<MichaelC> Hi everyone - zoom seems to have a widespread outage, we´re trying to find alternatives
<Rachael> Thank you for letting us know.
<Rachael> While we wait, it would help if everyone pre-reads the revised PRs for issue severity and equity. They are linked from the agenda at https://www.w3.org/WAI/GL/wiki/Meetings/TPAC_2022#Thursday.2C_15_September_2022
<MichaelC> let´s use Google meet
<MichaelC> To join the video meeting, click this link: https://meet.google.com/mkm-jwaa-dvf
<MichaelC> Otherwise, to join by phone, dial +1 613-916-7313 and enter this PIN: 897 737 467#
<MichaelC> To view more phone numbers, click this link: https://tel.meet/mkm-jwaa-dvf?hs=5
<MichaelC> To view more phone numbers, click this link: https://tel.meet/mkm-jwaa-dvf?hs=5
<MichaelC> To join the video meeting, click this link: https://meet.google.com/mkm-jwaa-dvf
<Rachael> While we wait, it would help if everyone pre-reads the revised PRs for issue severity and equity. They are linked from the agenda at https://www.w3.org/WAI/GL/wiki/Meetings/TPAC_2022#Thursday.2C_15_September_2022
<Ben_Tillyer> alastairc, we are in google meet
<Rachael> scribe: Poornima
<MichaelC> To join the video meeting, click this link: https://meet.google.com/mkm-jwaa-dvf
<Jennie> * 2nd scribe in 45 minutes or next session?
<Jennie> in 45 minutes, I am in!
<Rachael> https://www.w3.org/WAI/GL/wiki/Meetings/TPAC_2022#Thursday.2C_15_September_2022
Rachel: check the links if you have not attended the previous meetings
<AWK> +AWK
Rachel: We've got pull requests, suggestions, survey, so I moved the content to the doc. We'll not be doing maturing levels in the doc itself
<Rachael> Document version: https://rawgit.com/w3c/silver/issue-severity-main/guidelines/index.html#critical-errors
Rachel: People can review the document while we are in the meeting. https://rawgit.com/w3c/silver/issue-severity-main/guidelines/index.html#critical-errors
<Jennie> Captions are working for me
<maryjom> FYI, in the editor's note for critical errors there's a misspelling of "prioritization".
Rachel: Not necessarily making a decision until we come to a point of comfort, but open for discussion on questions or thoughts
<jaunita_george_> +1
<Lauriat> +1
<jeanne> +1 to moving it to WCAG3 - better fit than Requirements
Wilco: critical errors are non
normative, test completion is kind of dependent of defining
tests, both of these seem challenging for conformance
modal
... it's okay if this is non normative
<Wilco> Suggestion for the note: Whether or not critical errors will be defined in normative text, and so can be used in the conformance requirements needs further discussion.
<jaunita_george_> +1
<Rachael> https://rawgit.com/w3c/silver/issue-severity-main/guidelines/index.html#critical-errors
<Lauriat> I can share!
AWK: it would help to share what we're looking at. It'd be great if W3C schedule aligns with the discussion schedule.
Rachel: reading the 'Critical errors' section. People in the Issue Severity can capture the suggestion and others can ask the question
<Rachael> Poornima: What does it mean the A/AA/AAA at the test level
<Ben_Tillyer> alastairc, search doc for "prioritizatoin" if you can't see it
Cybele: How we are not replicating by changing the language for critical vs non-critical?
Rachel: My understanding is conceptually it don't have the test level, we can remove that.
<Zakim> Lauriat, you wanted to suggest maybe "This is a potential way to replace how A/AA/AAA levels represented severity by incorporating a mechanism to express severity as a part of
<AWK> maybe instead of "express" - "evaluate"?
Michael: The aspect of severity A/AA/AAA from the past can be rephrased to Severity level in a little different way
<Cyborg> is what is critical for some more or less critical for others? is there a subjectivity to what is critical or not?
Alastair: For each of these tests, categorizing both going to determine equity but giving a better measure of equity
<Cyborg> definition would help +1 to critical definition
Michael: The section says 'critical errors', but the definition is not included there. like to see the definition
<Ryladog> +1
<Ben_Tillyer> "An accessibility problem that will stop a user from being able to complete a process." from https://www.w3.org/TR/wcag-3.0/#dfn-critical-error
Rachel: doesn't have to approve today, as the definition needed. What survey is needed for future work?
Ben: It's already read in the document link
Rachel: It's from the public working draft, can that be used?
Jonathan: Critical issues are almost a matter, especially for multiple questions, there is a broader connection for overall scoring.
<Zakim> alastairc, you wanted to suggest Critical issue: An accessibility problem that is likely to stop a user from being able to complete a process.
Rachel: Probably we could add Jon's point as a question
<Cyborg> are there non-critical issues that in frequency becomes overwhelming? Non-critical friction over time can prevent access.
Alastair: we are trying to more
severity level rating that's below A/AA/AAA below the guideline
level
... Critical issues are the issues that completely stops the
user from continuing..
<jon_avila> Critical issue is good with me.
<Rachael> strawpoll: 1) critical error 2) critical failure 3) critical issue
Alastair: I feel critical issues are like critical errors
<Amanda> 3
<jon_avila> 3
<JakeAbma> 3
<Wilco> 3
<ShawnT> 3
3
<Rachael> 3
<Ben_Tillyer> 3
<Lauriat> 1,2,3 - fine with any
<alastairc> 3, don't like 2 as it crosses over with pass/fail.
<jeanne> any, lean toward 3
<AWK> 3
<Ryladog> 1
<maryjom> 1
<laura> 1
<ShawnT> +1 to alastairc
<Zakim> Lauriat, you wanted to suggest we frame things in terms of something about issue severity or impact to users, rather than critical vs. not (even if we end up with similar concepts
<jaunita_george_> 3
Rachel: 3. Critical Issue is strongly preferred, so we can make that wording change
<Cyborg> +1 to impact to users
Shawn: Just a suggestion, with the conversation around the critical vs non-critical issues, some kind of priority around users, positive impact to the users..
cybele: I have a concern on #5 - the use of Bronze, Silver, Gold will be tied to, but is the less representation of critical vs non-critical
<jon_avila> That's how I assumed it would be.
cybele: also the use of Bronze, is still problematic, but with A/AA/AAA many don't feel to be the Gold
Wilco: We know how to define tasks, it's still a question to figure it out, am not sure this advances where we are now..
<Cyborg> +1 to Wilco
<Zakim> Rachael, you wanted to go back to straw poll, "crtical barrier" and the 1 votes
Wilco: am hesitant on how this proposal to be framed
<Zakim> alastairc, you wanted to comment on the topic of bronze/silver etc.
Rachel: I'm not disregarding what Wilco said, just looking at people votes
Alaistair: Not having such a
blunt instrument in measuring the a11y complex interfaces
through testing, that probably get into people way...
... that's the problem, the goal essentially is to some form of
passing that allows levels to get into people way for testing..
it doesn't have to involve bronze, silver, gold..
... we can get agreement on which are the critical issues, we
need to explore more on reasonable confidence in creating the
matrix of critical issues
https://meet.google.com/mkm-jwaa-dvf?pli=1
Ryladog: one is providing the opportunity for companies easier way to get one over other, other is for a11y companies, say banking how they will use..
in wcag 2, the conformance requirements are A/AA.. organization are generally going to be using some level of bug tracking and ranking
scribe: Going to Question 1, saying 'Critical error' is important
Cybele: the items are possible defined as critical or non-critical or low value, there may be spectrum of categories that are not identified
<alastairc> We've reviewed it twice, the second time earlier this week.
Cybele: is there a item that's least critical / unimportant? could we identify those and then give that category as some kind of exception and say these are not being addressed at all
Alastairc: If you are pre-defining the severity at the test level, that's more of a process approach like silver/gold level for the evaluation
<Cyborg> +1 to Process Evaluation
<Rachael> https://docs.google.com/document/d/1OprBE6Lqb1iSKx_q-33fC-jH-Xh3YyTYAmeTLyp5j9I/edit#
<Cyborg> can we please add discussion of process evaluation to next steps for discussion?
<alastairc> it is, we already went through this on tuesday
Rachel: For people have questions, add that to parking lot for later discussion and voting
<jeanne> Equity proposal update https://rawgit.com/w3c/silver/equity-proposal/requirements/index.html#equity
<alastairc> NB: I saved changes to the issue severity PR.
<Rachael> https://rawgit.com/w3c/silver/equity-proposal/requirements/index.html#equity
<Jennie> * Ready to switch Poornima?
yes
<Jennie> scribe: Jennie
Jeanne: I tried to address the
more editorial things
... I started by removing the questions from the 1st editors
note and moved them to the end
... I put in the statements we talked about
... Next I moved the scope up to the 1st of the actual
paragraphs that is not an editor's note
... The primary scope is to address equity for persons with
disabilities
<scribe> ...(continues reading)
UNKNOWN_SPEAKER: What we are
doing is coordinating with the working group working on the
CEPC
... We have our writings in synch
... Next - the definition of equity
... Then, part of that is the 3 different types we discussed on
Monday
... Finally, I have the descriptive paragraphs about
equity
... and the equity centered progress
... I didn't make any changes to them
... I also taked notes from the meeting minutes
... I categorized them, and took a lot of the suggestions from
the survey
... I added a couple to include the groups of questions
... These include: how can we recruit more diverse people with
disabilities
... How can equity issues align better with civil
legislation
... How might the principles of equity be raised
proactively?
... (scribe missed a few)
Rachael: What has shifted from the survey - this is content that would be added
<AWK> Seems like "W3C guideline technical report" is an odd phrase - maybe remove "guideline"?
<Zakim> Ben_Tillyer, you wanted to say there is a small edit I'd like to see
Rachael: I am opening the cue for suggestions and thoughts
<Ben_Tillyer> The first sentence states the primary scope is equity for persons with disabilities.
