Silver Task Force & Community Group

12 August 2022


Azlan, janina, jeanne, Lauriat, maryjom, MichaelC, Poornima, ToddL
Jemma, Makoto
jeanne, Shawn

Meeting minutes

<jeanne> Do we write Outcomes before tests?

<jeanne> s/ Do we write Outcomes before tests?//

<Lauriat> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Scribe_List

AG WG agenda preview

jeanne: AG hosting an extended time mtg next week beginning 10:30 Boston, all on WCAG 2.2. So, FYI!

jeanne: Meeting will run until 13:30 with a break midway.

Chuck_: Notes no WCAG 3 content, but anyone with 2.2 interest, it's important to come!

announce digital publishing salon on 8 September

<jeanne> https://www.w3.org/2022/09/digpubsalon

jeanne: Looking at the future of digital publishing

<jeanne> Tuesday September 8, 8 AM - 12 PM PDT

jeanne: Tuesday 8 September, 08:00-Noon PT (UTC -700)

<Chuck_> janina: Is this during TPAC?

<Chuck_> Jeanne: No, this is the Thursday before TPAC.

janina: Notes confusing calendering!

jeanne: We'll have to find out!

<jeanne> Jeanne will look into the correct time/date

jeanne: Covering many innovative issues which are likely to have an impact on WCAG 3.

janina: +1; knowing about some of those issues!

Use functional needs or user needs to develop the Exploratory material?

<jeanne> Writing Process home page

jeanne: Notes above URI is writing process page

<jeanne> Overview of WCAG3 Writing Process by Maturity Levels

jeanne: Notes last week we stopped on a todo item deep in the weeds -- we'll get there shortly

jeanne: Instead look by maturity models and note grid:

<jeanne> Table

jeanne: Hopes this makes the process more understandable

<Lauriat> Corrected link to table

jeanne: Notes proposal is to use migration doc for 2.x content being migrated -- to bring in as placeholder content

jeanne: i.e. the work is already done

jeanne: Then for exploratory, we'll take user needs identified through migration review and refine them

jeanne: Notes we had specific user needs; very detailed

jeanne: Seems appropriate to Exploratory

jeanne: guidelines may migrate--so we do them later

jeanne: asks whether this explains things?

jeanne: Specifically asks SL

Lauriat: Table breakout is really helpful

Lauriat: Might also help to get into doing some by way of example

jeanne: Functional Needs or User Needs to develop examples?

<jeanne> Last weeks example 1.4.2

jeanne: notes 1.4.2 audio controls

jeanne: Am struggling with the functional needs as currently setup

<Chuck_> Janina: I think brought up a point that is a valid point and it's missing in 2.2.

<Chuck_> Janina: It wasn't specifically an audio control. Navigate semantically.

<Chuck_> Janina: We won't succeed if we have "next" and "previous".

<Chuck_> Janina: There's been more than this for the past 25 years.

<Chuck_> Janina: Not specifically an audio control. Video controls and media in controls in general, don't know if we have it broken out that way.

<Chuck_> Jeanne: Only audio controls by definition.

<Chuck_> Jeanne: Are you suggesting we add a new guideline or subguideline?

<Chuck_> Janina: If we are migrating existing content, let's leave it.

<Chuck_> Janina: We will need a better concept of semantic navigation at some point.

<Chuck_> Janina: CSS is ready to talk to APA about this.

<Chuck_> Jeanne: That gets away from the question.

<Chuck_> Janina: If we are migrating, let's stick to audio. We do need a plan for other stuff, but for now we are migrating what's available.

<Chuck_> Jeanne: For this project, it's to update the writing process. 1.4.2 is just an example for the writing process. to make it more agile. The original writing process was very waterfall.

<Chuck_> Jeanne: We want to make sure that how we are doing the writing process going forward is more agile that maps to the maturity levels that AG has agreed to for how we develop content for WCAG 3.

<Chuck_> Jeanne: 1.4.2 is an example.

<Chuck_> Jeanne: Are functional or user needs sufficient to be developing the user needs we need for the writing process?

jeanne: Notes looking at 1.4.2 as an example of how the writing process works

<Zakim> Lauriat, you wanted to say I think we need the intersections, not one or the other

Lauriat: think we need both functional and user needs in order to capture the important intersections

Lauriat: will help us test what needs coverage

<Chuck_> Shawn touched on my points.

<jeanne> Example of what Errors subgroup did on User Needs

<Chuck_> Janina: In the functional needs or user needs, do they capture...?

<Chuck_> Shawn: Getting more specific is boiling the ocean, and we may cover some things and miss others.

Lauriat: concerned that becoming more specific would be boiling the ocean

Chuck_: Do both have value?

jeanne: reads from user needs

janina: so users need to be able to navigate through semantic structures semantically; not just by time offsets

<Chuck_> +1

<Chuck_> Janina: My response is it's not boiling the ocean, there's a lot of art and specificity built in. If you have semantic structures, you need to cover the details.

<Chuck_> Janina: I can provide a specific example. I obtained a book that had only "next" and "previous" chapter of bible, it's not very usable.

<Chuck_> Jeanne: Are you recommending we write on the user process? Are you suggesting we need more detailed user needs?

