IRC log of ag on 2022-11-01

Timestamps are in UTC.

14:57:47 [RRSAgent]
RRSAgent has joined #ag
14:57:47 [RRSAgent]
logging to https://www.w3.org/2022/11/01-ag-irc
14:57:49 [Zakim]
RRSAgent, make logs Public
14:57:50 [Zakim]
Meeting: AGWG Teleconference
14:58:02 [Jay]
Jay has joined #ag
14:58:06 [MichaelC]
present+
14:58:11 [Jay]
Present+
14:58:19 [Jennie]
present+
14:58:22 [alastairc]
present+
14:58:24 [jeanne]
present+
14:58:25 [alastairc]
present+ jeanne
14:58:25 [Jennie]
scribe: Jennie
14:58:50 [iankersey]
present+
14:58:55 [Francis_Storr]
Francis_Storr has joined #ag
14:59:07 [GreggVan]
GreggVan has joined #ag
14:59:32 [ShawnT]
present+
15:00:24 [Jennie]
Alastair: If anyone is available to scribe for hour 2, please let me know
15:00:25 [ToddL]
ToddL has joined #ag
15:00:33 [Lauriat]
Lauriat has joined #ag
15:00:52 [Lauriat]
Present+
15:01:09 [Jennie]
Alastair: If anyone can scribe for hour 2, please let me know
15:01:12 [laura]
laura has joined #ag
15:01:16 [AWK]
AWK has joined #ag
15:01:24 [alastairc]
agenda?
15:01:27 [joweismantel]
joweismantel has joined #ag
15:01:29 [wendyreid]
wendyreid has joined #ag
15:01:31 [AWK]
q+ to ask about survey closing
15:01:45 [wendyreid]
present+
15:01:45 [alastairc]
ack AWK
15:01:45 [Zakim]
AWK, you wanted to ask about survey closing
15:02:01 [Jennie]
AWK: I know we have been closing surveys in advance of the meeting. This 1 was closed yesterday.
15:02:12 [Caryn]
Caryn has joined #ag
15:02:13 [Jennie]
Alastair: It was a quarter of an hour ago, but it appears like it was yesterday
15:02:20 [wendyreid]
q+
15:02:22 [Jennie]
AWK: It can be hard to get time to review things
15:02:30 [Jennie]
...I would love additional time.
15:02:38 [kirkwood]
present+
15:02:46 [Caryn]
present+
15:02:51 [laura]
present+ Laura_Carlson
15:02:54 [Jennie]
Wendy R: Apologies - I didn't know I was meant to write a full report.
15:03:05 [mbgower]
mbgower has joined #ag
15:03:07 [Jennie]
...I wasn't able to vote in the survey because I didn't have time to read all the proposals in the time I had
15:03:08 [mbgower]
present+
15:03:08 [kirkwood]
+1 to Michael Cooper regarding closing survey per prevous deadline standard
15:03:14 [Poornima]
Poornima has joined #ag
15:03:15 [SuzanneTaylor]
SuzanneTaylor has joined #ag
15:03:15 [Jennie]
Alastair: We appreciate that
15:03:21 [Poornima]
present+
15:03:27 [SuzanneTaylor]
present+
15:03:31 [Jennie]
...To Andrew's point: it is different for something like this because there is substantial content to review
15:03:32 [ToddL]
present+
15:03:41 [Jennie]
...We will do our best to gather comments today
15:03:49 [JustineP]
JustineP has joined #ag
15:03:52 [Jennie]
...We can share what we have today
15:03:55 [MichaelC]
s/Michael Cooper/@@/
15:04:02 [Jennie]
...That survey format is difficult to read through the results of
15:04:10 [Jennie]
...We are pasting into a separate document
15:04:16 [Jennie]
...Apologies for that
15:04:31 [Jennie]
...In general we are trying to stop the surveys about 30 minutes before the meeting to help structure the meeting
15:04:39 [JakeAbma]
JakeAbma has joined #ag
15:04:42 [JakeAbma]
present+
15:04:50 [Jennie]
...There is a lot to get through, we will do our best
15:04:54 [alastairc]
TOPIC: WCAG 2.2 status update
15:04:57 [Wilco]
Wilco has joined #ag
15:05:06 [Jennie]
Alastair: We are slipping a bit
15:05:14 [Jennie]
...We have implementations for most things
15:05:25 [Jennie]
...We have got through 1 AAA with Funka
15:05:39 [shawn]
shawn has joined #ag
15:05:47 [Jennie]
...If anyone has another site that is AAA to WCAG 2.2 please let us know
15:05:50 [JenStrickland]
JenStrickland has joined #ag
15:05:50 [sarahhorton]
sarahhorton has joined #ag
15:05:54 [JenStrickland]
present+
15:06:00 [sarahhorton]
present+
15:06:02 [Jennie]
...Also, if you are interested in helping with testing, please let us know
15:06:08 [shawn]
present+ ShawnLawtonHenry(first_part)
15:06:14 [alastairc]
q
15:06:15 [alastairc]
q?
15:06:17 [Jennie]
...We will use the WCAG 2.2 meeting on Friday to make further progress on that
15:06:19 [alastairc]
ack wendyreid
15:06:31 [Jennie]
Alastair: Any questions on the WCAG 2.2 side of things?
15:06:36 [shawn]
[ Shawn wonders about LFLegal for AAA ?]
15:06:41 [jaunita_george]
jaunita_george has joined #ag
15:07:01 [Jennie]
Alastair: Do we have anyone new to the group that wants to introduce themselves?
15:07:18 [Jennie]
...Any other topics you would like added to the potential topics list?
