<scribe> scribe: ChrisLoiselle

<Makoto> Alt Text Subgroup Wiki:

Sub-group check-ins

<jeanne> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Main_Page#Sub_Groups

<Makoto> Alt Text Subgroup Wiki: https://www.w3.org/WAI/GL/task-forces/silver/wiki/Alt_Text_Subgroup

Jeanne: I moved main information into sub group pages to declutter extra links.

<bruce_bailey> no activity to report for audio description

Makoto: I edited the wiki. Next steps would be write draft of guideline explainer and more methods.
... I will then share experiences in how to section and then move forward as a group. I've met Nicolas who wants to contribute to web accessibility, he is in university. I pointed him to WCAG references and will talk to him further if he is interested further.

Jeanne: It would be great to reference old participants in group as well.

<Chuck> +1 full name for me

<MikeCrabb> +1 full name for me too

ChrisLoiselle: For Visual contrast, Andy is working on transition of formulas into MathML and reviewing Bruce's work on the visual contrast table ratios.

<CharlesHall> +1 full

Jeanne: Would you want full name on the respective pages? +1 for yes

+1

<Fazio> Full name for me

<Francis_Storr> +1

<bruce_bailey> +1 full

<kirkwood> +1

Jeanne: You don't have to answer publicly, you can just send me a message.

<JF> +1 to full name

Jeanne: Clear language, is Jan here?

Jan: We have been meeting. A couple of new members are joining as well. John Rochford for example.
... We will continue our work and meet next week as well.

Jeanne: XR?

Mike Crabb: XR met yesterday. Third meeting in total. We are looking at guideline for captioning and mixed reality. We looked at user needs that Josh had put together on XR user requirements.

<MikeCrabb> XR Group - User Needs analysis on Github https://github.com/w3c/silver/projects/2

Mike Crabb: places link to github that walks through these. We will then will be moving on to functional outcomes and tests to meet user needs. Next meeting Monday 15th at 9am Boston time. Feel free to join.

Jeanne and Mike: Mike states he can showcase how to use the projects in github and record a session if needed for all to consume.

<JakeAbma> https://docs.google.com/spreadsheets/d/1iCJfyMtcsSq7GHmwnc4aTNguadRfGDa0H8FBZMaJpcQ/edit#gid=2091469352

Jake: On testing. I can talk to scoring. I tested JF's pages.

Jake presents Benchmark Scoring Experiment 2 - Silver google doc

My question to myself was if I test John's pages, do I get same results? Took a look at the heading tests. I got results back. The results were pretty convincing.

It seemed that it worked as expected, which was a good result. It also raised a lot of questions on approach, which we can talk to later on.

<Jan> I have to drop for another meeting. It's great to hear all of the progress!

MichaelCooper: We will get a meeting together this week to move forward. We will discuss how we work together. This will be a joint group with APA group so we are all coordinating together.

<Jan> +1 full name

Jeanne: To MichaelCooper: Can you list out who is attending? Yes. David F, Me, Charles H.

JF: Volunteers if needed, Michael will follow up later in process.

Jeanne: We have new people on call, lets talk to these people so they can introduce themselves.

<Fazio> It's in its infancy but yes

Sheri Byrne-Haber: Hi. I'm the lead for VMware. I'm interested in the maturity model framework. Jeanne: Welcome. Looking forward to include your interests and moving forward on those.

<CharlesHall> *waves enthusiastically at new participants*

Jeanne: For Maturity model work, lets discuss interest. JF is interested as DavidF noted.

Omar Bonilla: I'm been part of group for about of month, part of Thompson Reuters, I do a little bit of everything.

Francis Storr: I work at Intel. I started about 6 years ago with them. I work in UX and Design lead working to build their program. Jeanne: Feel free to review our work and volunteer where you are interested.

Jeanne: Any other new members on call?

Starting a new guideline

<OmarBonilla> Thanks!

Jeanne: Our commitment to AGWG , we'd have six guidelines for working draft. We have 5 currently. We need one more success criteria / guideline.

<Fazio> No thx

Jeanne: Would you want to talk to defining Inputs as a guideline?

Jake: To summarize...when I did the exercise when judging headings on JF's heading files, I came across different functional types. JF's big table on functional disability was present on each page on JF's proposal. From testing to adjectival rating questions, I tried to apply which questions applied to headings.

<Fazio> Yes, COGA

Jake: Some questions left out some disability types. To make it concrete, the more I asked questions about headings, the more uncomfortable I became about heading guideline. Low vision, no vision, learning disabilities, color blind , all have different needs. Opportunity of functional outcome needs to be present for all users.

<JF> Meeting color contrast is a functional outcome

<SBH> I am interested in headings also. VMware is working on a machine learning based tool to analyze headings automatically

Headings would need focus order, contrast, coga, etc.

<SBH> if we get it working, it will be released as open source

Jake: Outline for headings could be in structure guideline, but how would landmarks fit in to that guideline? If landmarks are optional , how does that become pass / fail test ?

So take Input guideline , it is not clear the structure of new guidelines would fit into. Maybe guideline conditions are needed? Color contrast is a condition of a guideline?