<Chuck_> Janina: In the audio example this whole concept was missing. This was a major enhancement that was missing.

<Chuck_> Janina: This was needed in 1996. We can go to specific semantic structures.

<Chuck_> Jeanne: We added this as a new sub guideline in 1.4.2.

<Chuck_> Jeanne: We are using this as an example of the writing process.

jeanne: reassures janina need for semantic nav is captured

<Chuck_> Jeanne: Do we want to use this more detailed user needs at the exploratory level? We could look at it and say maybe it needs to be in the developing level.

jeanne: question is whether this writing approach worksxz

<Chuck_> Jeanne: At some point we do need to go to that level. Where is the question.

Lauriat: looking to see what we had from ux research for existing sc; will be important when we go to wider review

Lauriat: ex. might be nontext contrast; e.g. recognizing buttons and their states

Lauriat: we need to avoid repeating former mistakes

<Zakim> Lauriat, you wanted to ask what research we have from the original SC that we can reference for this

Lauriat: we shouldn't need to redo research that has well informed sc; but also to do research not done

Chuck_: believe we're trying to answer whether we need more detailed user needs at the exploratory level

jeanne: suggesting exploratory should look at existing research

<Zakim> Chuck_, you wanted to ask the question I think we are trying to answer and to

jeanne: then for developing we're writing the detailed user needs

Chuck_: i expect no one individual can understand all the research; but we should have it very available

jeanne: notes section in howto that links to research--so we could mine that as needed

<Chuck_> +1 to shawn's reframing of the need for research

Lauriat: we need research to backup the guidance we put in wcag 3; we need to be evidence based; it's our principle and we need to stick to it

Lauriat: opinions are not as persuasive as research outcomes

Lauriat: even documenting "this sc has no research" is still helpful

<Chuck_> +1 that "no research" is a very useful data point

Lauriat: we will need to work in parallel, of course

<Zakim> Lauriat, you wanted to say we need it to show our work and have our work evidence and research-based, not just that it'll help to reference

<Chuck_> Janina: Want to point out that APA research questions task force has been doing this!

<Chuck_> Janina: I think they've been doing pretty good. Will put link in.

<Lauriat> Excellent! That'll help immensely.


janina: RQTF does this kind of work

Chuck_: We've had just a few of us in this conversation--other people here; any thoughts?

ToddL: Good so far!

jeanne: notes Todd lead errors group

ToddL: Says Sarah! But, I'll take some credit

<maryjom> I'm good so far too.

Do we write Outcomes before tests?

jeanne: what comes first? Outcomes? Test?

jeanne: this topic likely controversial

jeanne: how do we test a user need is met?

jeanne: we can also say there are certain outcomes based on the user need, i.e. what does the outcome need to be to satisfy the user need?

Chuck_: asks what are the pro/con arguments for tests first vs outcomes first

jeanne: original idea is user need; then ask how to tell it's fulfilled

jeanne: based on how it's fulfilled drives defining the outcome

jeanne: based on the user need, what's the outcome we desire and then how do we test that outcome

<Chuck_> Janina: That's where I'm at, I'm having trouble seeing the other approach.

<Zakim> Lauriat, you wanted to ask for a comparative example

jeanne: by example 1.4.2 -- users can customize functionality in settings

jeanne: who to test? then we derive an outcome of what we want

jeanne: or ...

Chuck_: chair hat off -- completely not understanding tests first

Chuck_: because i can see how that builds and works; but not if we say tests first

jeanne: reason is more historic; what we found during wcag 2.1

jeanne: the SC (approx outcomes); tests written later and that exposed errors in the sc

<Zakim> jeanne, you wanted to answer chuck

Lauriat: wanted to build on example because thought it incomplete

Lauriat: users can make audio settings; they're saved; they're applied when users go to something audio

<Chuck_> Under Shawn's example, you need the outcome first. That's how I'm interpreting the statement.

Lauriat: we need to understand what it is we want to test before we can write them; they will be context specific

Lauriat: will be different in trad website than online game

Lauriat: tests will need to be platform specific

<Chuck_> Janina: I think it's correct Shawn. I don't think it invalidates that we discovered gaps, I don't think that this invalidates coming up with outcomes first. APA wants to talk about verifiable credentials.

<Azlan> +1 to using tests to further refine outcomes

<Chuck_> +1 to using tests to further refine outcomes

<Chuck_> Janina: This is a very important conversation to have.

janina: suggests outcomes first doesn't invalidate tests exposing what was missed in defining outcomes

jeanne: agrees

<maryjom> +1 That was what I was going to say - do tests early after outcomes

<Zakim> Lauriat, you wanted to add to that example and to note early and also evolving


<ToddL> +1

<Poornima> +1

<Azlan> +1

maryjom: important to think about testing shortly after writing outcomes, to be sure the outcomes are fully fleshed out

<Zakim> Lauriat, you wanted to note early and also evolving

Lauriat: early writing tests; but we need to reiterate the process as tech evolves; it's a constant cycle

Minutes manually created (not a transcript), formatted by scribe.perl version 192 (Tue Jun 28 16:55:30 2022 UTC).


Failed: s/ Do we write Outcomes before tests?//

Maybe present: Chuck_