15:07:29 [alastairc]
TOPIC: Subgroup check-ins
15:07:59 [Jennie]
Francis: Issue Severity 2 subgroup met last week
15:08:03 [bruce_bailey]
bruce_bailey has joined #ag
15:08:23 [bruce_bailey]
present+
15:08:25 [Jennie]
...We were unsure on the functional needs information
15:08:32 [Jennie]
...We had Joshua O'Connor join us
15:08:40 [Jennie]
...He confirmed that there are a lot of functional needs
15:08:54 [Jennie]
...We decided that, since this is a proof of concept
15:09:12 [Jennie]
...to take the functional needs from the 11 301.549
15:09:19 [Jennie]
...to allow us to carry on
15:09:31 [Jennie]
...looking at issue severity, and working out how they can go together
15:09:37 [Jennie]
...We have another meeting tomorrow morning
15:09:49 [Jennie]
...We are keeping the UK time (normal time for UK people)
15:10:06 [Jennie]
Alastair: The calendar invite I has is on US time
15:10:17 [Jennie]
...members should confirm the time
15:10:27 [shawn]
shawn has left #ag
15:10:27 [Jennie]
Juanite: Testing requirements as methods subgroup met
15:10:35 [Jennie]
...We are making progress on the process types
15:10:42 [Jennie]
...Hopefully we will have more to present in a few weeks
15:10:57 [Jennie]
Jeanne: Silver task force is working on writing outcomes as user needs
15:11:02 [Jennie]
...to see if this is a viable direction
15:11:12 [Jennie]
...Last week's call was about granularity of outcomes
15:11:17 [Jennie]
...We took a homework assignment
15:11:25 [Jennie]
...Makoto did some image types
15:11:30 [Jennie]
...We are writing outcomes for that
15:11:41 [Jennie]
Alastair: Thank you
15:11:45 [Wilco]
q+
15:11:53 [alastairc]
ack Wilco
15:11:59 [Jennie]
Wilco: There is a design WCAG3 subgroup that started last week
15:12:08 [Jennie]
...They are working on mockups for the how-to and the methods
15:12:14 [Jennie]
...They hope to have 1st drafts next week
15:12:19 [alastairc]
TOPIC: Conformance conversation https://www.w3.org/2002/09/wbs/35422/conformance_models/
15:12:38 [Jennie]
Alastair: We have been through several models in the last few weeks
15:12:42 [Jennie]
...I will share my screen
15:12:59 [kirkwood]
present+
15:13:00 [Rachael]
Slide deck that Alastair is sharing: xhttps://docs.google.com/presentation/d/1CvtUu-3h0wx-j-p3MHTsTAuEC3imTU936yDLrLVwISs/edit#slide=id.g17b55e5217a_0_0
15:13:02 [Jennie]
...We will go through each option, and the comments on those options
15:13:11 [Jennie]
...I will start with what has been already added to the survey
15:13:17 [Jennie]
...Then we will open up for other comments
15:13:21 [Rachael]
Comments in a google doc: https://docs.google.com/document/d/1Ekgz4ultg0X8rd5SiYwJ4jbkew77pEULPvIWQrmZLqU/edit#
15:13:28 [Jennie]
...Commenters: please summarize your comments
15:13:55 [Jennie]
...(reads through summaries of models, listed by option number on slide 2)
15:15:01 [Jennie]
...(reads through summary of option 1)
15:15:11 [Detlev]
Detlev has joined #ag
15:15:13 [Jennie]
...David M had some concerns
15:15:26 [Jennie]
...(reads through David's comments)
15:15:50 [Rachael]
q+
15:15:55 [Jennie]
...I assume he is talking about Silver and Gold level tests
15:16:18 [Jennie]
Rachael: I think it probably applies to many of these because it is a base concept
15:16:25 [Jennie]
...I think they will be repeated in several places
15:16:43 [Jennie]
Alastair: We have discussed assertion based things in the past
15:16:55 [kirkwood]
David’s concerns seem valid to me
15:16:56 [Jennie]
...but it is still useful feedback
15:17:11 [Jennie]
Alastair: Moving to Gregg's comments
15:17:42 [Jennie]
GreggVan: the scores were yes or no.
15:17:56 [Jennie]
...I refer to these comments in later reviews
15:18:02 [Jennie]
...Several use "use of views"
15:18:23 [Jennie]
...We tried this the last time around, and it doesn't work because there are an infinite number
15:18:32 [Jennie]
...Example: toolbars pop up, expand and contract
15:18:41 [Jennie]
...Second 1: partial score, or good enough
15:18:49 [Jennie]
...In the example it didn't show any of this
15:19:07 [Jennie]
...This can be marked true if there is enough of them
15:19:12 [Jennie]
...and it counts for conformance
15:19:21 [Jennie]
...I think this is good for showing progress before conforming
15:19:26 [Jennie]
...Or for things beyond conformance
15:19:36 [Jennie]
...The 3rd thing: user needs base - which user?
15:19:48 [Jennie]
...Users have diverse needs. We should talk about specs for the page
15:20:02 [Jennie]
...We should talk about user needs, but most authors have no real understanding of the users
15:20:11 [Jennie]
...You have to tell them what their thing needs to do
15:20:21 [Jennie]
...Before they release a product they need to figure out if they pass
15:20:32 [Jennie]
...The next one had to do with context
15:20:36 [Jennie]
...and adaptive tests
15:20:44 [Jennie]
...Again, this falls into the same category
15:20:56 [Jennie]
...Say if this, then that - make it a conditional test
15:21:04 [Jennie]
...The authors are not going to know what these contexts are
15:21:10 [Jennie]
...And you don't want them defining the context
15:21:23 [Jennie]
...That is not accessibility
15:21:38 [Jennie]
...Extensibility talks about multiple specifications are available for the same tests
15:21:49 [Jennie]
...To say to the authors that there are multiple tests
15:22:03 [Jennie]
...Pounds vs metric is fine because of conversions
15:22:20 [Jennie]
...But different tests could come up with different results? You have to tell me which test to do
15:22:31 [Jennie]
...And maybe it is compound if you need to meet all 3 tests
15:22:40 [Jennie]
...Next: qualitative testing, which I think we need to have
15:22:49 [Jennie]
...But for a basic level of conformance I don't think we can have it
15:23:02 [Jennie]
...Once I ship something, it is frozen for a while. I need to know it is accessible
15:23:10 [Jennie]
...I have to be able to defend it
15:23:16 [Jennie]
...Then, user testing - love this
15:23:30 [Jennie]
...It should always be done, but the results are dependent on the user used to test
15:23:35 [Jennie]
...It is a process, not an outcome
15:23:51 [Jennie]
...To say the testers have to be able to use it, I can find testers that cannot
15:24:18 [Jennie]
...I have tested sites that fail WCAG, but some people that are blind can still use it
15:24:26 [Jennie]
...So I worry about this
15:24:37 [Jennie]
...Those are my concerns
15:24:51 [Jennie]
...There is so much good stuff in these
15:25:40 [Jennie]
Alastair: Bruce is next
15:25:56 [Jennie]
Bruce: What I was trying to get at in regards to extensible
15:26:05 [Jennie]
...When I first read it, I thought it would be easy to get to Silver
15:26:07 [SuzanneTaylor]
q+ to say that when trying to address combined functional needs requiring, for example, 2 out of 5 granular SC (which were called "methods" in option 4) allows us to include, for example, 3 more granular success criteria, rather than having those items completely left out: slightly non-uniform accessibility may make a more accessible internet overall
15:26:21 [Jennie]
...The other options perhaps don't call this out as well
15:26:28 [Jennie]
...I think I need to see more in order to get to Silver
15:26:32 [Jennie]
Alastair: Rachael?
