<kirkwood> can’t today

<Jennie_Delisi> * I can scribe in 2 minutes

<Jennie_Delisi> * ok! BRB

<julierawe> I could scribe for a couple minutes?

<Rachael> scribe+ julierawe

<ChrisLoiselle> scribe: ChrisLoiselle

Announcements

<Jennie_Delisi> * I am back and ready to scribe if needed

AlastairC: Opening new subgroups soon.

<alastairc> https://www.w3.org/2002/09/wbs/35422/AGWGSubgroupParticipation_Oct_24/

AlastairC: Views subgroup , voice control, text contrast , text to speech are all new sub groups

If you can meet outside of this meeting, please to sign up and join.

Questionnaire will open after the call.

<alastairc> scribe:Jennie_Delisi

Jennie: Scribe+

Requirements Publication Review Kickoff

scribe+ Jennie_Delisi

Chuck: Do we have anyone new on the call?

Alastair: Anyone new on the call?

Georgios: I am new, and from SAP

<Chuck> Welcome Georgios!

Alastair: Anyone else?

Frankie: I have changed affiliation - I am now an invited expert.

Alastair: Anyone else?

<Chuck> Welcome invited expert Frankie!

<alastairc> https://services.w3.org/htmldiff?doc1=https%3A%2F%2Fwww.w3.org%2FTR%2Fwcag-3.0-requirements%2F&doc2=https%3A%2F%2Fraw.githack.com%2Fw3c%2Fwcag3%2Frequirements-updates-2024-09%2Frequirements%2Findex.html

Alastair: Moving on to requirements
... (sharing screen)
... Started working on this for WCAG 3
... Issues have been raised since first working draft
... Previous meetings: reviewed changes. Nothing new in this version
... Might be 1st time reviewing all in one place ...Ignore: anything red or with strikethrough text; yes = changed ...Recommend: start at the introduction
... Issue Severity Area - probably not supposed to be in this document (it is marked in yellow)
... Editor's note in Silver Task Force Research should be deleted.

<alastairc> https://github.com/w3c/wcag3/pull/117 ...Recommend: Encourage you to: review the diff listed in Github
... Anything you spot around the changes please share
... Other items identified will probably be new issues unless a blocker leading to misinterpretation

Julie: Why do some things have arrows in different directions?

Alastair: Arrows down = removed
... green = added
... It has changed since the previous version
... I can give links to the various versions

Glenda: up and down arrows are for the adds, so this may be not only using color alone

Note from scribe (group tried to determine the arrow meanings, and did not appear to have a legend)

<julierawe> Thank you!

<kevin> deleted text is shown in pink with down-arrows (as styled for a <del> element)

<kevin> where there is replacement, it’s shown in green with bi-directional arrows,

<kevin> where there is newly inserted text, it’s yellow with up arrows (<ins> element)

Kevin: In the manual there is a declaration of the different stylings
... I added this into the IRC channel above

Alastair: Any other questions?
... This review will be open for a couple of weeks.
... We will reassess, and look forward to going to a CFC
... At TPAC we looked at a couple of conformance models.
... It would be useful to settle on 1 or 2 to use in the next publication
... This would have heavy caveats that this is in the early stages
... Going into TPAC we had 3 levels (reads from document)
... Prerequisite, Baseline, Enhanced

<kirkwood> “likely to prevent task completion” ?

Alastair: Baseline and Prerequisite are what we were considering required
... (reviewed the section Required plus percentage of enhanced requirements)
... Models differ: what do you do on top of that
... Question around what is applicable

<kirkwood> regarding level 1

Alastair: Would only specify percentage ...Today: want to get down the pros and cons of each model
... We could recommend certain requirements based on sector
... Cons for this one (reads from the document in same section)
... The 2nd bullet applies to all the ones listed in the document

John K: "likely to prevent task completion" - is that true?