<CharlesHall> the original intent of overlapping guidelines was to solve it via information architecture and tagging those things like landmarks and clear language within one guideline to link to another.

We are going for a flat organization vs. principle, operable, understandable, robust....will we have variant that ultimately pulls these principles back in.

<SBH> In the legal world, we call this a "condition precedent" (which I know is not plain language). Something that must exist before something else can take place.

Jake: I'd like to get some answers back , as if we are getting into scoring, the functional outcomes that are available for other guidelines, are scored in similar way. I.e. one guideline has 2 outcomes and another guideline has more than 2, how is that scored? Normalization of scoring is needed. Weighting needed?

<SBH> Toast messages have 11 guidelines implicated FWIW

Jeanne: Right now, I'd like see if people would be interested in addressing these issues in a new guideline called Inputs.

JF: One of the things that you confirmed to me is we need to figure out what a point is worth. Functional outcomes. Additive model of more things you do , increases a score. What does a point mean? How many points is needed to get to a functional success level?
... Binary of pass / fail vs. functional outcome is based on points. Forms , multiple things applied to forms. Some techniques are better than others. More mature technique that has better ability to increase accessibility for all users, that would equate to a better score by using that technique.

<JF> we are working on Guidelines ^AND^ a Conformance Model

Jake: Headings are binary, but some are more severe than others. We are working on guidelines, but I'm not sure of structure. Do we need a guideline? Or is it a guideline condition? Atomic to apply to headings.

<CharlesHall> could the unit of points be percentages instead and variable to a possible 100% within each guideline? {functional outcome number} / {tests} = possible percent?

Jake: Reasons for and against site map or table of contents, we don't have these for "X" reason(s). This is why they don't apply this is why they apply. The user may not always need a heading from a user standpoint.
... Are conditions the guidelines after the exercise I completed? JF: The scoring piece is needed but has not been built up yet.

JF: From Coga, it comes back to , does it work for intended audiences? The techniques and all the right things are done well, i.e. its a great chocolate cake, but what are the ingredients to build the cake?

Jake and JF: Atomic rules are needed and then assigning values to the rules.

Jake: Guideline conditions, like contrast, semantics, language etc. Are the conditions becoming guidelines?
... Say color contrast not being part of headings, perhaps outcomes become the guidelines?

DavidF: I like the addition of POUR principles to the testing / evaluations. Making something Perceivable to various senses, then functional needs are met. The more ways you do that, the greater the score should be. Has that done with currency or scoring ? Is this done with testing methods or AGWG? It is important

<JF> +1 to weighted scoring

DavidF: Let me know if I caught everything you stated.

CharlesH: I don't think the atomic unit has to be static or normalized. I think the percentage or point is variable based on guideline.

From CharlesH: could the unit of points be percentages instead and variable to a possible 100% within each guideline? {functional outcome number} / {tests} = possible percent?

CharlesH: I think that is a plausible approach to scoring.

JF: It comes down to points structure and value.

Which comes back to functional outcomes and contribution to accessibility.

<Lauriat> +1 to Jake's earlier point (for lack of a better term) to create the atomic tests first, before we figure out the details of points they award.

<SBH> I agree with one caveat - not all guideline conditions are created equally. There should be higher weighting for A than AA

I can't hear JF due to the screen reader reading. Sorry.

<Fazio> it has COGA impact too

<Fazio> must be able to decode sentences in a certain amount of time to string them together in comprehension

<Fazio> +1 SL

JF: Audio description is hard that should possible give more points to doing that vs. having same weight as lang equals English. ShawnL: It should be based on functional outcome. JF: The functional value of achievement is key.

<CharlesHall> +1 to impact on humans

<CharlesHall> -1 to impact on authors

<janina> I don't agree that page lang definition is superfluous. That is monolingual thinking, and we need to consider a multilingual universe.

<Fazio> I don't think they're divergent

<KimD> +1 to Janina.

Jeanne: Can you write something up on your proposal on where group is today? Do we have two divergent proposals or modifications to Jake's proposal? We would need to go through these and compare differences.

<Fazio> I still think a bullet point list from JF would help

JF: I think I'm asking same questions as Jake's questions. I.e. pass / fail vs. compound requirements. I.e. what is a point worth? Jeanne: We have been on percentages for a while. JF: Everything can be expressed as a percentages. Why is 17 a magic number? I.e. it is worth something and achieved something of functional value.

Jeanne: In current proposal, everything is worth "1". JF: I disagree. Jeanne: Please write it down so we can review proposal for core questions.

<Rachael> I think that we will likely in the long term weight guidelines but starting with everything at 1 point gives us a useful starting place.

<KimD> +1 to MC

MichaelCooper: I think writing down proposed answers for JF's questions would be helpful so we can talk through them. DavidF: I agree, it would make it more consumable

This is scribe.perl Revision of Date 
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/Guyidelines/Guidelines/
Present: CharlesHall ChrisLoiselle Chuck Fazio Francis_Storr JF JakeAbma Jan Joshue108 KimD Lauriat Makoto MikeCrabb OmarBonilla Rachael bruce_bailey janina jeanne kirkwood SheriByrne-Haber
Found Scribe: ChrisLoiselle
Inferring ScribeNick: ChrisLoiselle

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: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


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.)