<Ben_Tillyer> The paragraph then goes onto say we must also consider other parts of the human experience.
<Ben_Tillyer> One of the "Other parts" of the human experience mentioned is "disability (both visible and invisible)" which is surely a duplication and should be removed as that is not one of the "other parts", it is the previously stated primary scope.
<AWK> +1
<Rachael> +1
<jeanne> +1. Will fix
<jon_avila> We should frame in terms of intersection
<alastairc> +1, and I'm editing the PR now
<Ryladog> +1
<JakeAbma> +1
JenniferS: The tricky part is that the long list of identity characteristics come from the CEPC
<laura> +1
JenniferS: So Michael may have something to say
Jon_Avila: Seems like we are trying to solve the intersection between disabilities and these issues - how they can be compounded
Rachael: Is there a wording change you would suggestion?
Jon_Avila: Probably using "intersection" so people don't get the impression we are trying to solve them in and of themselves
<Lauriat> +1 to highlighting intersectionality
Rachael: Is there support for adding this?
<AWK> sure
<maryjom> +1
<alastairc> "... consider the spectrum of human experience and how it intersects with disability" ?
<David-Clarke> +1
<jenniferS> consider the intersection of spectrum
<ShawnT> +1
<jenniferS> +1
<jaunita_george_> +1
Rachael: Is anyone considered with making this clarification?
<Rachael> +1
<JakeAbma> +1
<Ryladog> +1
<Ben_Tillyer> +1 to alastairc
Rachael: (reads Alastair's suggestion)
<jon_avila> +1
<laura> +1
Rachael: There is general support for that
<Poornima> +1 to alastairc
<jenniferS> consider the intersection of the spectrum……
<jeanne> This includes but is not limited to intersections of socio-economic status, sexual orientation,etc.
Rachael: Are there other concerns or changes that are needed to this?
Ben: Do we need to move mental health into the primary scope as well?
Rachael: Good question. I will defer that to the subgroup
<Rachael> ack. jenniferS
JenniferS: I want to circle back
to something Andrew raised earlier this week
... About whether we should address equity in AGWG or if this
is something that should go higher to W3C as a general
thing
<AWK> Trying to remember...
JenniferS: If that is something we will discuss today, I think it is critical that AGWG discuss this because of the barriers people with disabilities face
<AWK> I do think that AGWG should, but I also think that the W3C should have overarching guidance for TR specs
JenniferS: In response to Ben - does AGWG tend to separate mental health from disability?
<Zakim> jeanne, you wanted to say that there are advantages to staying coordinated with CEPC
Jeanne: I think there are large
advantages to stay coordinated with the CEPC list.
... If we make changes, I think we should actually do as an
issue with CEPC so that we don't diverge
<Zakim> Rachael, you wanted to answer Jennifer
Rachael: Going back to Jennifer,
chair hat on
... We voted and agreed to have a group to start that
conversation
<jenniferS> Thank you!
Rachael: This will apply to what applies to AGWG and what is broader, so we won't revisit today unless it relates to this particular content
<AWK> I would say that AGWG might view mental health as causal for some types of disabilities, but not that mental health is "covered"
Cybele: I want to address the
definition question
... and +1 to coordinating with the CEPC
... Especially if they are amenable to some coordination
... My hesitation to removing mental health is that we need to
start addressing questions like neurotype
<Ben_Tillyer> Looks like it is already out of sync with CEPC. I can see gender expression and neurotype on that document, but not in our list.
Cybele: Within disability circles cognitive disabilities and mental health related are often neglected
<Ben_Tillyer> Ignore me, it's in a different order
Cybele: There are other groups also marginalized
<AWK> Gender expression is in our list
<jeanne> Ben, Janina has been filing issues with CEPC and they are discussing and adding in their meetins. When we are finished this week, we will update
Cybele: Coordinating with the group will help ensure the inclusion of the full spectrum of human experience
Rachael: I am trying to understand which way we would like to go
<alastairc> It seems reasonable for us to focus on disabilities
<jenniferS> +1 to Rachael as that intersection of disabilities with other disabilities is a good point.
<jaunita_george_> +1
Cybele: My 1st instinct: possibly we might consider expanding through CEPC that 1st line
<Lauriat> +1 just keeping it and having them represented in both places
Cybele: To think about how we
want to address diversity within disability too
... And if addressed, in a 2nd sentence, that this may impact
the whole list
<jeanne> +1 to keeping in both places
Cybele: And I think this is coordinated conversation with CEPC
Rachael: I am seeing +1s to keeping it in both locations
Jeanne: Could we do a straw poll that says we stay coordinated with CEPC
<Rachael> straw poll: Should we stay coordinated with CEPC?
Jeanne: Then when they are finished (if this passes) we can edit when we are finished with this week
<Lauriat> +1
<Cyborg> +1
<David-Clarke> +1
<Ben_Tillyer> +1
<jenniferS> +1
<jeanne> +1
<ShawnT> +1
<maryjom> +1
<laura> +1
<JakeAbma> +1
<Rachael> +1
<jaunita_george_> +1
<Amanda> +1
<Ryladog> +1
<jon_avila> +1
<alastairc> I don't mind, it just seem odd to say "for persons with disabilities" and then include that in the big list.
Wilco: I am not clear what stay coordinated means. What happens if they don't make changes?
Rachael: I believe it means the list portion that says "this includes..."
<AWK> +1
Wilco: If CEPC goes a direction we don't like, we are still going to go with ti?
Rachael: Do you want us to discuss each time?
<jeanne> So far, they have accepted all our suggestions
<Cyborg> suggestion: staying coordinated to me means we have some capacity to influence that as we need, the coordination goes both ways
Rachael: Right now if we noticed an update we would update this list
<Cyborg> intention makes sense to me
<Wilco> 0, I very much don't mind either direction
<Cyborg> +1 to intention
Rachael: But we could also say we bring it back to the group
David: I think it is also worth
actually referencing the CEPC document
... Or, we are not
... And link directly to it
JenniferS: I think I read it
differently
... Maybe it requires tone
... (reads)
... This, refers to the spectrum of human experience
... We could say "the spectrum includes but is not limited
to"
Rachael: Are you suggesting we make the change, then not update the text?
JenniferS: Correct. If we change
to "the spectrum" we would be fine
... We already have a pull request for the list of
characteristics, and it has been approved
... So our list is represented within CEPC
Rachael: OK, thank you
... I will do a couple of polls
<Rachael> straw poll: Add a link to CEPC in the text
<alastairc> Is it drawn from the first bullet here? https://www.w3.org/Consortium/cepc/#unacceptablebehavior
<David-Clarke> +1
<Ben_Tillyer> +1
<jeanne> +1
<maryjom> +1
<AWK> +1
<Cyborg> +1 but i like the "intention" word added in
<JakeAbma> +1
<ShawnT> +1
<Ryladog> +1
<jaunita_george_> +1
<Cyborg> intention to stay coordinated with...
Rachael: I will get to the word intention in a moment
<Ben_Tillyer> AWK is typing
<laura> +1
<AWK> I agree with the intent of including what the CEPC is stating, but the CEPC doesn't seem to be targeted to specs - is there a change planned?
<AWK> Seems like the CEPC focus is on PWE within the W3C WG's
Rachael: Not that I know of, it is more around referencing the list
<alastairc> Suggested text: This list draws from the _W3C's_CEPC_, and we intend to stay aligned with the CEPC.
Rachael: (reads Alastair's suggestion)
<jaunita_george_> +1
<jeanne> +1 Alastair
Rachael: any word smithing anyone wants to do?
<David-Clarke> +1
<laura> +1
<jenniferS> +1 as I think that is what Michael Cooper wanted us to do
<Ryladog> +1
<Rachael> Jennie: I woudl support if CEPC is written out fully the first time its used and has a link
<jeanne> +1 Jennie
Rachael: OK
<Rachael> +1
<alastairc> This list draws from the <a href="https://www.w3.org/Consortium/cepc/#unacceptablebehavior">W3C's Code of Ethics and Professional Conduct</a> (CEPC), and we intend to stay aligned with the CEPC.
Rachael: Anyone object to
Alastair's suggestion, plus the link
... OK, no objections so let's add that
... We have surveyed this once
<jenniferS> Here is the PR for the CEPC, for the record: https://github.com/w3c/PWETF/pull/209#pullrequestreview-1103949227
Rachael: Do we want to read this
and survey for a future meeting or vote?
... Who wants a chance to read it again before voting?
<Cyborg> +1
<Poornima> I am
<AWK> It is exploratory so can be changed
Rachael: OK, then we will postpone a final decision until we survey a second time
<AWK> no?
<Cyborg> oh i didn't mean postpone
Rachael: No, not exploratory
<Cyborg> sorry i was confused
<Cyborg> -1
<AWK> got it
Rachael: It is in the requirements document
<jenniferS> -1
Wilco: Process question
... This seems like the type of thing we might need a CFC
for?
Rachael: It is an informative
document
... I defer to Michael and Alastair
<jenniferS> Do we do a CFC in this stage of the maturity model process?
<Ben_Tillyer> MichaelC is not in the room
<Cyborg> +1 to voting
<AWK> +1 to Wilco - A requirements doc change should be CFC'd
Alastair: It is a TR
document
... We would have to come back for a survey, then come back for
CFC
... But we do need a survey first
Rachael: Thank you
... Yes, we would CFC it
Wilco: I think it would be good to survey it
Rachael: OK, that is what we will do
Cybele: If I have additional suggestions, where would I do that?
<jenniferS> Reviewing the Editor's Draft Maturity Levels — it would be helpful to itemize what survey / CFC requirements are needed for Placeholder, Exploratory, Developing, etc.