Alastair: At that level, might prevent task completion, even with assistive technology support

<Zakim> Rachael, you wanted to speak to changing level definitions

Rachael: We have been continuing to explore the levels in different ways among the chairs
... If we decide to follow the levels idea, we would come back with different proposals on how to define the levels
... Please keep the concept but don't get attached to the definitions

Gregg: If you have 3 different things, and in 3 different levels instead of in one line it will be easier to read
... (under the "Conformance model pros and cons")

Alastair: OK

Graham: There is a section not really needed
... The onus is on us to make the baseline balance

Gregg: I want to reinforce Graham's comment
... You can't have a conformance statement because nothing ever conforms 100%
... This is true of every standard.
... Will you count as not passing because of a bug?
... That is in a regulation area. Ruler vs rule
... If there is a criteria you have to pass to conform...this needs to be considered
... 1 thing wrong shouldn't disqualify

<Zakim> alastairc, you wanted to comment on the con

Alastair: Chair hat off
... Problem of being 1 thing wrong in a complex website, and failing, I think that is a con of these models
... It might be one we want to "live with"
... It was one of our requirements to tackle this problem.
... We could have a document like WCAG EM to outline how you might claim conformance for a product or website
... as opposed to the view concept

<kirkwood> level of effort is not our job

Gregg: I don't think we ever said the unit of evaluation
... previously we had it at "web page"
... You never have an error causing an entire site to fail, unless there is a bug that happens on every page.

Alastair: It has been a page.
... And the way this is forming, it is likely to be the same again
... except for those which are site-wide or product-wide
... This conversation comes up regularly.
... We can take it out as a con.

<Detlev> Single pig fails can fail "complete processes"!

Alastair: It is something we need to be aware of for all of them.

<Detlev> Page not pig

Alastair: That's the 1st model - purely percentage based.
... Levels based one (reading from the section "Levels")
... We would be selecting how people need to progress through the requirements
... Which are part of each level. ...Cons: (reads from the document)

<kirkwood> are we concerned about the potential reduction of WCAG compliance adoption due to increased complexity?

Alastair: (reads John K's question above)
... That might be another con
... (adds that into the document)
... Any other pros and cons?

Gregg: A pro: tiny steps are more motivating, easier to advance to the next level.

Graham: The levels building on each other
... (hard to hear)

Alastair: There may be something at level 2 you cannot meet, which prevents you progressing to the next level

<Zakim> Rachael, you wanted to say con of defined levels is that we run into the issue with A & AA again

Rachael: I think with this approach we risk running into the A, AA problem
... If those setting laws approach similarly, the higher levels get approached more similarly to AAA
... (chair hat off) this is why I am less certain about this approach

Alastair: Moving to the next one "Hybrid"
... (reads from the document
... ) ...Example: you must meet 10% of the enhanced requirements for level 2
... and a higher percentage for the next level, plus everything that met for level 2
... We could pick out requirements that address functional needs
... (reads from the document)

Rachael: If we are at level 1
... and people are working there and the next level says requirement a, b, and c are at level 2
... I would probably work on those in level 1 which are related

Alastair: I will add that

Gregg: Higher levels or future levels?
... Future sounds like something we will do

Rachael: Will we see people pick just the required, even though we haven't structured it that way.

Alastair: If we had 3 outcomes where you need another 10% they would still have to pick a certain number of other ones
... The last of the 4: "Required plus percentage across functional needs"
... Each outcome would need to be categorized by functional need.
... (reads from document pros and cons)

Gregg: term means you take whatever you are doing and carefully select to just do the easiest thing, not the most important thing.

Alastair: Wouldn't it be easier in an open percentage model, our 1st option, compared to this one?
... This one requires some balance across functional needs.

Gregg: Yes and no
... There is also something in each functional category which will be "also helpful"
... I don't think there is any functional area which is all easy stuff
... It may be a bit harder, I can't tell until I see the final.

Alastair: Any other pros or cons?

<Zakim> ChrisLoiselle, you wanted to comment on Return on investment ? Is it different for each need?

Chris L: Looking at these pros and cons

scribe: If I am in a small or large organization is there a return on investment?
... Looking at the difference between 1, 2, and 3
... Is it a combination of getting to 1 and 2?
... Or is it something I need to do to make my product better?
... Or if my product is just a media player are there things that wouldn't be part of the baseline
... but does this restrict my outcome since my product only looks at requirements related to my product.
... Has that been captured?

Alastair: We did have a discussion about that at TPAC.
... It is good to restate.
... Not applicable can massively influence how you do percentages.
... We are talking about how many outcomes are passed or failed.
... We either need to exclude them, or count as possible
... We probably need to look at failures.
... Not counting not applicables would make it impossible for smaller sites to pass

<Graham> The answer to that is that "not applicable" has to be a pass if we want to go in the positive direction of scoring.

Glenda: I like having the functional needs represented.
... In both conformance and WCAG.
... It is not clear enough currently in WCAG. It is in other regulations.
... The biggest pro is to get a score that shows you that you are harming an entire group with a particular disability will motivate behaviour
... Don't people want their product to be useable?

