AGWG meeting -- 01 Sep 2020

<laura> Scribe: Laura

https://www.w3.org/WAI/GL/wiki/Scribing_Commands_and_Related_Info

https://www.w3.org/WAI/GL/wiki/Scribe_List

ac: one topic today.

WCAG 3.0 survey

<alastairc> https://w3c.github.io/silver/guidelines/

ac: looking at fpwd
... first round with the group. we are looking for feedback.

<alastairc> https://www.w3.org/2002/09/wbs/35422/silver-fpwd/results

ac: survey is fairly open.
... follows the structure of the document.
... we will discuss comments

Structure

ac: look at substantive comments first.comments first.

<alastairc> https://www.w3.org/2002/09/wbs/35422/silver-fpwd/results#xq3

ac: mostly positive. one person looking for significant updates.
... starting with structure.

Jeanne was positive.

scribe: Jon had several areas that should be updated.
... move pass this one for now.
... jake had a lot of feedback.

jake: sometimes it is wording sometimes more than that.
... some substantial.

MC: focus on substantive today.

ac: Sarah had some significant areas that must be updated before we publish the FPWD.
... thought it was close.

sh: not that significant. ordering and some things are not included. primarily editorial.

ac: Peter mentioned graphic needs alternative text.

<PeterKorn> Thanks Bruce.

<PeterKorn> Sorry; I meant Michael.

mc: I have a branch with the update.

<MichaelC> Pending edits not yet merged: https://raw.githack.com/w3c/silver/editorial_30_aug_2020/guidelines/index.html

ac: df had a comment. ac reads it.
... df says overall structure not easy to understand.

df: haven’t had much time to look at the draft.
... not easy to understand what the structure is.
... doesn’t come out clearly yet as everything is still in flux.

rm: if we move graphic to the top and give an intro would that help?

df: maybe.

<CharlesHall> functional categories are headings of a master list of functional needs. the full list of categories should be added to §2.1 instead of just a single example.

mc: that change is also changed in the new branch.

ac: jake put in a lot of changes thatmay have been addessed.

mg: may need a specific example.
... could be heading versions

wilco: thought we were going for a simpler structure. Why do we need a diagram?

rm: it didn’t get more complicated. but you are seeing the full picture.
... not technical documents are included.
... 2 tracts one for tech folks and one for others.
... you are seeing more of the details.

df: looked at the branch. problem with layers.
... at which point do we need functional needs?

<jeanne> +1 to skip functional needs, they are used in the writing, not conformance.

<Rachael> +1 to remove functional needs from the main hierarchy

mc: some of these my not show up in the spec. we will make info available with tags.
... functional needs will be part of the model.

<Zakim> MichaelC, you wanted to mention tagging, searching

<Zakim> mbgower, you wanted to say Functional Needs has many to many. Why?

mg: many to many relationships. Is that what we mean?

<MichaelC> I´m ok with removing functional needs from the doc, but consider it important to the background materials

rm: Yes, it is.

mg: may get some friction on that.

ac: We have some support to removing functional needs from this document.

df: functional outcome more fine grained.
... 2 different functional outcome could map to one functional need.

js: would be simpler if it could work.
... don’t want to repeat but we have intersectional needs.
... thought we should not be repeating the same information.

jake: we might end up with a lot of duplicate methods.
... need to show that it works this way.
... don’t bloat the document too much.

rm: clear sense of where we are going.

Evaluation

<alastairc> https://www.w3.org/2002/09/wbs/35422/silver-fpwd/results#xq6

ac: yes jake has missed some stuff.

rm: we have not written all of the methods. so we need something at the functional outcone level.
... we will address this.

ac: functional outcomes seem to be akin to SCs

rm: agree.

ac: peter wanted some form of example
... wonder where that would be appropriate. Could be overwhelming.

peter: not sure where it would be.
... need to see where it all works.

<david-macdonald> presnt+

ac: Sarah said “Traditional Tests” as a term isn’t clearly defined, except by referencing WCAG 2.x tests, which assumes familiarity with WCAG 2.x tests. A few examples would help, as is done in the Critical Errors section.

<Rachael> +1 to a different term.

ac: might be able to do that later.
... need a few examples.

peter: need to look at abstract and status.
... should have a survey on status.

mc: abstract has been drafted. but is not on the survey. The abstract shouldn’t say anything that is not in the doc.
... will write status section later.

ac: editors will take a look at Sarah’s comments.

<Sheri_B-H__> +1 to that comment

ac: peter’s comments. view" is a very important, new concept in WCAG 3 and needs deeper definition
... comment on large websites.