<Cyborg> We invite, on an ongoing basis, the identification of equity-based concerns resulting from and amplified by WCAG's application, which will then be addressed within an equity-based process.
Rachael You could put it in the survey or paste it now
Cybele: The paste above is a suggestion for the paragraph on structural equity at the end of the paragraph on structural equity
<jenniferS> Ah, additional pages…I see it in the details.
Rachael: chair hat off, this
doesn't feel like the right place for it for me
... I am not against the concept, just not sure where it will
go
... Do others have thoughts on integrating this?
Alastair: For invitations of that nature they tend to go out in calls for review, or potentially the introduction to the spec itself
<Rachael> akc jenniferS
JenniferS: I am trying to
empathize and imagine the use case
... If someone reads this and wants to contribute, they are
welcome to
<alastairc> the requirements are mostly for us, if we want contributors from outside the group, the calls for review would be the best place.
JenniferS: This can be difficult for people to understand
Cybele: This is in alignment for
what you suggested
... If we are inviting equity converations then there needs to
be a challenge for those to raise use cases
... to determine if they are in scope
Rachael: I will suggest that for
the parking lot
... Cybele - you are welcome to add that in the survey as
well
... Thank you for raising that
<Cyborg> thanks and thanks Jennifer for saying that
Rachael: I am moving to slide
3
... To set the stage for the day's conversations which are on
how we want to build WCAG 3
... We started a conversation on tests and severity
... There were requests for a dive into that
... We will spend all of today talking about that level
... We will then move on to a conversation about the
conformance approach
... Or approaches that we feel comfortable exploring then we
will talk about level
... I would like to avoid conversations about levels until we
have worked through issue severity
... Levels are an equity concern
... There are a number of ways to divide them
... And sometimes we get stuck. We will try to avoid them
... Normative vs informative will be after TPAC, as well as
writing content and organizing content
... Are there questions on that?
Shadi: Are you going through the slides we didn't finish on Tuesday?
Rachael: If the group feels it is useful we can put it in later today
<alastairc> My memory doesn't go back that far, which ones?
<AWK> Test Types
Rachael: The conformance architecture definitions, and regulatory slides
<AWK> No preference, as long as we do at some point
<Rachael> Straw poll: Do you want to revisit the Conformnace,Accessibility, and Regulatory slides https://docs.google.com/presentation/d/1Hp460oEvI6dkCQJ3XsLDkvrGsnh9Ub8O/edit#slide=id.g13957fdd223_11_21
Rachael: There was some great after meeting conversation
<Rachael> Strawpoll: Reivisit today or revisit after TPAC
<jaunita_george_> either
<AWK> today or after TPAC - either
<jeanne> after tpac
<alastairc> I don't think it's required for understanding todays pres, happy to do later
<maryjom> I'll leave it to the chairs when its best to revisit. Since you already have your schedule.
<Wilco> After, I guess. Not sure what more there is to go through though
<Lauriat> 0
Rachael: There has been discussion on the list about defining things and ironing them out
<Ryladog> 0
Rachael: The goals was to raise
people's awareness and better understand it
... Going on to slide 4
... Silver did a number of years of work trying out different
approaches, prototyping, and other work
... We looked at all the things discussed
... We came up with 5 approaches
... WCAG 2 approach: everything passes or fails
... It's current status: We are discussing keeping the testing
level at pass or fail, but are exploring other options
... We know it has limitations
... The 2nd approach: percentage based, and in the 1st public
working draft
... Current state: level of effort is too high
... It is viable, but not practical
... The chairs suggestion: throw this out, but we will staw
poll this later
... Option 3: points
... John's proposal is an example of using points
... This is still in "possible, for exploration"
... Option 4: adjectival
... This came from Bruce
... This is still possible for exploration
<Cyborg> sorry what is adjectival scoring?
Rachael: Option 5: evaluating
severity and context
... This was a little implemented in the working draft
... This is also possible for exploration
<alastairc> https://docs.google.com/presentation/d/1dV1moNnq-56sS1o84UCKkc_g-gE10X6Y/edit#slide=id.p1
<jeanne> Cybele, it is poor,
fair, good, best ...Adjectival: you create some
level of categories, assign passing at that level
... The concept of using words to grade conformance instead of
trying to do it numerically
<alastairc> broader than percentages, but similar idea ...Adjectival: Those categories would have a series of requirements to help people place the content in those categories
<Zakim> Amanda, you wanted to discuss Shadi
<Cyborg> yes - my proposal?
Rachael: My first question - have
we missed one we should have on this list?
... I can go to slide 5
<Cyborg> Rachael - i resent my proposal - does that fit here?
Rachael: These overlap with those
different approaches
... 1 is prioritizing by functional needs
... 1 is requiring a minimum score
... and 1 is weighting
... Waiting was prototyped and examined. The Silver group
revisited several times and agreed this is not viable
Wilco: I am wondering if this is
the thing we want outcomes to have?
... Or, is that different from what we are doing?
Rachael: I am going back to slide
4
... The short answer: having experimented, from my perspective,
all of these can apply at any level
... Outcome, overall conformance
... And we have tried it at different levels
<Cyborg> maybe i'll just repost it here...
<Rachael> Group decided to explore keeping the test level at pass/fail.
<Rachael> These can apply at outcome or overall level.
Cybele: I don't know what
happened. I resent the proposal that brings equity into
this
... But it doesn't seem to be reflected here
Rachael: This is not every proposal we received, it is an aggregate
Cybele: This is dramatically
different, and I resent it, and it may not have been received
in tim
... time, I will readd it in IRC
Rachael: I will review during the
break
... These are the possibilities on the table
... Not a specific proposal
... If we have missed something it would be good to know
<Cyborg> Different conformance proposal: https://docs.google.com/document/d/1ccKlaPMaVvazbSqMPgttvMesy9D0KAjGY01pAQES2K0/edit#heading=h.8v4ivoroyy0s This is the accompanying spreadsheet (from page 12 of that Google doc): https://docs.google.com/spreadsheets/d/1gW1PPI5pQ-yUNxR16XbLiPxTJzrBjqB3EroLfmLfe0s/edit#gid=1455183688
Rachael: When we come back from
break I would like to get a formal decision that we are not
going to pursue percentages or waiting
... If we are not, then I want to identify that we are keeping
those on the table ...Reminder: the reasons why
are on slide 6
... I will ask Jeanne to talk to this slide before we get
started again
Wilco: I think that this is not only about outcomes - they are not mutually exclusive and I am not sure we should rule any of them out ...Example: if something works best as a percentage, then we shouldn't rule that out
*Note: current scribe needs to drop in 1 minute
Rachael: Why don't we discuss that when we come back?
Wilco: OK
<AWK> can we switch to zoom?
<AWK> Michaell not here
<Ben_Tillyer> Michael isnt in the room
<Ben_Tillyer> +1 for zoom
* I will be back after lunch
<Cyborg> do we have the zoom link?
<Ben_Tillyer> Michael is back
<Cyborg> thanks
<AWK> +AWK
<alastairc> scribe?
<alastairc> scribe: alastairc
Rachael: There were factors we discussed previously, with some background. [missed a bit]
<Zakim> jeanne, you wanted to say that using percentage for everything didn't work. That doesn't take it off the table for specific tesets
Ben_Tillyer: Rather than removing an approach due to difficulty, should we keep multiple ones which fit them best?
<Wilco> That sounds like something I've heard pass
jeanne: When we look at something
as an approach, we agreed that using percentages to measure
everything got a lot of negative feedback.
... so we're talking about an overall approach rather than
individual tests
<Rachael> ack +
jeanne: so we could remove those at the overall level, but keep at the test level.
<shadi_> +1 to Jeanne
<wendyreid> scribe+
<wendyreid> Cyborg: I have questions about whether one of these might be taken off the table
<wendyreid> ... what is the rubric for doing that
<wendyreid> ... a factor of consistency, does it work across large or smal scale organizations
<wendyreid> ... does it need to apply consistently
<wendyreid> ... level, does it work at individual or aggregate levels
<wendyreid> ... impact or task based
<wendyreid> ... equity, wouldit support a culture shift
<wendyreid> ... is it worth developing a rubric to determine which of these is best
<Zakim> AWK, you wanted to say that we need to figure out what our end goal is to decide this perhaps?
<wendyreid> ... what is the process for determining what we take off the table
<wendyreid> AWK: Part of the discussion on tuesday afternoon, we discussed regulators vs other motivations
<wendyreid> ... how to define what is in WCAG
<wendyreid> ... if we are deciding on levels, regulators are going to look at that and want to adopt one
<wendyreid> ... if we find out that regulators preferred a sliding scale, 50% is the bare minimum, and goes up to 100
<wendyreid> ... different orgs or countries may adopt different approaches
<wendyreid> ... if we had better clarity on the end goal, it might make it easier to decide how to reach that
<wendyreid> Wilco: There has been a discussion of multiple conformance models in WCAG, that was brief
<wendyreid> ... I am with Andrew on reframing this conversation
<wendyreid> ... what are the problems we are trying to solve
<wendyreid> ... what are we trying to use this for?
<wendyreid> ... why do we need any of these, or anything more complicated than pass/fail
AWK's end goal, suggestion: A way to measure accessibility that aligned with the lived experience and is compatible for use by organisations and regulators.