15:26:49 [Jennie]
Rachael: The hardest part of this is shifting the decision making and prioritizing to the working group
15:26:57 [Jennie]
...I think the tree structure would be an extensive lift
15:27:01 [Jennie]
...having to create it
15:27:10 [Jennie]
...I think the biggest gap is not motivating people between levels
15:27:19 [Jennie]
...Something would need to be fixed with that to make this work
15:27:25 [alastairc]
regrets+ Makoto
15:27:37 [Jennie]
Alastair: Makoto said he wasn't sure if this was realistic or not
15:27:49 [Jennie]
...(reads from Makoto's comments)
15:28:27 [alastairc]
https://docs.google.com/document/d/1Ekgz4ultg0X8rd5SiYwJ4jbkew77pEULPvIWQrmZLqU/edit#
15:28:47 [Jennie]
Alastair: This is in the shared folder as well
15:29:11 [Jennie]
...Next is Rain's comment (reads Rain's comment)
15:29:49 [Jennie]
...I think this is a good point, and might be something that could be improved and worked on
15:29:54 [Jennie]
...Jeanne?
15:29:56 [kirkwood]
+1 to Rain’s comment
15:30:18 [Jennie]
Jeanne: In all of my comments, we don't know what the validity, reliability, and complexity is for all of these because we didn't test it
15:30:25 [Jennie]
...option 1 has good ideas
15:30:32 [Jennie]
...It is complex and needs to be simplified
15:30:43 [Jennie]
...I like Bronze, Silver and Gold, but worry about using test types
15:30:59 [Jennie]
...When we did the Silver research, we interviewed people doing Civil Rights in the US
15:31:04 [GreggVan]
q+
15:31:09 [Jennie]
...They felt this was a weakness of WCAG 2
15:31:43 [Jennie]
...When the Civil Rights office was doing their own studies of WCAG for use in legal cases
15:31:55 [alastairc]
q+ to suggest something at silver / gold could over-ride the results at bronze, if it is robust enough (e.g. an industry standard on the same topic).
15:31:56 [Jennie]
...The lawyers thought it was a flaw that testing could be used to stratify
15:32:00 [kirkwood]
+1 to Jeanne’s comment. regarding OCR. Found that as well with my interaction.
15:32:00 [Jennie]
...A, AA, and AAA
15:32:07 [Jennie]
...That it was inherently not a civil rights issue
15:32:14 [Jennie]
...It was more of a convenience, or cost issue
15:32:24 [Jennie]
...A number of things may or may not have been used to decide
15:32:31 [Jennie]
...Their concern was testing, and ease of testing
15:32:37 [GreggVan]
ah ease of testing got it thanks
15:32:38 [bruce_bailey]
@Jeanne -- is this DOJ OCR and is any of the OCR concerns with testing public facing?
15:32:42 [Jennie]
...Because something is hard to test, it shouldn't be AAA
15:33:04 [Jennie]
Kirkwood: I have had this exact interaction as well and +1 to Jeanne
15:33:11 [Jennie]
Alastair: OK Wilco
15:33:25 [Jennie]
Wilco: I think it comes down to: I am not sure this is going to work well
15:33:34 [Jennie]
...I think the 3 errors will be difficult
15:33:46 [Jennie]
...I am not sure this is a good place to put this bar - seems like a policy question
15:33:56 [Jennie]
...Knowing how many errors something is is very tricky
15:34:03 [bruce_bailey]
IMHO there are few AAA SC that are at AAA (and not AA) because they are hard to test.
15:34:11 [Jennie]
...ACT has steered away from this because we are not sure this can be done reliably and consistently
15:34:16 [Jennie]
Alastair: I will put this as a topic
15:34:28 [Jennie]
...It would be useful for people to think about how we talk about perfection being the
15:34:33 [Jennie]
...enemy of the good
15:34:42 [Jennie]
...It becomes a fail, and you cannot progress any further
15:34:59 [Jennie]
...How do we meet requirements around those aspects not becoming a complete blocker to an organization making progress
15:35:05 [Jennie]
...Michael Gower?
15:35:19 [Jennie]
MichaelG: I did not find the ranking (like Gregg) very useful
15:35:26 [Jennie]
...I made my own spreadsheet offline
15:35:34 [Jennie]
...Option 1: main concern is the no more than 3 errors
15:35:42 [Jennie]
...Page or site, that scale is pretty problematic
15:35:44 [mbgower]
https://webaim.org/projects/million/
15:35:56 [Jennie]
...I think most people are familiar with the WebAIM 1 million, just for home pages
15:36:03 [Jennie]
...They have done this every year for many years
15:36:13 [Jennie]
...Average of 50.8 errors per page
15:36:23 [Jennie]
... (reads stats)
15:36:27 [Jennie]
...Remember, this is one page
15:36:33 [Jennie]
...Lots of interesting information
15:36:43 [Jennie]
...If we come up with an error system, we have to assess what is realistic
15:36:48 [alastairc]
q?
15:36:59 [Jennie]
...My overall comment: I have spent several weeks to see if a site can pass 2.2 to AAA
15:37:13 [AWK]
+1 to Mike Gower's point about grounding in reality for conformance.