<Sheri_B-H__> doesn't the sampling verbiage take care of the complex sites?

ac: what kind of things should it address?
... develop a score card for task conformance. Not seeing important comments from silver research.
... can’t be silent on all the comments.
... need to say here are the things that have not been done. and that we will address A B and C.
... will help to put in notes.
... how does bronze relate?

<Zakim> MichaelC, you wanted to say welcome additional material and to say comparison with WCAG 2 should mostly be in ednotes

ac: how it is scored may not be an equivalent to 2.x

mc: if peter has vision for a statement we can work with it.
... don’t want to put the doc in a defensive posture.

jake: good, better, best and not bronze silver, gold.
... when is the time to mention that?

rm: it you conform to 2.0 you will likely conform to bronze.
... some tests are different.

not 1 to 1 relationship.

scribe: if jake want to write an alternate model we can link to it.

<Zakim> PeterKorn, you wanted to whether/how Silver is presented in contrast to WCAG 2

ac: a site that may be pretty good but fails 2.0 but passes 3.0.

<Rachael> +1 to making that more clear

peter: 6.6.1 It is not clear that it is not backwards compatible. but later says may conform.

<PeterKorn> "Content that conforms to WCAG 2.x generally can conform to the Bronze level of WCAG 3.0, though WCAG 3-specific evaluation is still needed."

peter: needs to say not backwards compatible.
... happy to draft text. What is the timeline?

<jeanne> +1 to clarify add backward compatible to 6.6.1. It's on my list.

mc: 1 to 2 weeks.

dm: hoping that a site that passes wcag 2 would pass 3.0. we have put in a lot of work into wcag 2.

<Lauriat> The Requirements call this out explicitly at https://w3c.github.io/silver/requirements/#comparison-to-wcag-2-x-requirements

ac: Some aspects will be the same but some may be new.

<Francis_Storr> @alastairc I've not scribed before but will do it if we can do intros at the top of the hour :)

rm: we had substantial conformance but took it out.

<Levon> I'm also new here :)

<alastairc> scribe:Francis_Storr

<alastairc> https://www.w3.org/WAI/GL/wiki/Scribing_Commands_and_Related_Info

<alastairc> AC: Comment goes here

<alastairc> ... carry on the comment

pk: not sure we need a substantially conformed below bronze but need something that shows the lived experience of people using the web.
... it's impractical or impossible to test a massive site.

<CharlesHall> lead (the alchemy version of a metal that aspires to be another metal)

pk: we need a new term to describe not meeting bronze level.

<Zakim> Rachael, you wanted to say that is the purpose of the holistic tests and critical failures. Do we need text to explain that?

rm: we have the concept in WCAG 3 that it addresses real-life issues.
... it's the "massive site" question: how to we make sure we've assessed enough of a massive site.

<laura> s/that may have been addressed/that may have been addressed/

<laura> s/simpler stucture/simpler structure/

pk: appreciates the work done to create critical path concepts.

js: do we need a new conformance level?

pk: as written today, bronze says "passes traditional tests". If we change bronze to say "in situations where you can't achieve bronze, here's an alternative way to do that".

js: wants to get into more detail with this but not right now.

pk: we should look at the Spoons metaphor as we look further at conformance.

<Zakim> Lauriat, you wanted to mention the Regulatory Environment and Scope requirements: https://w3c.github.io/silver/requirements/#regulatory-environment

sl: brings up the regulatory environment. If someone has a large thing (site, etc.) we still need to support this.
... we can to the conclusion that we'll need supplementary documentation to describe support for "differently shaped and sized" things.

<Zakim> alastairc, you wanted to ask if site-wide conformance (large site) is more of a silver conformance level thing?

ja: it almost seems like you can do all the AAA criteria at the moment and not meet Bronze level. We should look at this more.

pk: what sampling does is gives you confidence that the whole site is like your sample.

<alastairc> ac: Suggest that 'substantial conformance' would come with extra scoping and be applied at a Silver/Gold level, rather than being 'sub-bronze'.

pk: it says that no matter how large the site, no page can be below a score. This doesn't match reality. We need a term that says something like "the majority of the site meets this level but there will be areas where it doesn't".

<jeanne> Peter, the Methods and the their scoring address Peter's concern about perfection.

<Zakim> Rachael, you wanted to answer Jake.

sl: agrees that we need some additional documentation like WCAG-EM, updated for Silver.

rm: there's no A, AA, AAA in WCAG 3.
... if we want to add this back in, we need more tests.

ja: at the moment it says "traditional tests" which are similar to A, AA, AAA.