<wendyreid> David-Clarke: I would say that if the scoring system or whatever gets rejected, if its eliminated with a reason
<wendyreid> ... if we don't have one, it might be re-suggested later on by people who didn't have the history
<wendyreid> mbgower_: Something that got mentioned in A11y at the Edge by Manu
<Zakim> mbgower_, you wanted to say that I would like to discuss the goal of development iteration and motivation
<wendyreid> ... he said that accessibility is an iterative process
<Cyborg> what David just said aligns with what i was saying - we need a rubric and documentation to assess conformance models and to document that
<wendyreid> ... it's not set up like that now though
<wendyreid> ... questions around motivation and iteration
<wendyreid> ... if we can develop a mechanism that as product A I have x amount of accessibility, and there's an increase in usability/a11y, that's a benefit
<wendyreid> ... it's anathema to the idea that products should just be accessible
<wendyreid> ... no product is complete, it's always improving
<wendyreid> ... a11y may need to fit into that model
<wendyreid> +1
<wendyreid> jaunita_george_: What we could do, we have orgs with critical paths, critical workflows
<wendyreid> ... these must not have any failures
<wendyreid> ... but tolerate failure in more incidental sections of the site
<wendyreid> ... we would want to help develop tooling for people who need to implement this
<Rachael> draft RESOLUTION: Develop a process to eliminate options and document what is removed and why
<Wilco> -1
<jeanne> +1
<jaunita_george_> +1
<Ben_Tillyer> +1
<wendyreid> Rachael: Draft resolution, it sounds like there is a proposal to document how and why we eliminate options
<David-Clarke> +1
<wendyreid> mbgower_: I don't understand the context for eliminate
<wendyreid> Rachael: As we move forward we need to make some decisions
<Cyborg> my suggestion is we develop a rubric
<wendyreid> ... whether we should remove weighting
<wendyreid> ... document a process based on criteria, we document the elimination and why
<Cyborg> if we have a rubric and agree on the rubric then yes it is a useful thing to do
<wendyreid> ... is it useful for us to do
<ShawnT> +1
<wendyreid> Wilco: The process we should have is one where we figure out the best approach for each scenario
<Cyborg> an assessment process to evaluate
<wendyreid> ... as opposed to eliminating
<wendyreid> Rachael: Or not eliminate and move forward
<wendyreid> Cyborg: I am confused as well, I suggest we develop a rubric
<wendyreid> ... based on the rubric, which can iterate, that process of assessment is documented
<wendyreid> ... if the model can adjust to adapt to that rubric, then it can be used
<wendyreid> Rachael: It sounds like we won't get agreement on this issue, it goes in the parking lot
<wendyreid> ... moving to slide 6
<wendyreid> jeanne: 2017, which feels like a long time ago, we did 18 months of research
<wendyreid> ... partnering with academic and corporate research departments
<wendyreid> ... list of questions, how to improve WCAG 2
<wendyreid> ... various academics and corporate researchers, gave us the results, and we analyzed it all
<wendyreid> ... what we boiled down to are the issues that WCAG 3 is trying to solve
<wendyreid> ... usability, WCAG 2 was too difficult to understand, read, translate
<wendyreid> ... too difficult to understand thelanguage
<wendyreid> ... conformance, strictly testable costraints, we couldn't include anything that wasn't strictly testable
<wendyreid> ... human vs machine testable
<wendyreid> ... we need to address these things
<wendyreid> ... how do we accommodate accessibility support
<wendyreid> ... how do we include evolving technologies
<wendyreid> ... maintenance, how do we maintain WCAG 3, we need flexibility, scalability
<wendyreid> ... governance issues, developing needs of disability groups experiencing issues with new tech
<wendyreid> ... takes too long to get them into the standards
<wendyreid> mbgower_: To Jeanne, is it your sense that 5 years later, the findings on conformance, are those still all good questions for use to be solving
<wendyreid> ... are these proving harder to solve
<wendyreid> jeanne: we're starting to come up with solutions
<wendyreid> ... they're all still valid
<wendyreid> ... if you look at success criteria, the user needs identified by groups that can't go in WCAG 2
<wendyreid> ... some orgs want only automated testing
<wendyreid> ... others want to have human testing
<wendyreid> ... evolving technology, we know today, we have new tech coming out all the time and it's hard to include them in WCAG 2
<wendyreid> ... we've possibly added more
<wendyreid> ... starting point backed by research
<wendyreid> AWK: Probably covered before, out of the research, this slide mentions the gaps
<wendyreid> ... is there a list that came out that lists what was essential about WCAG
<wendyreid> jeanne: It's not in this slide
<wendyreid> ... the first revelation, people love the guidance of WCAG 2
<wendyreid> ... they like being told what to do
<wendyreid> ... the structural parts were where there were complaints
<wendyreid> ... people like having instructions, but they need to understand, measure, and implement
<wendyreid> ... WCAG 2 does a good job, but needs improvement
<wendyreid> jeanne: Some of the things we've been talking about this week
<wendyreid> ... more equity between disability groups
<wendyreid> ... work after the research as we built prototypes
<wendyreid> ... discovered along the way
<wendyreid> ... A AA AAA has a lot of requirements for disabilities at a higher level
<wendyreid> ... the Deaf community has a lot of requirements at AAA
<wendyreid> ... you either pass 100% or fail
<wendyreid> ... perfection or failure
<wendyreid> ... that's another area we need to address
<wendyreid> Rachael: Move on to slide 8, the topic of architecture
<wendyreid> ... a conversation to see if we are all on the same page
<wendyreid> ... some questions on architecture
<wendyreid> ... timebox discussion for 30 minutes
<wendyreid> ... what are the perceived gaps
<wendyreid> ... last slide is what are we missing
<wendyreid> ... structure for WCAG 3 is guidelines, outcomes, and methods
<wendyreid> ... three levels, similar to WCAG 2
<wendyreid> ... like captions, translating text into audio, or alternative formats
<wendyreid> ... provide text equivalents, or display
<wendyreid> ... three different levels, tied to that, the method has content to test
<wendyreid> ... intro, background, samples, a glossary
<wendyreid> ... slide 11, a how to
<wendyreid> ... planning, designing, developing
<wendyreid> ... designed for people new to a11y guidance
<wendyreid> ... slide12, two things to do on architecture
<wendyreid> ... methods and how to look alike
<wendyreid> ... form a group to make the styles more organized
<wendyreid> ... can members volunteer designers/engineers to help
<wendyreid> ... how will the guidelines be organized
<wendyreid> ... tagging to organize based on a user's needs
<wendyreid> ... default may be confusing
<wendyreid> ... request to visit this in open conversation
<wendyreid> ... open the floor
<Zakim> jeanne, you wanted to say that the reason we didn't want to keep POUR is that so many SC apply to both
<wendyreid> jeanne: The reason we did not keep POUR in the FPWD, we found so many SCs applied to multiple categories
<wendyreid> ... it would be much more useful to organize with tagging
<wendyreid> ... anyone who wanted to see what applied to Understandable, would get a more robust list
<wendyreid> ... move from the strict hierarchy
<wendyreid> MichaelC: Picking up on the design thing
<wendyreid> ... we need help here
<wendyreid> ... I am not a designer, it's not just visual, structure
<wendyreid> ... design the information flow
<wendyreid> ... how to go through the different flows
<wendyreid> MichaelC: Let's look at tagging very broadly
<wendyreid> ... I hope that we will be rich in our tagging
<jeanne> +1 to rich tagging
<wendyreid> ... unused is better than missing
<Zakim> AWK, you wanted to ask if we've decided what is normative
<wendyreid> AWK: Architecture level question, have we determined what is normative?
<wendyreid> Rachael: No, intent was to do so after determining conformance
<wendyreid> ... but if we need to do that now to move things ahead we can revisit
<Zakim> mbgower_, you wanted to say that I'd like to understand the degree to which the presentation/structure can be malleable
<wendyreid> mbgower_: I think this falls along the lines of what Michael was talking about
<wendyreid> ... I'm not sure how much we can revamp the presentation
<wendyreid> ... of the existing document without affecting the process
<wendyreid> ... I do get bewildered with the current format
<AWK> +1000
<wendyreid> ... we need obviously some structure to tackle the information
<wendyreid> ... I'm not clear how much the information design is solid and we can move ahead
<wendyreid> ... it's hard for me to know how much we can explore the decisions we've made
<Zakim> alastairc, you wanted to suggest a starting point for re-structure
<wendyreid> alastairc: I agree with Wendy, we need an IA
<wendyreid> ... information architect
<wendyreid> ... guidelines to outcomes to method
<wendyreid> ... they all look similar
<wendyreid> ... that would be my starting point
<wendyreid> ... take the IA approach to make the relationships clear
<wendyreid> ... make it understandable
<mbgower_> +1 for IA, although the architect is going to want clear requirements and business rules!
<wendyreid> ... agree it would be good to have a group working on this
<Zakim> jeanne, you wanted to talk about the database
<wendyreid> ... starting from IA to the way components are structured and disaplyed
<wendyreid> jeanne: Going to the original slide, what we did, we had a design sprint
<wendyreid> ... we invited a group, 37 people, representing stakeholder groups
<wendyreid> ... a broad spectrum of the a11y industry
<wendyreid> ... advocates, lawyers, corporate, etc
<wendyreid> ... we gave them problems, they brainstormed solutions
<wendyreid> ... one solution was to put things into a database
<wendyreid> ... that people can transform as they need it
<wendyreid> ... with the direction it's going, people won't be reading the spec
<wendyreid> ... they'll go to wcag3.org and look at an opening page
<wendyreid> ... look at a starting page of a database
<wendyreid> ... we are starting with a spec, then mocking up a database
<wendyreid> ... we did have an IA, who helped us make some decisions, but he wanted us to drop a level
<wendyreid> ... AGWG wanted a measurable outcome level
<wendyreid> ... we did work with an IA to get some of that work done
<AWK> Scribe: AWK
MichaelC: Support help from an
IA
... a strict hierarchy will be constraining
<jeanne> +1 for tagging
MichaelC: there will be IA
questions on managing that
... also need interaction design - how exposed and how working
with it
<alastairc> I wasn't suggesting it has to be a strict heirarchy, but making sure the relationship of guidelines > outcomes > methods is clear.