Alastair: This last model - percentage across functional need
... Each outcome would be assigned to 1 or more functional need.
... In the 1st model: just percentage, that is not included.
... It would be up to the author or claimant
... In the levels one: we could balance the functional needs.
... Also we could decide to make that transparent.
... The hybrid is in the middle.

<Zakim> alastairc, you wanted to comment on the org perspective

Alastair: (chair hat off) on the organizational perspective raised by Chris L
... I don't think any model will be different except complexity to meet it, motivation to increase your score

Gregg: Re your last comment
... Separating conformance from reporting - we could say you pass at level x and have a percentage of the rest done
... 5% would be a name, 10% would be a name
... Instead you could just have a percent.
... Many of these things apply to many categories.
... I worry about the functional categories.

<kirkwood> +1 to Glanda. conversely, we may find future regulations to be written around constituent groups (lobby) for particular functional needs. ...1: a huge amount of time deciding what the functional categories are. Example: vision, low vision, etc.
... If I have to pick I could possibly pick specific areas only
... Then avoid a particular area.

<Glenda> I propose using the same functional categories as EN 301 549 ...2: if we have categories, and I notice one areas has more than another, I might pick based on that.

<alastairc> Draft Poll: Preference for conformance model - 1 Percentage, 2 (levels), 3 Hybrid, 4 % and Functional needs ...Also: programmatically available doesn't impact certain areas
... This will get us into battles of counts.

<alastairc> Draft Poll: Preference for conformance model - 1 Percentage, 2 Levels, 3 Hybrid, 4 Percentage with Functional needs

<Detlev> +1 toGregg

Alastair: I have put in a draft poll
... to take the temperature

<Chuck> 3,2,1,4

Alastair: please enter your preference and order of preference

Graham: All 4 use percentages
... While subtle, I think it should change to a point system

<Detlev> other

Graham: some people struggle with percentage
... We should think about this in the future
... That may align to what Gregg was saying.

Alastair: You would give each outcome which passes a number that passes?

<Zakim> ChrisLoiselle, you wanted to comment on category

Graham: yes. For example a score of over 100 gets you to a particular score.

Chris L: to Gregg's point

scribe: If we measure task success
... and the ultimate goal is access for a varied number of users.
... You could have multiple types of users
... in the categories and intersectional needs
... related to one task's success

<Rachael> 1, 3

scribe: for level 1 or 2 are we expecting it is measured that task's success against these user needs?

<Graham> 3,1,4,2

scribe: Or is that up to the organization that is asserting all of us.
... Is that layered in?

<Glenda> +1 to Graham’s idea of points. We could get way more control of impact of meeting specitic requirements. And…the points would not need to = 100 when added together.

scribe: There is a lot which could get into level 1 or 2.

<filippo-zorzi> 4,3,1,2

<filippo-zorzi> +1 on Graham's idea

Gregg: On percent - if there is a WCAG 3.1 with provisions by percent, everything that passed in 3.0 could fail
... because there could be more items
... This supports the idea of points over percentage.
... We could poll separately
... On the points side, same issue
... It could become easier with future versions to get points.

<Glenda> 4, 1, 2, 3

Gregg: And yes, percentage is really difficult for a large portion of the population
... and fractions

<Zakim> alastairc, you wanted to comment on points

Alastair: With points if done in a positive way (more points = better) but you have a simple site it might be difficult to pass
... We may be penalizing simplicity
... I think we have to somehow mitigate that.

<Zakim> Rachael, you wanted to comment on points being harder to write and grow over percent

<kirkwood> +1 to need to not penailize simplicity

Rachael: We can mitigate by giving credit for "not applicable"
... (chair hat off) I think some of this can get very complicated
... If every outcome is comparable to each other we have to think about the technical statements
... It doesn't mean don't do it, we just have to explore it
... We could try to explore with one model

Graham: I agree if you flip it around
... It would have the same outcome effectively
... Re different points for different things
... If we do similar points for similar items it would be identical to percentages

Gregg: I 100% agree we don't want different amounts of points to different things
... Points and percentage are the same

<alastairc> I'd like a principle of "minumising barriers", so that more points is bad.

Gregg: One requires calculation and 1 requires addition
... You can disallow not applicable, or allow on a simple site, and it passes even if not accessible
... Both methods around "not applicable" cause issues
... Sites with fewer items become easier to pass or fail.

