat> Don't break: https://www.w3.org/TR/WCAG21/#multiple-ways

jf: would be easy to argue reviews of products are supplamentary to ecommerce model
... we need to be concerned with all that's on screen -- not saying pass/fail -- just consider all the flows

js: moves on in comments ...
... now at Judy's comment ...
... notes jb says doesn't match current draft scope ...

sl: yes, draft was meant to replace that

js: should clarify this is a replacement

js notes jb rationale to apply to path; but not why path needs to be more inclusive

sl: yes, we need more comprehensive examples

<Lauriat> …more comprehensive examples that demonstrate how the "path" concept does not exclude any kind of content.

js: now the comment about substantially limited how much of a site is covered by claims

jf: agree accomplishing things is big part of web
... but ... makes it easy to game a report

sl: points out that we're not trying to prevent lies, that's not our remit

<CharlesHall> while I think the “reducing incentives” comment is subjective, that can be a specific testing goal to conduct public survey and usability tests on scoping.

jf: not all the truth isn't that

sl: then you have transparency of the scope

jf: regulators will need to weigh in on scoping
... orgs will want a11y dashboard health reporting

sl: we allow that
... points out 2.x doesn't require, so why does silver?

jf: because we're trying to improve

sl: defining a complete thing is impossible for us
... we can give people the building blocks

<KimD> I think this is tied to the transparency info - that's our due diligence

jf: guess i'll wait for examples

sl: that's fair we need examples

<Rachael> https://docs.google.com/presentation/d/1IceTYOyGitApczya4vat4gPk9_I-7hIwpqTtXGusEZk/edit#slide=id.g8bebb5a093_2_11

This is scribe.perl Revision of Date 
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/sear h/search/
Present: jeanne Francis_Storr OmarBonilla Lauriat CharlesHall MichaelC KimD AngelaAccessForAll bruce_bailey Rachael
Regrets: Jan Shari Peter
Found Scribe: sajkaj
Inferring ScribeNick: 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.)