15:37:18 [Jennie]
...and I am concerned that we are adding stuff
15:37:20 [bruce_bailey]
q+ to ask mg about recent webaim millions. Is 10 errors on a page characterized as pretty good?
15:37:25 [Jennie]
...and people cannot pass now with their home page
15:37:50 [Jennie]
...I think we have to concentrate on how can we make more things automatically testable and understandable
15:38:03 [Jennie]
...And for the WebAIM report - that is with their own automated testing tool
15:38:15 [Jennie]
...And, how can we reward people for exceeding that?
15:38:23 [alastairc]
q?
15:38:26 [Jennie]
Alastair: That's everyone's comments in the survey on option 1
15:38:30 [alastairc]
ack ra
15:38:31 [Rachael]
q-
15:38:35 [alastairc]
ack SuzanneTaylor
15:38:35 [Zakim]
SuzanneTaylor, you wanted to say that when trying to address combined functional needs requiring, for example, 2 out of 5 granular SC (which were called "methods" in option 4)
15:38:36 [bruce_bailey]
+1 to mg wrt need to give props for orgs *starting* with accessiblity
15:38:39 [Zakim]
... allows us to include, for example, 3 more granular success criteria, rather than having those items completely left out: slightly non-uniform accessibility may make a more
15:38:39 [Zakim]
... accessible internet overall
15:38:51 [bruce_bailey]
q-
15:38:55 [Jennie]
SuzanneTaylor: On the concept of not being allowed to say you need to choose 2 out of 5, or 4 out of 10
15:39:05 [Jennie]
...When we are trying to address intersectional needs
15:39:24 [Jennie]
...we might want to have other techniques like visual instructions, cartoon like instructions
15:39:43 [Jennie]
...To do that, there would be so many more success criteria and people are already struggling with the number of success criteria
15:39:57 [Jennie]
...If you have to do 2 out of 5 - and you are the users needing the other 3
15:40:08 [Jennie]
...You will find sites that meet your needs
15:40:23 [alastairc]
q?
15:40:26 [Jennie]
...Rather than leaving things completely out, we should not completely through this out
15:40:27 [alastairc]
ack GreggVan
15:40:46 [Jennie]
GreggVan: One of the things that comes up all the time: a few small errors blocking conformance
15:40:54 [Jennie]
...I think it is important to separate bugs from omits
15:41:11 [Jennie]
...I don't see bugs being a thing for conformance
15:41:18 [Jennie]
...especially for people that fix these bugs
15:41:22 [Jennie]
...But omits are different
15:41:36 [Jennie]
...I think we should admit that there are bugs
15:41:46 [Jennie]
...But we try to repair them as we find them.
15:42:07 [Jennie]
...Omits are people saying I know I am not doing everything
15:42:14 [alastairc]
ack alastairc
15:42:14 [Zakim]
alastairc, you wanted to suggest something at silver / gold could over-ride the results at bronze, if it is robust enough (e.g. an industry standard on the same topic).
15:42:34 [Jennie]
Alastair: Responding to Gregg - I am fairly sure I have heard about people
15:42:49 [Jennie]
...having difficulties even if they are bugs
15:43:08 [Jennie]
...Regarding building on top of bronze - I wonder if this would work in option 1's tree structure
15:43:16 [Jennie]
...Could something in gold
15:43:33 [Jennie]
...might enable you to pass even if you have got a few bugs or whatever at the bronze level
15:43:50 [Jennie]
...Maybe we could construct conformance if you are doing a more difficult process at the silver level
15:43:58 [Jennie]
...to provide flexibility at the bronze level
15:44:10 [Jennie]
...This has taken us 45 minutes for option 1
15:44:18 [Jennie]
...Some of the comments are applicable to the others
15:44:24 [Jennie]
...Any other comments about option 1?
15:44:51 [alastairc]
TOPIC: Conformance option 2
15:45:02 [Jennie]
Alastair: This was around tests and assertions
15:45:15 [Jennie]
...(reads from slide 2 of slide presentation Exploring Conformance Models)
15:45:23 [Jennie]
...(reads like 4)
15:46:31 [Jennie]
...moving on to the comments
15:46:42 [Jennie]
...Gregg's comments are in the previous option
15:47:01 [Jennie]
Bruce: I assume the thresholds are up for discussion everywhere
15:47:04 [Jennie]
Alastair: Definitely true
15:47:23 [Jennie]
Rachael: The combination of percentages and adjectival could hide issues within the complexity
15:47:33 [Jennie]
...I think at the lower level it needs to be really clear
15:47:40 [GreggVan]
q+
15:47:45 [Jennie]
...I am also concerned that the COGA issues would be repeated
15:47:55 [ToddL]
+1 w/Rachael
15:48:02 [Jennie]
...Jeanne"
15:48:11 [kirkwood]
+1 to Rachael’s comment
15:48:22 [Jennie]
...(reads Rain's comment)
15:48:58 [Jennie]
Jeanne: I think there are some good ideas. I like the scoring of the adjectival
15:49:05 [Jennie]
...and separating it out at the end of the process
15:49:10 [Jennie]
...I would like to see that go forward
15:49:21 [Jennie]
...I think a lot of the other testing is being forced into binary/pass faill
15:49:32 [Jennie]
...I found it difficult to untangle how this would work
15:49:40 [Jennie]
...I think it needs to be simplified, and be made consistent
15:50:01 [Jennie]
...The bronze, silver, gold would probably be approved by Civil Rights but needs to be tested
15:50:04 [alastairc]
q?