<Zakim> Rachael, you wanted to say that we would have to decide whether to include the current AAA as a requirement or drop it

ja: it seems like now we can have AAA criteria in WCAG 3 but it will only give a bronze level of conformance.

rm: that concept of above and beyond testing to get to bronze isn't in the current document.

<JakeAbma> Traditional tests: Those are similar to the tests for A, AA, and AAA Success Criteria in WCAG 2.x. (These tests are used to reach a Bronze rating.) Holistic tests: User-centered design methods, assistive technology testing, user and expert usability testing. (These tests are used to reach a Silver or Gold rating.)

<Lauriat> -1, "traditional tests" ≠ "WCAG tests"

dm: is bronze the same as AAA?

<Rachael> +1 to rewording that sentence.

<alastairc> ackala

<Zakim> alastairc, you wanted to ask about within-method scoring

<jeanne> +1 to reword the sentence

<jeanne> +1 to alastair - flexibility in Methods scoring

<Lauriat> +1

<JakeAbma> +1 to Alaistair, the adjectival rating like tests would be my proposal

ac: it's worth going through the detail of the methods to see if we need some kind of AAA requirement.

js: it's important to note that we're moving away from the "100% or nothing" from WCAG 2.

<laura> s/misspelling/correctspelling/

js: e.g. captioning in XR: some of the outcomes we're recommending are in the future based on where the industry is going. There's a much more flexible scoring system.
... we're going to include all of the AAA but we're not going to force people to do them all.
... if something is on a critical path if you hit a critical failure then you fail. If you don't then there are more nuances: did you do a good, average, outstanding, etc. job?
... there's lots of detail still to work out with these.

Conformance

<laura> s/first.comments first /first. /

<alastairc> https://www.w3.org/2002/09/wbs/35422/silver-fpwd/results#xq7

ac: at the bronze level would you need to use any specific testing technology? Or is it just needed in Silver and Gold?

rm: we need to flush this out more and have an editor's note to say this.

sh: need to make sure the language is as clear as possible in the document.

ac: if there's anything that stands out, let Jeanne, Shaun know. Best thing is to email the chairs so they can forward.

<kirkwood> +1 to Janina regarding conformance claims

JS: Technology changes so much so quickly. We need to think more deeply on how conformance functions. Don't think section 6 has enough detail.

<alastairc> https://raw.githack.com/w3c/silver/editorial_30_aug_2020/guidelines/index.html#example-conformance-claim

<Glenda> I totally agree. A conformance claim is a “snapshot” and the date it was conducted is critical to include.

ac: one of the things we discussed in the deep dives is whether conformance deprecates over time. Decided that a conformance claim is a snapshot in time.

<Lauriat> +1 to Alastair, transparency should cover this.

<kirkwood> Think Janina addressed the issue of ongoing conformance. They are making a repsentation that they are tracking conformance.

JS: it won't be possible to reconstruct a site so it can be checked to see if the conformance claim was accurate.

gs: conformance claim date is critical.

<Rachael> https://www.w3.org/TR/WCAG20/

pk: the example conformance claim would be much more useful to have a claim example where there's a site large enough that all the pages can't be seen.

<kirkwood> +1 to Peter for example of conformance claim

<Lauriat> +1 to examples, especially in relation to the "view" type concept expanding how that can work.

<alastairc> Editor's note for having a larger example

rm: we can add an editor's note for this.

Introduction

ac: what's the next step for public review? Do we send people to github to raise issues?

pk: the statement that we worked with policy makers is an unsupported assertion.

ac: can we add a link to supporting content for this?

<bruce_bailey> @Peter, Access Board makes policy (regulation).

<bruce_bailey> Agencies implement policy, the Access Board does not really *implement* policies (except at our own agency).

<alastairc> Any ideas for replacement term for "traditional" tests?

dm: traditional vs. holistic tests. Traditional sounds unhappy, not progressive. Holistic sounds better. Suggests replacement for "traditional"?

<laura> WCAG 2X tests?

<PeterKorn> I need to drop a couple of minutes early.

<sarahhorton> "core tests"?

<Lauriat> We had many +1s to replacing the terms, especially since we'll create more of both (so both will have new tests).

<kirkwood> 2.X

<laura> classic?

rm: we had "atomic" but people didn't understand that so we went with "traditional" but we can change this.

<Rachael> +1 to including suggested text and edits.

dm: where's the best place to make suggestions?

rm: email chairs list and copy Jeanne and Shaun.

js: also email Silver list so they're part of the conversation.

- DRAFT -

AGWG meeting

01 Sep 2020

Attendees

Contents

WCAG 3.0 survey

Structure

Evaluation

Conformance

Introduction

Summary of Action Items

Summary of Resolutions

Scribe.perl diagnostic output