MichaelC: another thing that is challenging is for content to be normative it needs to be published as a static technical report
<jeanne> +1 for needing a TR version
MichaelC: if the GLs become long as I suspect it will make that harder to use but will be needed
wendyreid: use WCAG every
day
... coworkers are UX pros
... should consider user journeys
<jeanne> We did the Stakeholder analysis in 2016
wendyreid: there are multiple
stateholders - lawyers, accessibility leads, etc also
... would love to see results of design sprint
<jeanne> Stakeholder Job Stories https://www.w3.org/WAI/GL/task-forces/silver/wiki/Job_Stories_for_Stakeholders
<MichaelC> +1
<Rachael> draft RESOLUTION: Set up a subgroup post TPAC to iterate information architecture and page design
<jeanne> Silver Design Sprint Final Report https://www.w3.org/community/silver/draft-final-report-of-silver/
Rachael: propose draft resolution above
<jaunita_george_> +1
Wilco: may be making more
complicated than needed
... maybe the IA decisions come after we know what is in WCAG
3
<alastairc> +1, we do need something to look at
MichaelC: in principle agree with wilco in practice people need something to react to
<mbgower_> +1 that we both don't have to design the end product now AND need information now to be able to work from
Rachael: as a group are hearing confusion and need to resolve it
<jaunita_george_> +1
<Rachael> draft RESOLUTION: Set up a subgroup post TPAC to iterate information architecture and page design
<ShawnT> +1
<jaunita_george_> +1
<MichaelC> +1
<alastairc> +1
<David-Clarke> +1
-1
<Ben_Tillyer> +1
<jeanne> +1
<Wilco> -.5 I think it's too early
mbgower_: can you clarify more about page design?
Rachael: the CSS between slides 10 and 11
<mbgower_> +1 to doing what alastair just said
alastairc: more than just the CSS - it is how the pieces link together and the user experience when viewing
<Rachael> draft RESOLUTION: Set up a subgroup post TPAC to iterate information architecture, css, and how the documents link to each other
Wilco: think it is too soon
<jeanne> Folder of information on the Information Architecture work done in 2018-2019 https://drive.google.com/drive/folders/1VtNVXe8PaM6NA8pZLyUDPv4Hh99IG59K
<mbgower_> 0 I'm inclined to leave IA off of this resolution right now
need more clarity on what is being designed and architected
<alastairc> AWK: If i went to one of our designers and asked for help, to look & feel & behave, they would ask us what we are trying to express.
<Rachael> AWK: If I went to one of our designers and said help us make this look and feel and express the intent of what we are doing here, they would ask what are you trying to express. That is what we are trying to work out.
<alastairc> ... I don't have that clarity on WCAG 3 now, the clarity doesn't help me. Need to answer foundational questions first.
<alastairc> ... then someone with those skills could do a better job.
<Rachael> ..better to figure out foundational questions and then create IA from that. If we have those skills, someone can say this is what this should look like, this is what that should look like.
MichaelC: hear that it is too
early but
... but we need it also
... assume that we aren't going to develop the IA in 8 weeks in
a subgroup
... perhaps we could get an IA to guide us in our decisions
Rachael: the reason we are having
this discussion is that people couldn't approve this without
discussing
... seeing a lot of support but also some concerns
<alastairc> I suggest that if we have people who want to work on it, we facilitate that, and see if they can provide a useful itteration at this time.
Rachael: propose changing resolution
<Rachael> draft RESOLUTION: Set up a subgroup post TPAC to iterate css, and how the documents link to each other. IA rework at a future time
<mbgower_> ...to improve the connection between the guidelines and other documents linked to from the existing PWD
<jaunita_george_> +1
<Rachael> draft RESOLUTION: Set up a subgroup post TPAC to iterate style and how the documents link to each other. IA rework at a future time
<mbgower_> +1
<jeanne> +1
<ShawnT> +1
<Amanda_> +1
<MichaelC> +1
<Ben_Tillyer> +1
<Wilco> 0
<alastairc> +1
<David-Clarke> +1
+.5
<mbgower_> It should help us work with it!
<Rachael> +1 mbgower
RESOLUTION: Set up a subgroup post TPAC to iterate style and how the documents link to each other. IA rework at a future time
mbgower_: do we need resources?
<alastairc> I think Jeanne posted them in...
Rachael: YES
... moving on
Rachael: exploratory
conversation
... rest of meeting on this
... slide 14 is key points
... test based and task based
... slide 15
... types of tests
... computational - results don't vary
... qualitative - results may vary
... four types of requirements
... slide 17
... pulls these all into one table
... slide 18 shows example for image alternatives test
<Zakim> mbgower_, you wanted to say can we change "adaptive" to "contextual" requirements?
mbgower_: an example of
computational vs qualitative extensible might be related to
closed captioning
... also wants to change "adaptive" to "contextual"
<Zakim> AWK, you wanted to agree with Mike Gower
<alastairc> +1, contextual makes sense to me for that concept
Wilco: slide 18
... what was interesting was that all of these might be things
that an org might need to do
... was thinking that orgs might use just one of them(?)
<MichaelC> +
<Zakim> shadi, you wanted to respond on contextual and to respond on slide 17 (no difference in test types)
Wilco: as written this seems that they are building on top of each other
shadi: The reason that we didn't
use contextual is because qualitative tests are also
contextual
... silde 17 was also interesting - didn't think that there
would be any differences
... basically two types of tests, some requiring tester
knowledge and some that don't.
Cyborg: slide 17
... what Im seeing is that what is likely to be different is
the difference between the main categories
me https://docs.google.com/presentation/d/18W8M21acEeb5Z1ARdafTQc01_mzOvvjsS184_Qoma0c/edit#slide=id.g1544c217c09_0_133/me https://docs.google.com/presentation/d/18W8M21acEeb5Z1ARdafTQc01_mzOvvjsS184_Qoma0c/edit#slide=id.g1544c217c09_0_133
Cyborg: a different parallel
matrix might make it clearer
... As you move toward the bottom and more to the right there
is more complexity
MichaelC: assuming that the
different test types will map to requirements in different
ways
... there will be requirements that need >1 test type and
others that don't
<shadi> +1 to MichaelC
<Cyborg> is it helpful if i write my question here too?
<mbgower_> scribe+
<Cyborg> i'm not seeing it reflected in the notes
<mbgower_> AWK: Types of tests slide says "requirements". What does that mean in the context of this slide? Is it 'outcome'?
<mbgower_> ... it feels like we have a type of a test and then a type of a requirement.
<mbgower_> ... on slide 15
<mbgower_> scribe -1
<mbgower_> scribe-
<Cyborg> the question is: can we imagine complexity of test types as a matrix, where the upper left is more simple, clear, doable at volume in automated way, and the bottom right is more complex, customizable, but requires more human testing. both have their uses, depending on the needs of end user and the type of organization doing them.
jaunita_george_: The requirements are like the flavors of outcomes
<Zakim> Rachael, you wanted to point out differences between 18 and 19
<Cyborg> so my question is that true or not and why?
<Cyborg> i don't feel like i heard an answer, so i'm not sure if that is the case.
Rachael: slide 19 shows that some outcomes are largely computational
<Zakim> MichaelC, you wanted to say we´ll need to answer when two test types equally test the same thing
<Rachael> +1 to both "and" and "or"
<Zakim> Wilco, you wanted to respond to AWK
MichaelC: Possible that we have different test types that are duplicative of each other and one might be "better"
Wilco: We could think of the
requirements as outcomes or methods, hasn't been
determined
... (wilco, missed that last point)
jaunita_george_: Looks liike I have some volunteers for the design
<Zakim> Rachael, you wanted to argue against using these on outcomes since they are tech neutral
<alastairc> scribe-
<alastairc> We had a whole big discussion about "tests", apparently the four requirements aren't tests.
Rachael: if the mid level is tech neutral then these shouldn't be requirements targeting that level
Cyborg: The bottom right is more complex, more broadly impacting users, and harder to test - this area may need more work as upper left is more like the current WCAG
<Amanda_> +1 to ben
<mbgower_> +1
<wendyreid> +1000 as an ex QA
Ben_Tillyer: If had to ask a UX pro if they used a test with these combinations expect a blank stare
<jeanne> It's what we have today with technology neutral success criteria and and informative tests and techniques
<Cyborg> my suggestion was that bottom right is used in more complex, personalized/customized situations by organizations that are more nimble addressing more complex needs (for example of people with multiple disabilities), and therefore a shift from WCAG 2. So might the next step to work on be the most bottom right, to work on testing based on complex needs of people with multiple disabilities, in nimble environments, by smaller and more nimble organizations
<Cyborg> (cont)
<mbgower_> To Ben's point, computational and qualitative are useful for everyone thinking about test formation. Are the rest necessary for most involved to know?
Rachael: we do need to figure the normative/informative question out
<Cyborg> to advance more accessibility and then to slide those solutions towards the upper right, to make them more adaptable to larger orgs who currently benefit from prescriptive tests of WCAG 2.
Rachael: suggest for the group is
that we try to fill out 1-2 of these charts
... and then come back to the questions
... Back in 1 hr
... on zoom
<Cyborg> sorry (correction) to start bottom right and to slide towards top left.
<Cyborg> in other words, as with principle of inclusive design, focus first on designing tests for more outlying cases and then consider how those might be simplified
<Cyborg> more complex use case would be a list of criteria, including possibly: task-based, needing customization, highly contextual, impacts people with multiple disabilities more than an individual disability, has friction more than a critical failure, and is more used by smaller orgs with more diverse clients than by a larger org with more mass markets.
<Cyborg> would that be more bottom right?
<Cyborg> is the cost/benefit tradeoff of designing tests for those situations that they are time consuming, need to be human led, are hard to scale.
<Cyborg> so then can we design for those bottom right situations and identify ways to make them less time consuming, less human led and easier to scale.