15:50:07 [alastairc]
ack GreggVan
15:50:09 [Jennie]
Alastair: Gregg
15:50:25 [Jennie]
Gregg: The comment on COGA - one of the major reasons for WCAG 3 -
15:50:35 [alastairc]
q+ on the coga / assertions issue
15:50:39 [Jennie]
...we couldn't figure out ways to get multiple groups' needs included
15:50:55 [Jennie]
...COGA's is so much broader - we need to go beyond what we have known how to do
15:51:07 [Jennie]
...1. We make sure that all of the rest of the things we can't require
15:51:17 [Jennie]
...we get into the guidelines mixed in with everything else
15:51:29 [Jennie]
...2. They don't get in here in a different place, like a AAA
15:51:36 [Jennie]
...You could have the adjectival mixed in
15:51:49 [Jennie]
...Is it there? Is it clear? Can be a higher adjectival in the same provision
15:51:57 [Jennie]
...to make sure the COGA stuff gets in front of them
15:52:10 [Jennie]
...People will consider them because they are right in front of their eyes
15:52:15 [alastairc]
ack alastairc
15:52:15 [Zakim]
alastairc, you wanted to comment on the coga / assertions issue
15:52:26 [Jennie]
Alastair: I have a similar suggestion to Gregg
15:52:42 [Jennie]
...I think there is a danger that a lot of COGA requirements could end up not in the bronze conformance area
15:52:50 [Jennie]
...but it doesn't need to be just the objective test
15:53:00 [alastairc]
q?
15:53:05 [Jennie]
...It shouldn't be part of the conformance model that makes it so
15:53:05 [mbgower]
q+
15:53:11 [alastairc]
ack mbgower
15:53:22 [Jennie]
Mbgower: I want to highlight things I like in this option
15:53:36 [Jennie]
...1. The idea that the focus is on automated, then human as needed
15:53:55 [Jennie]
...2. Using assertion rather than protocol - I am not sure if it is testing, but assertion is easier to grock
15:54:06 [Jennie]
...I had a lot of difficulty understanding the range of measurements
15:54:13 [Jennie]
...I also thought the use of very poor and poor
15:54:26 [Jennie]
...I would like to get to something like poor and insufficient to more clearly show progression
15:54:33 [GreggVan]
q+
15:54:39 [alastairc]
ack GreggVan
15:54:45 [Jennie]
GreggVan: That is for large and small text.
15:54:56 [bruce_bailey]
q+ about adjectives
15:54:57 [Jennie]
...To answer his question about the measurements
15:55:01 [alastairc]
ack bruce_bailey
15:55:08 [alastairc]
ack about
15:55:11 [Jennie]
Bruce: I think I made this comment on a couple of them
15:55:12 [alastairc]
ack adject
15:55:21 [Jennie]
...I know what unacceptable means, but I don't know what poor means
15:55:21 [kirkwood]
+1 to insufficient rather than ‘poor’ don’t like it either
15:55:23 [GreggVan]
q+
15:55:26 [alastairc]
ack GreggVan
15:55:29 [Jennie]
GreggVan: Pass is passing
15:55:33 [Jennie]
...The other 2 are not passing
15:55:48 [Jennie]
...We talked about needing people who haven't yet made it can measure progress towards passing
15:55:53 [bruce_bailey]
+1 for need to measure progress towards passing
15:56:14 [Jennie]
...Example: had many poors before, now have only one poor - to help people see reports internally or externally
15:56:16 [mbgower]
q+ to say that 4.2.3 Recommendations has potential, but I worry about if it can scale
15:56:17 [bruce_bailey]
+1 for rating below pass
15:56:25 [alastairc]
ack mbgower
15:56:25 [Zakim]
mbgower, you wanted to say that 4.2.3 Recommendations has potential, but I worry about if it can scale
15:56:30 [Jennie]
Mbgower: Section 4.2.3 recommendations
15:56:39 [Jennie]
...That includes things that are beneficial but couldn't be required
15:56:47 [Jennie]
...I think this has potential
15:56:57 [Jennie]
...I worry about having some way that scales to how people follow looser guidelines
15:57:04 [Jennie]
...as opposed to specific requirements
15:57:11 [Jennie]
Alastair: That's all the comments on option 2
15:57:20 [Jennie]
...Looking for a scribe change
15:57:23 [mbgower]
scribe: mbgower
15:57:23 [Jennie]
...Can anyone scribe?
15:57:40 [alastairc]
TOPIC: Option 3
15:57:55 [Jennie]
* Thanks Bruce!
15:58:21 [mbgower]
Alastairc: I collapsed very poor and poor into one level
15:58:45 [mbgower]
... although I take Bruce's point about capturing improvement
15:59:09 [mbgower]
... It was also trying to allow for some things not passing and that not instantly being a fail
16:00:22 [mbgower]
Gregg: It was similar to 1, with the changes you talked about. I think there should be 2 below pass, because some have terrible sites. If getting all the way up to pass is too daunting. Gives opportunity to show progress.
16:00:49 [mbgower]
... There may not be 2 levels below because we may not be able to think of 5 levels, but we may have to collapse
16:01:13 [bruce_bailey]
+1 to idea that we need a couple adjectival levels below passing. I just do not think "poor" is one to use.
16:01:16 [mbgower]
Alastair: [reads out Bruce's comment]
16:01:35 [mbgower]
Alastair: Do we think adjectival is a useful way to progress?
16:01:49 [Ben_Tillyer]
Ben_Tillyer has joined #ag
16:01:53 [Ben_Tillyer]
present+
16:02:03 [mbgower]
Rachael: My primary concern is using adjectival at the test level. It can be effective to speed up, but it is hard to get consistency
16:02:16 [Wilco]
+1
16:02:23 [mbgower]
... We played with it using percentages in the FPWD and it didn't seem to work
16:02:42 [mbgower]
Alastair: [Reads Rain's comment]
16:03:04 [mbgower]
Jeanne: The idea of picking the lowest score for the outcome would probably not be valid.
16:03:37 [mbgower]
... I spent a lot of time studying validity. It's much easier to understand, and the bronze/silver/gold would probably be accepted by civil rights experts.
16:03:53 [mbgower]
Alastair: Just so I understand, there's a gated approach where you have to score at least "okay"
16:04:11 [mbgower]
Alastair: Why wouldn't that be valid?
16:04:21 [mbgower]
Jeanne: Validity is how well it reflects the real world.
16:04:41 [mbgower]
... If you're always picking the lowest score, then you're always going to be dragged down.
16:04:57 [Detlev]
I don't get it
16:05:03 [mbgower]
... When you look at the validity, it's going to be too low. You have to have some way of picking the middle. The average.
