W3C

Accessibility Conformance Testing Teleconference

18 Jan 2017

See also: IRC log

Attendees

Present
Alan, Shadi, Charu, Wilco, Moe, MaryJo, Alistair, Kathy
Regrets
Katie_Haritos-Shea, Detlev_Fischer
Chair
Wilco, MaryJo
Scribe
Charu

Contents


Availability survey https://www.w3.org/2002/09/wbs/93339/availability/

Wilco: we have the survey, if you have not please do so

Approval of last week's meeting minutes https://www.w3.org/2017/01/11-wcag-act-minutes.html

<maryjom> https://www.w3.org/2002/09/wbs/93339/ACTTF18Jan2017/results

minutes approved

Draft Section 4.2 Accessibility Support Data https://github.com/w3c/wcag-act/pull/32/files?diff=split

<MoeKraft> https://www.w3.org/2002/09/wbs/93339/ACTTF18Jan2017/

Wilco: how do we define accessibility support?

Moe: is there a defination or a clause in WCAG we can refer to?

<MoeKraft> https://www.w3.org/TR/UNDERSTANDING-WCAG20/conformance.html#uc-accessibility-support-head

<Wilco> Using a technology in a way that is accessibility supported means that it works with assistive technologies (AT) and the accessibility features of operating systems, browsers, and other user agents. Technology features can only be relied upon to conform to WCAG 2.0 success criteria if they are used in a way that is "accessibility supported". Technology features can be used in ways that are not accessibility supported (do not work with assistive technologies[CUT]

Wilco: let me grab the WCAG definition

Shadi: i have not seen any work on the notation of accessibility support

Wilco: i want to standardize the input

Shadi: do we need this at this stage
... we could end having the notation in text later, we may not have to worry about it now

Wilco: you would set up accessibility support to compare 2 tools or implementation of the rule, you have to compare input matches for the same output
... this is tricky, we could set up support function that would check what is supported, the rule might go if aria-describedby is supported on text, a support clause in the rule to support the outcome
... what that means is check if x is supported
... if we have to do it it has to be super light so it can adapt

Alistair: It is very complicated, i would not worry about it

Wilco: the alternative would be to hard code it

Alistair: Over thinking this, blending WCAG and ACT, not a big issue, not writing a light weight support
... We just have to maintain what is useful and what is not
... Allen writes a test after research and i don't have that rule, then i would just take Allans, then we find out it is not accessibility supported then we all dump it collectively

Shadi: i am all for trying to find simpler ways and may be need to look at this later
... the assumption that this is accessibility supported is always there, it could change, it could be supported in one region or with one screen reader but may not be uniform

Alistair: if i adopt WCAG then we have to write the test in the same way, if we pool resources we make the determination of what test

Kathy: Different countries have different standards, based on te content any rule can have compatibility issues
... i agree with Alistair, people will choose what rules to pick and support based on the customer and content

Shadi: Picking and choosing, we have to make some assumptions, we have to agree that people can make selections
... What i understood, are there test written independently and then make the assumption if it is accessibility supported

Alistair: i don't think it is useful, it will be too much to maintain, lot of extra work with no benefit

Shadi: Interesting proposal, 2 tools run on the same site with different assumptions and have different results
... the important question is will the user understand why the results are different
... The more tranperency we can provide the more clear the results will be or reasons why they not same

Alistair: it is common sense to select the same tests for same results

Wilco: fasinating discussion, we have to make some meta data available on which rules are accessibility supported, if that has to be explicit or can be in the rules description

Alistair: we write the test in plain english and not say anything about accessibility supported, has that will change

Wilco: That was what i am trying to get at, it can be implicit

Moe: We can identify the clause for accessibility supported data and rule can be independent of that

Wilco: that sounding like leave it implicit and not explicitly state

Moe: Can you give me an example
... may be i am misunderstanding

Wilco: Input field have label rule, implicit as well as explicit label and certain AT supports one

Wilco: it has to be defined in the baseline
... the code is being tested

Moe: if the client supports a configuration

Wilco: We can not outline what user agent or what AT will the rule supports

<MoeKraft> I think the rule should be independent of the assistive technologies.

Shadi: The goal is to have all tools interpret standards same

<MoeKraft> We are trying to confirm that the code meets success criteria and not validate the AT.

<MoeKraft> Sorry, need to drop.

Shadi: one could write a test can raise an issue on a code and someone else will write a rule that passes based on their interpretation
... the end user will not know the difference, how can we make that more transparent

Alistair: the results can show why one fails one the other does not

Shadi: the report will say 2 tools ran and one is failing and other is passing, it will not be clear

Wilco: i would say if it is different rule then it will show different results
... 3 options, we say, you must explicitly define what accessibility supports the rule, or we can make it optional and let it be implicit
... last option is don't include accessibility support

Alistair: lot of overhead to optionaly define

Wilco: input text has a label, what are the accessibility supported assumptions, implicit labels, explicit labels, aria-label ..etc

Shadi: another use case is, if there is a different way to test, rule, description, assumptions, steps, expected results
... what if someone has different procedure or steps to test, so that has to be defined

<shadi> +1 to specific examples!

Wilco: lets work this through via email and end it here
... i want to mention F2F at CSUN
... like to hear, do we want to meet or not and the second thing is where? we need a host
... We have talked to IBM and Deque, with no luck
... Alistair, can you explore if we can host, or any other ideas for half day or full day meeting
... Jumping toopen issues

not sure what happened to my audio

still trying to understand what we want to include in the section of exception

Wilco: thanks everyone, talk soon

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.148 (CVS log)
$Date: 2017/01/18 18:21:13 $