<Cyborg> that's what i mean by starting bottom right and moving to upper left.
<Cyborg> good luck to everyone sticking around for afternoon.
+AWK
<Ben_Tillyer> +1
<Rachael> mbgower: Discussion after morning session. For purposes of people involved in creating WCAG 3.0. Only need to worry about computational and qualitative. Anything more than that is likely an impediment.
<Rachael> ....We should maybe stop worrying about publishing. We are only talking abotu terminology used in house.
<mbgower> scribe: mbgower
<jeanne> -1 because it doesn't include good potential solutions for adding more content that can't be added like protocols
shadi: this is exploratory material. sometimes it's useful to show the world what we're thinking about, even though it might not be there in the end
<alastairc> You sometimes need to go through complexity to get the simple version.
shadi: It might go into a glossary or something, if needed.
<jeanne> +1 shadi to explore the terms
<Wilco> +1
<jaunita_george_> +1
shadi: I am wondering if putting
all these tables together is creating more confusion.
... I did find it useful to see how things build on each other.
But it's not what it was intended to be.
... Maybe we can be a bit clearer: 2 test types. Four ways of
thinking they can be applied, which are not mutually
exclusive.
... Maybe we didn't explain that this wasn't designed to be
comprehensive.
Wilco: The thing I've been
wondering about is whether we need computational and
qualitative.
... We could potentially identify the computational tests
David-Clarke: Really computational could be better substitutable with automated
<AWK> mbgower: value of computational / qualitatitive - by being aware of it we can take things that are qualitative and figure out how to minimize disagreement and target efforts to make them more quantitative
scribe+
Rachael: l'm concerned about the
idea of getting rid of these.
... If we don't have those concepts, how is this different than
wcag 2?
... How do we extend coverage?
<Wilco> +1
<jeanne> +1 -- we need more coverage of disability categories
shadi: we didn't get that far to answer any of these questions.
<jaunita_george_> +1 to shadi
<alastairc> +1, we don't have to use them yet, but the concepts seem useful to expand coverage
shadi: we started with the 4
categories that were there. We dissected them.
... We came back to the fact that in all these types there is
an underlying step that are either computational or
qualitative.
... So we reduced the 4 into 2.
... But there are different ways of applying them in methods
and requirements. But we didn't explore that entirely.
... No keyboard trap is prescriptive.
... But there can be multiple contrast algorithms.
... But testing, it's going to be computational or
qualitative
So there is a separation between testing and outcomes/method writing.
<Zakim> alastairc, you wanted to talk about next steps - finding/creating examples
alastairc: it's useful to have
these concepts. I think they are primarily internally
facing.
... It would be good to take both wcag 2.x materials and areas
we want to extend other outcomes and come up with some
examples.
AWK: I think it was Rachael who
said 'we need these' or it will constrain us to what we have in
wcag 2.
... The biggest category is 'more qualitative'
... I assume what we do with those 2 is we say, 'okay, these
are the outcomes'
<alastairc> qualitative could be a very wide concept, so far we've used it for the qualitative aspects of WCAG 2.x SCs. The adaptive / extensive/ protocol all seem to fit in qualitative to me.#
AWK: I don't think it will constrain us. If we have a bunch of requirements, we can expand. But I like the idea of simplifying.
Jennie: I'm looking at the homework slides. Do we have a working definition of 'test'? It could be contributing to confusion.
<Wilco> I don't
<alastairc> +1, I got confused where the requirements were separated from tests, and I haven't fully understood why
<jeanne> -1 to AWK, because I think it will constrain us. We have good proposals for solving some of the inclusion issues, like protocols.
[ no one had definition]
AWK: We do have a definition for test in wcag 3
<Rachael> Test: Mechanism to evaluate implementation of a method. Tests can include true / false evaluation or various types of rating scales as appropriate for the guideline, outcome, or technology.
<jaunita_george_> +1
<David-Clarke> +1
alastairc: The requirements got
separated from test. I didn't understand why.
... Could anyone explain?
<AWK> @jeanne, I'm not saying that we can't use protocols, just that we don't need this additional categorization right now
<jeanne> A test is a measurement of an outcome that is included in a Method.
Wilco: A method could have more than one test in it.
<Rachael> Test: Mechanism(s) to evaluate the implementation of a method.
Rachael: Alastair, do you have anything specific you were referring to?
alastairc: Originally we had 4 types of tests. Then we had the big discussion where we decided the other things weren't tests and were called requirments
shadi: I think there were
refernces where 'conventional' referenced conditional or
unconditional. They weren't really distinct. The final one was
called composite or something.
... We only identified 2 test types.
<alastairc> Ok, that makes sense, but didn't help with the definition of test as I'd hoped.
shadi: Then there were ideas of how to apply these tests.
<Rachael> draft RESOLUTION: A test is a mechanism(s) to evaluate the implementation of a method.
shadi: I agree
<jaunita_george_> +1
<Wilco> +1, good enough for now :-)
<jeanne> +1 good enough for now
Wilco: I was suggesting that a method could have more than 1 test.
awk: so in the definition should it be plural?
<Rachael> Test: Mechanism to evaluate the implementation of a method. Note: there may be more than one test to evaluate a method.
+1
<alastairc> +1
<Ben_Tillyer> +1
<jaunita_george_> +1
<AWK> +1
<jeanne> +1
<ShawnT> +1
<Amanda> +1
Jennie: I'm still stuck a bit on "method" but if the group is comfortable, that's fine.
<Rachael> method: Detailed information, either technology-specific or technology-agnostic, on ways to meet the outcome as well as tests and scoring information.
<Jennie> Thank you Rachael
<Rachael> method: Detailed information, either technology-specific or technology-agnostic, on ways to meet the outcome.
<alastairc> scribe:alastairc
Rachael: This slide is one to
try, we can do an exercise to see if this works. Does it makes
sense as a way to break things up?
... this example is for captions for audio content.
mbgower: The idea that the
captions, being open or closed, the values you assess on closed
captions are different from open captions, from the perspective
that open captions are in place and so you can't change
them.
... there's not a set of qualities you'd assess for open
captions, but are for closed.
Rachael: So that would be
different evaluations for the context.
... I had thought adaptive was different contrast modes, but
extensible would be different contrast algorythms.
Wilco: Yes, that's how I've
thought of it. trying to apply these 4 to the same type of user
need might not be best way to do this. We're trying to extend
the user needs to what we want to test.
... we might not need other types if it works with the
prescriptive one.
Rachael: Question back - does that negate the excercise, if things can be prescriptive then we don't need the others?
Wilco: E.g. for colour contrast, you could also do usability testing, but why would you?
shadi: Part of the issue is that
the whole purpose was to go beyond WCAG 2. What we can
currently express in WCAG 2, there isn't a strong motivation to
change because we can. Point is to have more coverage. As soon
as we do that, we don't have worked out ideas for it.
... so it gets hypothetical, we have to imagine ideas.
... in WCAG we don't cover the quality of captions, would such
a requirement be prescriptive? Is it specific to different
languages, genres?
... we have more options, can create more things, but needs to
be thought through to be convincing. Sub-groups could work on
that. For things not in WCAG 2.
... e.g. for quality criteria for captions.
... Rain has been working on language examples.
... and there are more.
Rachael: I agree that quality of captions, captions in different environments. There could be a principle: things should be met with the simplest things possible. Go beyond that if the simple methods don't meet the user needs.
<Zakim> mbgower, you wanted to say that this is where the goal of thinking up assessment approaches enters
mbgower: The goal gets
interesting. There's a lot of different ways you could think of
for testing things. imagine you're created a decision tree, and
you want to expose the tree to different outcomes to reduce
effort.
... e.g. start with does the page contain video? If no, for
this page, we've eleminating a bunch of criteria.
... If yes, we have a whole set of questions. does it contain
audio? If no, no author action required.
... also get into the idea of the player, so there's video, so
is there a caption button? A lot of the qualitative questions
around language & meaning are the same.
... there may be more tests, but all from the idea that we use
decision trees to output the tests.
... really interested in that area, but applicable to
testing.
... what do we have, as a result of what we want to convey
should come from testing.
<Zakim> jeanne, you wanted to say you would do a usability study if your corporate colors didn't meet color contrast. it has orange in it, for example.
jeanne: lot of ways of slicing
these. Working with an org that has orange, and generally has
good contrast, but can't pass.
... if they could use a user-test to pass, why not let
them?
... when we've tried to apply plain language, people say it
won't work as it doesn't apply to non-english languages.
<Zakim> Jennie, you wanted to discuss use case with pixels
jeanne: need the flexibility, not get locked into things because it has to work everywhere.
<Ben_Tillyer> 1.4.3 Logotypes Text that is part of a logo or brand name has no contrast requirement.
Jennie: We use to have an acronym logo where the lowercase "i" wouldn't pass contrast because it wasn't distributed across the screen the same way. There are very specific use-cases for these.
Wilco: I think for any outcome
there would be some qualitative testing. I don't think any ACT
tests are free of qualitative components.
... to Jeanne's point, there are use-cases where you'd want to
run usability tests to show it works.
... it would take some convincing of people, but really
interesting idea.
... sounds like we're converging these requirement types, but
are they really method types?
shadi: I think the best way to
convince is to prototype.
... I'd love to see it. We should t develop prototypes, for the
ones not in WCAG 2.
<jeanne> +1 to prototyping
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test types and requirements as methods
<Ben_Tillyer> +1
<jaunita_george_> +1
<Poornima> +1
<Amanda> +1
<Wilco> +1
<jeanne> +1
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test types and test vrequirements as methods
<Rachael> draft RESOLUTION: Set up several subgroups to prototype the test types and test requirements as methods
shadi: Might be too much for 1
subgroup.
... the protocols one is it's own beast, each of these needs
it's own focused work.