16:05:22 [mbgower]
Wilco: I'm essentially in agreement with Jeanne and Rachael
16:05:36 [SuzanneTaylor]
q+ to say that lower level quality tiers may not be helpful - It's difficult to imagine a team re-writing alt text to move from very poor to poor, for example
16:06:19 [mbgower]
Alastair: This had skipping methods. The tests would be normative.
16:06:35 [alastairc]
ack SuzanneTaylor
16:06:35 [Zakim]
SuzanneTaylor, you wanted to say that lower level quality tiers may not be helpful - It's difficult to imagine a team re-writing alt text to move from very poor to poor, for
16:06:38 [Zakim]
... example
16:06:39 [mbgower]
... It's a lot to get through with the number of documents we had.
16:06:52 [GreggVan]
q+
16:06:59 [mbgower]
Suzanne: Multiple levels below pass would not be motivating.
16:07:09 [alastairc]
ack GreggVan
16:07:14 [mbgower]
... I don't think they'd alter their alt text to go from very poor to poor.
16:07:17 [Francis_Storr]
present+
16:07:36 [mbgower]
Gregg: If you think about it, even your proposal has 2 levels. You didn't get to poor. You got to poor.
16:07:51 [Jennie]
q+
16:07:52 [mbgower]
... You used percentages. '75% is poor, not passing'
16:07:53 [bruce_bailey]
q+ to say mediocre sites are where the work is, i think
16:08:28 [mbgower]
... when we talk about percentage, we have to be careful. Percentage of what? For contrast, paragraphs, wordings, letters?
16:08:44 [SuzanneTaylor]
I didn't have any percents in my proposal, but my comments are not designed to steer toward my proposal
16:08:45 [mbgower]
... I think they're solvable, but there can be wierdnesses
16:08:47 [alastairc]
q?
16:08:52 [alastairc]
ack Jennie
16:09:30 [mbgower]
Jennie: Responding to the comment on needing more categories below pass, it's important to remember that people could be using these measurements in other ways.
16:09:32 [alastairc]
ack bruce_bailey
16:09:32 [Zakim]
bruce_bailey, you wanted to say mediocre sites are where the work is, i think
16:10:00 [mbgower]
Bruce: I think the opportunity for growth is at the mediocre. That's where we need to juice the standards. To motivate people.
16:10:20 [mbgower]
Bruce: I'm not sure we need to focus on rewarding people who are doing great jobs.
16:10:32 [alastairc]
TOPIC: Option 4
16:10:47 [mbgower]
Alastair: This is the badges option.
16:11:02 [mbgower]
... the example is Access to audio
16:12:00 [mbgower]
... At silver level you need to provide quality outcomes of the bronze levels
16:12:42 [mbgower]
At gold, you'd need to do more things, or gold-specific qualities, like atmospheric music descriptions.
16:12:59 [Regina]
Regina has joined #ag
16:13:07 [mbgower]
Gregg: There's lots of things to like here.
16:13:25 [mbgower]
... It restricts scoring to quality. It introduces badges.
16:13:46 [mbgower]
... It still has views in it, which is a problem.
16:14:11 [mbgower]
... It includes adaptive and extensible, which are problems I talked about before.
16:14:43 [mbgower]
... We should talk about what should be true, not how they do it.
16:15:01 [mbgower]
... I didn't unerstand what 'aggregate outcomes' meant as used.
16:15:13 [mbgower]
s/unerstand/understand
16:15:44 [mbgower]
Alastair: i'm not sure whose option this was.
16:16:27 [mbgower]
Suzanne: The aggregate outcomes are a special kind of outcome where if you have 1 outcome, that's fine, but 20 starts to build up on cognitive load.
16:16:54 [mbgower]
... One irritatingn bump in a road is okay. A bumpy road becomes exhausting.
16:17:14 [mbgower]
Gregg: That was helpful.
16:17:37 [mbgower]
Alastair: Bruce said this approach seems the most distinct.
16:17:55 [mbgower]
Alastair: Rachael your concerns were similar to last time.
16:18:06 [mbgower]
Alastair: [Reads Makoto's comment]
16:18:20 [mbgower]
Alastair: [Reads Rain's comment]
16:18:29 [Ben_Tillyer]
+1 to Rain's comment
16:18:49 [mbgower]
... That sounds like an interesting shift I hadn't thought of.
16:19:03 [mbgower]
Jeanne: This is an innovative appraoch. It has a lot of good ideas.
16:19:09 [kirkwood]
+1 to Rain’s reframing language
16:19:10 [mbgower]
... I love the badges.
16:19:18 [mbgower]
... I think it's more motiviational.
16:19:32 [GreggVan]
+1 to Rain comment about shifting language from what people can or cannot do - to what they need - or needs to be true about the content
16:19:43 [mbgower]
... It might motivate people who are just meeting WCAG today.
16:20:00 [mbgower]
... In 5.1 the scoring test ideas were very strongly disparage in issues filed in FPWD.
16:20:19 [mbgower]
... We probably shouldn't pursue that piece of it
16:20:27 [sarahhorton]
+1 to focusing on guidance for designers and developers
16:20:40 [mbgower]
Wilco: I feel like I don't fully understand the proposal.
16:20:44 [GreggVan]
q+
16:21:32 [mbgower]
... I really liked that if you tried to conform, it became 'required'. So it allowed WCAG 3 to be extensible. So if you make it accessible to kids, you declare it, and you are judged for that. Or for gaming.
16:22:01 [mbgower]
... There is a bit of a risk to that, in that people may not always want to do more. So we need a strong baseline still.
16:22:01 [alastairc]
ack GreggVan
16:22:28 [mbgower]
Gregg: Just one thing. I misread the badges as being a step between silver and gold. I like badges being for particular provisions.
16:22:43 [mbgower]
... Can we make something that encourages people to do steps in the right direction?
16:23:01 [alastairc]
q?
16:23:01 [mbgower]
... I like that badges being at almost a provision level.
16:23:30 [alastairc]
TOPIC: Option 5
16:23:47 [mbgower]
Alastair: "Required or optimised"
16:23:59 [mbgower]
... it uses captions as an example
16:24:30 [mbgower]
For gold, all methods pass.