<alastairc> q>

<scot> big +1 to gregg's comments right now. largely overlapping what i was going to saqy

<Zakim> Chuck, you wanted to analyze the (limited) results so far

Chuck: we are not making a resolution, just testing to see the preferences of the group
... We had 5 responses
... Option 3 and ?

<scot> i thought we were still talking, not voting yet?

Chuck: Others were honourable limited.
... honorable mentions

<alastairc> Poll: Preference for conformance model: 1 Percentage, 2 Levels, 3 Hybrid, 4 Percentage with Functional needs

* Scribe change?

* Thank yoU!

<Glenda> scribe: Glenda

<GreggVan> 3

Scott: leaning more towards levels instead of percentages (due to ability to game the system with N/A).

<Zakim> Rachael, you wanted to say it is not quite identical because it doesn't account for the number of technical statements

Rachael: we should explore both a points method and a percent method to see how it works out.

<Zakim> alastairc, you wanted to comment on reversing the numbers so it is barriers

Chris: on keyboard subgroup, keyboard needs visible focus (for motor) and keyboard needs name, role, value for screen reader. So impact on passing some things but not others will cause problems for different disabilities.

Alastair: having a score that equates to lack of barriers

Gregg: proposed a new model: Required (A/AA from current WCAG). Enhanced: things you can’t require

<Graham> "required" is our baseline in all of the models though right?

<dan_bjorge> yeah this doesn't sound new compared to every model we've been discussing, maybe I'm misunderstanding

Gregg: get focus on the enhancements - give higher scores for doing enhanced things

Alastair: That is what the models do.

Gregg: need to replace the words “Enhanced requirements” to something not using the words “requirements”

Rachael: we just don’t have a term for that thing yet. We were using “outcomes” but didn’t decide on a new word yet.

Gregg: Assertions are testable. I want to include things that are beyond assertions and beyond requirements.
... suggest the word replacement could be “provisions”
... we are losing recommendations with such a strong focus on reporting on requirements

Break-outs for outcomes and guidelines.

Alastair: if you have any alternatives or pros/cons to have, please do so in the Conformance Model pros & cons

<Chuck> Scribing is now complete, we are now moving into breakout rooms

<alastairc> Outcomes terminology: https://docs.google.com/document/d/1ZAC75CJPRXaplQh3ekxIYrfaZ2jkyTsDSJroxR5F5xM/edit?tab=t.0

<alastairc> Refining: https://docs.google.com/spreadsheets/d/1JgjDonZEvJMc3_k_R6r3siNexeBjigQsqVxvQRhLEN8/edit?gid=0#gid=0

<alastairc> Cateopgirsation: https://docs.google.com/spreadsheets/d/1QwSUxRm96Ez7RWbyqJEtiUa0j0ZbkfvvTpdK8x9Ww1c/edit?gid=1777266264#gid=1777266264

This is scribe.perl Revision VERSION of 2020-12-31
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Default Present: Laura_Carlson, ChrisLoiselle, Francis_Storr, GreggVan, rscano, filippo-zorzi, MJ, giacomo-petri, Azlan, alastairc, Jennie_Delisi, kevin, mbgower, jtoles, dan_bjorge, julierawe, ShawnT, sarahhorton, Glenda, Graham, Detlev, Jen_G, Kimberly, Rain, Rachael, kirkwood, scott, Frankie, wendyreid, ljoakley, JenStrickland, Chuck, no, regrets, AlinaV, nina, maryjom, scot
Present: Laura_Carlson, ChrisLoiselle, Francis_Storr, GreggVan, rscano, filippo-zorzi, MJ, giacomo-petri, Azlan, alastairc, Jennie_Delisi, kevin, mbgower, jtoles, dan_bjorge, julierawe, ShawnT, sarahhorton, Glenda, Graham, Detlev, Jen_G, Kimberly, Rain, Rachael, kirkwood, scott, Frankie, wendyreid, ljoakley, JenStrickland, Chuck, no, regrets, AlinaV, nina, maryjom, scot, jaunita_george
Regrets: RobertoS, Makoto, BruceB, SarahH, RainM, dj
Found Scribe: ChrisLoiselle
Inferring ScribeNick: ChrisLoiselle
Found Scribe: Jennie_Delisi
Inferring ScribeNick: Jennie_Delisi
Found Scribe: Glenda
Inferring ScribeNick: Glenda
Scribes: ChrisLoiselle, Jennie_Delisi, Glenda
ScribeNicks: ChrisLoiselle, Jennie_Delisi, Glenda

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: 

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)