... seeing how quickly 8 weeks passed, needs multiple
<Rachael> draft RESOLUTION: Set up several subgroups to prototype the test matrix (slide 17) as methods
jaunita_george_: Happy to do another protocols group, but want to see if there are resources to conduct usability test if a mock protocol would work against some sort of requirement.
Rachael: let's circle back, not sure we mean the same thing.
<Jennie> -q
Wilco: Not sure about multiple sub-groups. Not sure it's baked enough to go into multiple tracks and not diverge in odd ways.
Rachael: Are you thinking of a sub-group that would take one concept (e.g. captions) and do multiple ones. Or write up sample methods abstractly for each concept?
<Rachael> ack ack Rachael
mbgower: I thought I heard that the 4 requirements are actually methods? Combining that with test-type assessments. Q to wilco, is that somehting we need to answer first?
Wilco: yes, I think we explore
the test requirements should be types of methods, and the
test-types are within that.
... a method with different tests, computational some
qualitative.
mbgower: And do we have something broken up enough by methods that we can do this assessment with?
<jeanne> I agree with Wilco that it could have multiple tests in the Method. That is what came out of the Test Reliability work last spring
Rachael: Do we have enough outcomes broken into tests that we could do that?
jeanne: we could take the outcomes we have, but only have 7-8.
AWK: Text alts have 5, the others have one to three.
jeanne: The language aparts one as well. Some are no longer appropriate. It's a limited number. for an initial prototype I think we could go with what we have.
<Rachael> +1 to clear words or error prevention
Wilco: Clear words might be the obvious place to start, or error prevention.
<jaunita_george_> +1
Rachael: COGA started clear words, might be interesting to try both.
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods, likely using clear language or error prevention
Rachael: setting up a single sub-group, using slide 17, using clear lang / error prevention.
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods, likely using clear words or error prevention
Jennie: Does the sub-group work along side the COGA one? or in parallel.
Rachael: If it was taking clear-words, we'd ask people from COGA to help, but pause the work in coga.
Jennie: It may also be helpful if
they can have some orientation call to the work prior to
starting the new sub-group.
... 8 weeks is a great concept, they may need some prep.
mbgower: I'm sensing scope creep. The exersize is to check if the tests work, rather than create the final tests for those guidelines?
AWK: Suggested edit for the resolution.
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods and tests for an outcome, likely using clear language or error prevention
AWK: one axis is types of methods, the other is types of tests. So suggest "Set up a subgroup to prototype the test matrix (slide 17) as methods and tests for an outcome, likely using clear language or error prevention"
<jeanne> +1
Rachael: If we pick up a COGA outcome that they've been working on it, need their expertise in group.
<jeanne> +1 to build on COGA expertise
Jennie: If we're looking to build consensus, what this does is inform where the two groups don't understand all the components, might help us move forward.
<jeanne> +1 Jennie to include COGA to identify gaps
<Amanda> +1 to include COGA
Jennie: not as good to check little things off, but helps wider scale.
Wilco: In addition, how do we check the quality of these informative methods? If informative methods are the way to test, how do we check quality? How do we know that clear words for english works? How about the Hebrew, german ones?
jaunita_george_: Should we try to
get a panel of volunteers /partner(s) in the disability
community to validate these?
... we should probably work on setting that up
mbgower: Now I definitely feel the scope creep! I thought we were trying to carry out an experiment on the 4 requirement types works as methods.
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods and tests for an outcome, likely using sample methods
<Rachael> q
mbgower: that seems doable in 8 weeks, but coming up with something that works across languages, that seems like a different scale of work..
<Wilco> +1, lets go with it
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods and tests
Rachael: New draft resolution, focused on the tests
+1
<David-Clarke> +1
<ShawnT> +1
<jaunita_george_> +1
<Jennie> +1
AWK: Just trying to categorise the methods?
Rachael: Take a sample of each in the matrix, and flesh it out.
<maryjom> +1
Rachael: if that works well, follow up with quality checks, usability testing, then take one then apply to different outcomes.
<AWK> +?
<jeanne> +1
<Rachael> draft RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods and tests
<David-Clarke> +1
<Poornima> +1
Rachael: We'd have a detailed example of each type of method. both for computational and qualitative.
RESOLUTION: Set up a subgroup to prototype the test matrix (slide 17) as methods and tests
<Wilco> move forward
Rachael: Do we want to go back to adaptive vs contextual?
<Ben_Tillyer> +1 to moving forward
<AWK> +1 to jennie
<Rachael> +1
<shadi> +1 to Jennie
Jennie: I have to read those words with the definition. I don't spend a lot of time on it, but they aren't descriptive enough in my head.
<ShawnT> +1 to Jennie
Rachael: The two main ones I have trouble with are adaptable and extensible.
<Rachael> Strawpoll; Which terms are causing people challenges?
<jeanne> extensible and adaptable
<AWK> Adaptive, extensible
<Rachael> Adaptive nad extensible
<mbgower> same
<jaunita_george_> same
Adaptive, extenesible a bit.
<Jennie> same
<Amanda> same
<ShawnT> all of them, hard to remember
<maryjom> Adaptive extensible. I have trouble with terms beyond that too - method, etc.
<AWK> Variable instead of adaptive?
ShawnT: I'm not spending much time trying to remember them until they are more final.
<Rachael> ack q+
Ben_Tillyer: Is extensible a collection of tests? It must have multiple, therefore not a test type?
Rachael: Adaptive are different contexts, extensible is different rulers.
Ben_Tillyer: Wouldn't you then have prescriptive tests, e.g. there are open captions? yes/no. Or there are extensible...
Wilco: Prescriptive is where
there is only one way to meet the need, extensible means there
are multiple ways.
... a test can be a more atomic thing. A prescriptive thing
could be that the lang needs to be defined, e.g. lang attr
needs to be defined, and needs to be correct.
... can be qualitive + prescriptive.
mbgower: We're saying that a
prescribed outcome, which we scrutinize against a set of
methods... so on these types there is clarity at the method
level.
... if you're going to have a method with open captions,
wouldn't that be at the method level?
... that's where I'm struggling.
Jennie: In my mind, if
prescriptive, "do this". If it is one or multiple things it can
be prescriptive.
... it's the way it's written for me I'm struggling with.
<Rachael> + to wilco
Wilco: I've been thinking about it is that tests can be more granular. E.g. multiple caption tests that go to a single method. Or maybe we allow different styles which then requires we make it extensible.
Rachael: I'd been looking at this
as outcomes being the goal, e.g. images have a text alt. The
methods are the different ways that come together than meet
that goal. Could be 'and', do several things. Some things might
be optional.
... that's why we started at the test level, so we can do this
more informed.
... going back to text-alts. For that it has to have a name, is
it decorative, is the text alt sufficient? You can build on
that all the way up to usability-testing to see if they support
the task. They all come under having alt text, but all build /
interactive with each other.
... may have different steps to meet the tests... really a
decision tree, that might be the way to approach it.
<Wilco> I don't think we are
mbgower: It's a decision tree point, it's a computational test, but I thought it wouldn't fit that model. It leads to a series of questions... just want to make sure we're not ruling this out as an approahc.
AWK: where there's a test for open captions vs closed depends on what the outcome is asking for.
Jennie: So when we do protocols
in therapy, is that a single step or multiple step?
... when we have a multi step process where it's all
prescriptive, or could be combination. We should consider the
complexity the person is doing in that process.
... or maybe we're too late on the terminology boat.
Rachael: We have a decision to
make about whether open captions are acceptable, but ties into
how we deal with a sub-optimal solution.
... I don't want to loose the fact we struggled with those
terms, do we always use the definitions with them.
mbgower: Whoever does the 8 weeks on it will tell us.
<ShawnT> +1 to mbgower
<Rachael> +1
<shadi> +1 to mbgower
<Jennie> +1 to Mike
mbgower: I suspect we aren't going to care about the terms once identified. It helps us define these things, but probalby won't use them long term.
<AWK> +1
<Wilco> That works
<Ben_Tillyer> +1
Rachael: So we can leave these temrs for now?
<ShawnT> +1
+1
<Jennie> +1 as long as COGA knows it will be rereviewed
<jeanne> +1
<Amanda> +1
<mbgower> +1
<jaunita_george_> +1
Jennie: Do we mean - once we've tried these tests, we'll either rename them, or re-review that? COGA need to know ahead of time. don't want them to spend time thinking about that.
<ShawnT> Not me
<Jennie> Not me, sorry
Rachael: Our next round is to talk about different options in conformance. The next session is moderated.
<maryjom> +1
<jeanne> Not jeanne, but I may listen
<maryjom> I've got to go...
<Wilco> +1 to normative
<Rachael> Discussion of whether to have a working session based on number of people here.
<Rachael> Option 1) Start the testing subgroup work
<Rachael> 2) Whiteboard normative and informative
<Rachael> 3) Brainstorm ideas for maintaining the guidance
<mbgower> 1
<jeanne> 1
<Rachael> 4) We can continue talk
<MichaelC> 2 or 3 but no strong feeling
<Rachael> 1 or 2
<AWK> 1/2
<Ben_Tillyer> 2, 3
<Amanda> 1 or 2
<Wilco> 2, 3, 4, 1 in order of pref
<mbgower> yep
<jeanne> 2 is fine
<Rachael> We are going to whiteboard normative and informative.
<Rachael> https://jamboard.google.com/d/1b3MZ6ToJja5vbKiRREpWBpCnReFRz4BO37EtR03IlGo/viewer
<Rachael> scribe: Rachael
How Tos and Other method content is informative
Not what sure if principles, functional nees and tests are.
Right now we have non-normative sufficient technique with a prescriptive test that allows you to meet a requirement. So in WCAG 2 its non-normative because its under a non-normative content
But you don't have to use a sufficient technique.