16:24:49 [mbgower]
Gregg: Highlights: required items bronze; silver and gold progress
16:24:58 [mbgower]
... mixes required with optimized, which I like
16:25:02 [wendyreid]
q+
16:25:13 [mbgower]
... My concern is it passes about 'pass conditional'. I'm not sure what that means.
16:25:20 [mbgower]
... I couldn't find a full write-up
16:25:21 [SuzanneTaylor]
+1 to optimal at the bronze level
16:25:45 [mbgower]
Wendy: I can address if you want to go through comments first
16:26:09 [mbgower]
Bruce: This is the one that was in the speaker notes only?
16:26:19 [mbgower]
Alastair: Yes
16:26:47 [mbgower]
Bruce: I could see how that could go into a Chinese menu approach, which could be very accurate. I do worry about complexity.
16:27:05 [mbgower]
Rachael: I'd like to hear about 'pass conditional'. I'd like to see it fleshed out more.
16:27:20 [mbgower]
Alastair: Rain wasn't sure how scoring worked.
16:27:37 [mbgower]
Jeanne: I really liked this proposal. I want more details.
16:27:49 [mbgower]
... I like the required/optimized categorization
16:28:13 [mbgower]
... We need to keep a careful balance between 'required' and 'optimized'.
16:28:36 [mbgower]
... I didn't like the bronze/silver/gold. I think there are better versions.
16:28:41 [sarahhorton]
+1 to "required" concept as prerequisites to any conformance claim
16:28:49 [mbgower]
... I think this has great potentail as a structural base
16:28:54 [alastairc]
ack wendyreid
16:28:58 [ToddL]
+1 w/Jeanne
16:29:03 [mbgower]
Wendy: Apologies for not fleshing this out more.
16:29:09 [mbgower]
... I can do that.
16:29:21 [mbgower]
... To clarify pass conditional, it is separate from required.
16:29:44 [mbgower]
... I envisioned all things between required and optimized. It helps address disability groups.
16:29:57 [mbgower]
... If it's not present, it has an outsized impact on a user group.
16:30:04 [mbgower]
... so it should be required.
16:30:31 [mbgower]
... Pass conditional. I definitely see many where we have pass/fail. Something is present or not present.
16:30:47 [mbgower]
... There are tests where nuance is required. That's where I see pass conditional coming in.
16:31:09 [mbgower]
... Yes, this thing is here but there is an element of measurability. Yes, this image has alt text, but is this actually good?
16:31:25 [mbgower]
... There is an element of subjectivity, but it allows us to assess quality.
16:31:42 [mbgower]
... And I think it could be on a scale.
16:31:59 [mbgower]
... It would allow implementers to set up their own schemes.
16:32:14 [mbgower]
... I'd love to combine with some of the other proposals.
16:32:35 [mbgower]
... I can definitely flesh this out more. Thanks for the feedback.
16:33:05 [alastairc]
TOPIC: Best aspects to take forward
16:33:54 [mbgower]
Alastair: the best outcome was 'some combination'. I think comments helped with that. What are the best things we can start with and take forward?
16:34:42 [Rachael]
scribe: Rachael
16:35:04 [jeanne]
scribe+ jeanne
16:35:52 [jeanne]
Laura: Option 2 simplicity is more like WCAG2.x, continuing A, AA, AAA but that might be an equity problem
16:36:22 [jeanne]
... possibilty a solution from a proposal from John Foliot.
16:36:51 [jeanne]
Gregg: Best is Option 2. Wants to add Badgees from Option 4
16:37:17 [jeanne]
... I'm leaning toward having only one level below pass
16:37:43 [jeanne]
... Looking at Option 5 of mixing Required and Optimal
16:37:52 [jeanne]
... liked "
16:38:02 [jeanne]
s/... liked "//
16:38:29 [jeanne]
... explore the "criteria of evidence to support the assertion" from OPTION 6
16:39:11 [jeanne]
Alastair: If we use the Badges, then we don't need the levels below Pass because that introduces two motivation systems
16:39:53 [jeanne]
Bruce: Option 4 highest. LIked the changes to adjectival.
16:40:55 [jeanne]
Suzanne: Haptics, 3D printable files, are examples of different types of things at each level. Any of these is a great start and put concepts into the whatever one we do as a base.
16:41:45 [jeanne]
... I see it as a grid. Optimal tests are next with each other. @@
16:42:38 [jeanne]
Alastair reads from Makoto: After I'm done with this survey, I'm feeling that I'd like to see how the WCAG 3 Outcomes/Methods will look like to discuss the conformance model. If we can see how the existing SC in WCAG 2 and additional/new ones will look like in WCAG 3, we might be able to discuss the suitable conformance model more effectively. That's what I'm feeling now.
16:42:38 [jeanne]
...Plus, they might need a lower level than Bronze to promote making digital contents more accessible in countries where web/digital accessibility has not become legal requirements
16:43:26 [jeanne]
Alastair reads from Rain: Unclear on Model 5 so not marking as "don't want" or ranking. There may be something to the conformance that is being suggested here if it is more clearly defined.
16:43:26 [jeanne]
...In general, for *all* of these, I think it would also be helpful to have the conformance rating based more on success with entire tasks or jobs within the product rather than units. Instead of trying to conform to a % of optimal methods, moving more towards, "users can complete all tasks with x-ease" or something more closely tied to success with the actual purpose of the product than with
16:43:27 [jeanne]
individual bits or components.
16:43:27 [jeanne]
...Essentially, it would be great to see the conformance and scoring based more on *outcome* and less on *component interactions.*.
16:43:50 [alastairc]
Jeanne: Option 1 needs simplification but overall structure is excellent. Option 2 has a very good idea for scoring adjectival and for Bronze, Silver, Gold. Option 3 has a very good approach to organizing the testing. Option 4 badges are more motivating than any other proposal and should go forward. The edits to the intro to make them more designer and developer oriented should also go forward. Option 5 required and optimized is a simpler and
16:43:50 [alastairc]
easier to understand way of organizing the guidelines and outcomes.
16:44:56 [SuzanneTaylor]
s/@@/If those setting policy for a specific organization, country etc, are not able to adopt the quality-type tests, they can use the true-false side of the grid for each level. This way the levels are not based on testing. Others might be able to use both sides of the grid.
16:45:28 [jeanne]
Wilco: I like the opt-in requirements in Option 4
16:46:02 [jeanne]
... I worry about lowering the bar. Full conformance should be full conformance, not lower the bar. There are other and better ways that should be in regulations not guidelines
16:46:03 [mbgower]
q+
16:46:15 [kirkwood]
+1 to Wilco
16:46:15 [alastairc]
ack mbgower
16:46:35 [Jay]
+1 to wilco "do not lower the bar"
16:46:47 [jeanne]
jeanne: -1 to Wilco and regulation. It wrecks harmonization and ultimately makes it harder for multi-national organizations
16:47:15 [jaunita_george]
+1
16:47:47 [jeanne]
Mike: This would be a good time for user studies on when things become barries "when a bump becomes a bumpy road".
16:47:58 [jeanne]
... test for things that exceed the baseline
16:48:11 [jeanne]
... testing to help define the baseline
16:48:20 [jeanne]
... thinking about a lot of things
16:49:06 [Rachael]
q+
16:49:09 [jeanne]
Alastair: What of the options should we use as a base? We have support for 2 and 5
16:49:23 [jeanne]
... and drawing in other things like the badges
16:49:26 [alastairc]
ack Rachael
16:49:45 [jeanne]
Rachael: I agree with 2 and 5 and pulling ideas in.
16:49:47 [mbgower]
To summarize what I said, can we do user studies that assess accessible qualities of sites across functional needs, including intersections
16:50:21 [jeanne]
... Suzanne, would you like to explore badges on your own, or do you want to do it within 2 and 5?
16:50:40 [jeanne]
Suzanne: I would like to work within 2 and 5 and develop two different proposals.
16:50:42 [mbgower]
And can we work from that information to come up with computational tests that assess not just the baseline, but help assess attaining results beyond baseline?
16:50:59 [jeanne]
ALastair: We can start subgroups in January
16:51:21 [jeanne]
Rachael: We don't want to lose momentum on this. Survey next week on people who want to work on it.
16:51:40 [jeanne]
Alastair: I have some themes that I think we can continue this discussion.
16:52:12 [jeanne]
... Gregg's comments on Views need to be sorted out separately from this because it can be used in any of the options. It's not a requirement for this conversations
16:52:21 [mbgower]
I found 2, 5 and 6 in particular had ideas worth incorporating
16:52:41 [jeanne]
... the theme of COGA ending up as higher level assertions. Mike's comments can be a solution for it.
16:52:54 [GreggVan]
q+
16:53:03 [jeanne]
... I found @@ an attractive concept
16:53:08 [Rachael]
q+
16:53:36 [alastairc]
ack GreggVan
16:53:43 [jeanne]
... cumulative errors concept would also be helpful
16:54:18 [kirkwood]
+1 to Gregg about COGA language
16:54:39 [jeanne]
Gregg: I want to stop using the language that "COGA is pushed to a higher level". It has nothing to do with them being COGA, it has to do with ease of testing.
16:55:10 [jeanne]
... it is good that COGA and other disabilities get some things that are difficult to test before the eyes of developers.
16:55:29 [alastairc]
The adjectival rating was an attractive, I'd like to find out about people's previous experience.
16:55:31 [alastairc]
q+
16:55:37 [jeanne]
... levels below. I am moving to the camp of onely 1 level below.
16:55:57 [alastairc]
ack Rachael
16:56:12 [jeanne]
... I didn't think that Badges were a way of acknowledging below.
16:56:43 [jeanne]
Rachael: If you have proposals that aren't based on 2 or 5, please propose it. I don't want to close out ideas.
16:56:56 [alastairc]
ack alastairc
16:56:59 [SuzanneTaylor]
q+ to say yes badges were above in the initial proposal, but meeting comments last Tuesday resulted in a comment in the doc about using them for a path from a level below basic compliance to build up to basic compliance
16:57:00 [jeanne]
Alastair: If we started out with 2 and 5, what other ideas would be helpful
16:57:43 [jeanne]
... to answer Gregg, I tend to use COGA as shorthand for the difficulty of making things testable.
16:57:50 [alastairc]
q?
16:57:56 [alastairc]
ack SuzanneTaylor
16:57:56 [Zakim]
SuzanneTaylor, you wanted to say yes badges were above in the initial proposal, but meeting comments last Tuesday resulted in a comment in the doc about using them for a path from
16:57:59 [Zakim]
... a level below basic compliance to build up to basic compliance
16:58:07 [Rachael]
q+ about thanksgiving and christmas schedule
16:58:09 [jeanne]
... I thought using badges for lower rating was a soft thought
16:58:24 [alastairc]
ack Rachael
16:58:29 [jeanne]
Suzanne: I can write a clean doc on badges and base it on comments.
16:58:52 [jeanne]
Topic: Holiday scheduling - please check your calendars
16:58:56 [alastairc]
https://www.w3.org/WAI/GL/wiki/Upcoming_agendas
16:59:27 [jeanne]
Rachael: We will survey next week on US Thanksgiving holiday, and we also are thinking of being on holiday the last two weeks in December.
16:59:34 [Detlev]
present+
16:59:40 [GreggVan]
present+
16:59:41 [ShawnT]
present+
16:59:45 [alastairc]
RRSAgent, make minutes
16:59:45 [RRSAgent]
I have made the request to generate https://www.w3.org/2022/11/01-ag-minutes.html alastairc
16:59:55 [Rachael]
s/we are also thinking of being/we will be
17:08:58 [jamesn]
jamesn has joined #ag
18:23:19 [kirkwood_]
kirkwood_ has joined #AG
18:29:47 [kirkwood]
kirkwood has joined #ag
18:47:46 [jon_avila]
jon_avila has joined #ag
18:48:03 [jon_avila]
rrsagent, draft minutes
18:48:03 [RRSAgent]
I have made the request to generate https://www.w3.org/2022/11/01-ag-minutes.html jon_avila
19:09:18 [kirkwood]
kirkwood has joined #ag
21:02:26 [shawn2]
shawn2 has joined #ag
23:17:13 [SuzanneTaylor]
SuzanneTaylor has joined #ag