Andrew: In WCAG 2 time period,
the working group was concerned it couldn't cover the scope of
possibility.
... not sure we've covered everything.
... even if we only have to review the techniques, its more
than we have historically been able to do.
If tests are normative, we would not be able to handle it.
Jeanne: Question whether the tests prove the method. I think tests prove the outcome. I think we could keep tests and methods as informative
I don't think we can maintain the methods and tests as normative.
Michael: I woudl like to have only one normative layer. If outcomes are normative, then the guidelines are ?? anyway.
If methods are normative, then they restrict how people can meet outcomes.
<Ben_Tillyer> https://www.w3.org/TR/wcag-3.0/#approximate-mapping-of-wcag-2-and-wcag-3-documentation
mbgower: Follow up on Michael, Is
a sufficient technique in WCAG 2 comparable to methods right
now?
... Right now the guideline is the normative part and the
methods can be used to meet the guideline. Can Wilco go through
how ACT tests relate to teh current working model.
Wilco: ACT tests are informative.
Intended them to be that way because we need to update them
regularly. tech specific. Have to update ...can't be
normative.
... when it comes to outcomes, as long as they are testable
statements, I think the WCAG 3current model works pretty well.
If we want more abstract outcomes that may be difficult.
<Zakim> mbgower, you wanted to say how do act tests fit into the model?
mbgower: It feels like a bit of gymnastics. Informative test that is a valid check of a sufficient test of normative text.
Andrew: Jeanne, you said that tests are on the outcomes. But in WCAG 3 today, they are on methods.
AWK: they refer to the method which refers to the outcome.
Jeanne: Method is technology specific. Outcome is tech neutral.
AWK: The tests can't rely on the outcomes being normative.
Jeanne: Not what I meant. Stick with current model.
Wilco: Something we may not have fully settled on. What is the relationship between methods and outcomes? If you fail a method, do you fail the outcome?
<mbgower> decision tree?
<Zakim> MichaelC, you wanted to talk about sourcing and to talk about relationships of methods
Michael: 2 things. How we are
going to maintain guidleines. We have trouble maintaining
techniques ourselves. Have tried and failed crowdsourcing
techniques.
... we need to be able to maintain normative.
Maintining methods. I'm not sure if we have ands or orss for a sufficent outcome.
<Zakim> jeanne, you wanted to say that the outcome can be precise without being true/false. and to say that we can get corrections and code samples from crowdsourceing. I think it would
<Wilco> +1 I agree, that wasn't what I was trying to say
Jeanne: the outcome can be precise without being true/false. We will always have pass/fail
<Zakim> Rachael, you wanted to speak to relationship
Rachael: We had discussed the possibility of defining where methods break by what would be "anded" to meet the outcome and tests wtihin the methods woudl be ors.
jenniferS_: Difficulty with outcome concept.
<Zakim> AWK, you wanted to say that it is possible that writing methods becomes easier if the Outcomes are simpler? and to say that there is a little-known relationship between SC and
AWK: Its possible that writing
methods becomes easier if we are supper successful at making
the outcomes simpler. Related to Wilco's question about the
relationship between methods and outcomes. We coudl do a 1 to 1
relationship or what we do today where it is not. A secret
relationship between sufficient techniques which I didn't learn
about till a chair.
... right now some have and and some dont'.
<jeanne> That won't handle different platforms
AWK: Need to make relationship clearer.
Right now I am leaning towards tests are normative.
scribe: My understanding is that
informative help in understanding and normative are how to meet
requirements.
... is a test going to help meet conformance or just aid the
understanding.
... can a normative method have no tests?
AWK: For the last part, I don't think anyone will agree to a method wihtout a test.
<Zakim> MichaelC, you wanted to say the test types discussion changes the method and / or IMO and to not want to repeat WCAG 2 relations and to see method detail vs test detail
Michael: In FPWD, we said Methods must be in OR relationship and you need whatever tests are needed.
<Wilco> 1+
Michael: the and/or relationships are not helpful.
+1 to straightforward model
Fazio: 2 years ago when we put out the draft, we all decided it shoudl map to functional needs. That is why we decided on outcomes. They meet functional needs. The methods are how to acheive that. The tests proved it did that. If we aren't doing that anymore.... but that's how we go here.
<Zakim> jeanne, you wanted to say that the methods don't reuse tests
<jenniferS_> * took a screenshot of what David Fazio said and will use as a reminder going forward. Thank you!
Jeanne: Test reliability work.
Because methods are tech specific, they have to be an Or
relationship unless we restrict by platform.
... we need to avoid repeating tests for outcomes.
... we currently use techniques across SC. We want to keep
outcomes. If complex, break down.
... The method can say what steps are required to meet
outcome.
Shadi: Can I have normative specs wihtout normative tests?
Michael: Yes.
... the requirements are written to be testable but they are
not tests.
... if the requirements are testable that is what conformance
is based on.
... Whenever we take something ot the director, I have to
convince its valid.
<MichaelC> +1
Rachael: Slide 2 shows WCAG 3 model. Still think that is really good. Test might cross methods. But as soon as you or a Test you create a new method.
Wilco: TEsting at page level adds complexity so the scope here matters because at page level this can be diffcult.
Shoudl we talk abotu expectations.
<Zakim> AWK, you wanted to say that another substantial advantage of informative methods is the ability for the set of methods to grow/change/evolve and to ask how we will keep methods
AWK: another substantial
advantage of informative methods is the ability for the set of
methods to grow/change/evolve and to ask how we will keep
methods
... Jeanne said we currently use a lot of techniques to answer
different SC. I think its hard to set a technique that doesn't
meet more than one outcome. For example headings meets more
than 1.
... You are providing headings but also doing bypass blocks. We
will likely have those challenges again. Maybe we can minimize
it by making methods hyper focused.
<Zakim> MichaelC, you wanted to ask about reuse and to ask if outcomes can have scope and to clarify process
Michael: I agree. There are so
many things we can test that will affect mutliple outcomes.
Avoid duplicative methods. May not be able to avoid multiple
tests.
... When AWK says whole process, that means the whole W3C
process. Big lift. Bigger than AG process.
... I wonder if we could say this is an element level outcome.
Process level outcome.
Maybe it solves some problems.
+1 to including scope in method
<Zakim> mbgower, you wanted to say there are ways to still meet an outcome with one technology and to say there are mutliple methhods to meet an outcome using one technology
<jeanne> +1 to including the scope in the method
mbgower: I just added test set next to test 1. May be a logic flow.
<MichaelC> +1 to test set though we´re making a deeper hierarchy ;)
mbgower: I also thought something you said I wanted to clarify. There is going to be more than one method sometimes to reach an outcome.
<jeanne> +1 Mike, because we need to prototype that. So far, it has worked for 1 Method per technology.
mbgower: Michael - One test may support multiple outcomes. I can see parts of the test set being reused but not the full test set.
<Zakim> Rachael, you wanted to answer aWK
AWK: Create a list of potential conflicts to work through whenever we add an outcome.
Michael: Example: test that text
of an image is there may support both image has text and link
purpose is clear.
... I am intrigued by test sets and the possibility that tests
may apply in multple places but test sets do not. I think we
can do so more easily than just tests.
mbgower: I do think methods may overlap outcomes.
AWK: Jeanne was saying we want to avoid that and I agree. Not sure how achievable.
Ack: Wilco
Wilco: How does this impact our ability to update the standard? We've struggled with this in WCAG 2.
<jenniferS_> +1 to Wilco
Wilco: it feels like we are
setting oursleves up for the same problems.
... no suggestions on that.
MichaelC: I agree that is a
problem.
... I'm not sure if its a different problem than writing the
content the first time. Writing or maintaining its a problem.
We lack the comprehensive set of expertise to create
everything. For example we do not speak eveyr language of the
world. We need ways to bring content in that we didn' write. It
needs to be not so burdensome that we can't do it.
AWK: One answer might be that the normative outcomes need to make sure they are more foundational than they are today. So that its easier ot apply broadly.
We tried to add in additional criteria and some were veyr specific.Page markers for example. If we had a broad outcome that it woudl map in so it allows for external additions, it might decrease the time
<jenniferS_> I volunteer to scribe
<jenniferS_> I can hold it for 30m
scribe: also need to update more often.
<Ben_Tillyer> +1
<Wilco> +1
<Zakim> mbgower, you wanted to say the made up method 'Clearly describes' could apply to headings, labels, alts, etc
<jenniferS_> do you want me to scribe this?
Mbgower: A plain language method could cover multiple places and not a bad thing.
<scribe> scribe: jenniferS_
<Zakim> MichaelC, you wanted to say we thought WCAG2 SC were foundational too
<Zakim> jeanne, you wanted to ask what about making the Functional Needs the normative layer?
AWK in wcag2 we though we had written foundational success criteria and not going to need to update it.
AWK we have some foundational concepts but realized it's harder than it sounds.
Jeanne: I'm not totally getting it but think it could work.
AWK I was thinking… there's an outcome that is instead of info & relationships, maybe breaks out into different types of content explicitly…
could we make it clearer to basically say… the semantic structures are correctly conveyed or whatever.
we had principles on the line… in wcag2 principles are normative, well, pretty much all of wcag2 is normative except notes.
MichaelC: well, normative but don't have conformance statements.
Rachael: we just just embrace their non-normative status.
AWK: some want it to be normative… [scribe didn't follow]
referring to jamboard
"other method stuff" is part of methods
Rachael: tests are part of methods, but it's on the other side
AWK: so do all the things associated with methods belong together and methods are normative---orinformative?
Jeanne: might want to pull the tests out and make it normative
it sounds like there are 2 question items: one is of much greater consequence than another in terms of which side they fell for everyone
Rachael: first question for tomorrow!
Shadi: I don't know how that would be testable for the functional needs of the normative req.
* hey! I was still typing minutes!