Silver Community Group Teleconference -- 15 Sep 2019

<Chuck> https://www.w3.org/2017/08/telecon-info_silver-tpac

<Lauriat> Link to the agenda, call-in info, etc.: https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan

<JohnRochford> https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan

<AWK> +AWK

<Chuck> scribe: Chuck

Lauriat: We will build energy during the day.
... Before agenda, I'm linking to agenda and reviewing our plan for today and tomorrow.

<Lauriat> Agenda and stuff: https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan

Lauriat: This has plans for the week. One of the things is agenda. How to assess conformance model prototypes.
... <reviews written agenda>
... Thank you Jeanne for taking part from your timezone.

<Lauriat> How to evaluate Silver conformance proposals: https://docs.google.com/document/d/13A8mGMnQujfEVqcw_LmAUYT8DDq_qW0TNcHxmCHd0io/edit

Lauriat: This is our working document.
... We've been looking (catching everybody up) working on silver conformance model types. We need to create a framework to evaluate the models.
... Models aren't straight forward. We want to make sure that the model we make meets the silver requirements.

<Lauriat> Silver Requirements: https://w3c.github.io/silver/requirements/

Lauriat: We've taken each requirement from the doc that has something to do with conformance and added to working doc.
... We'll assess how well the model supports multiple ways to measure, for example, or how we test that.
... Jeanne jump in where you left off.

Leonie: On multiple ways to measure...

<Lauriat> Multiple ways to measure requirement: All Silver guidance has tests or procedures so that the results can be verified. In addition to the current true/false success criteria, other ways of measuring (for example, rubrics, sliding scale, task-completion, user research with people with disabilities, and more) can be used where appropriate so that more needs of people with disabilities can be included.

Lauriat: from the silver requirements.

https://w3c.github.io/silver/requirements/

Lauriat: One of the things we noted, some examples are true false, cognitive walkthrough, etc.
... We need to test that we serve more people with disabilities.
... Test that 4 the sc from coga will work in silver.

Charles: Do we need to define more needs? We are trying to stay way from measuring silver vs. wcag. If wcag did 2, does silver have to do 3?

Lauriat: Grey area.

Charles: Benefits and understanding docs quantified ... pointing out "more needs"... is a requirement. What does "more" mean?

Lauriat: Should we add that as a third thing? We have 2 things to test. More ways to measure than just true false. Second is that we serve more.

Alastiar: You can include wcag and more that meets other/more needs.

Katie: "At least" instead of adding a number. You don't know. We may end up with guidance that's user specific.

Charles: That's another thing. A particular group of sc's may have been for one functional need.

Lauriat: We can add another bullet to compare what's in the understanding doc and assess if we meet more than that.
... With content migration to silver, we've been working to take current sc's, group them for related functional needs (captions for pre-recorded and captions for live as an example).
... We are building up functional needs for listing who is helped and in what way.
... The understanding docs have good content to guide us.

Katie: Can we leave that open to address XR?
... Even though we are using existing content, we want to allow growth.

Charles: B statement can be modified to say the measurement of the conformance model is that it serves needs of more people and more context.

Katie: That covers it. I like that.

Lauriat: <updating doc immediately>
... For A, we have ... if we have additional tests beyond true/false, then we can say we met this criteria.
... For b: ... ask for feedback from task forces, and compare wcag and silver and see if we have met more.
... Anything else?
... Done.

RESOLUTION: For multiple ways to measure, we have the things to look at to test.

Andrew: Struggling with multiple ways to measure and B, that we serve more people and more context.

Lauriat: That calls out the last part of that requirement. Can be used where appriopropriate so that we meet more needs.

Katie: This is choosing which conformance model.

Lauriat: This is the test for how we test things.

Andrew: Related to Charles' comment, maybe what we need to do is clearly document with the tests what end user needs are being met. But that's what you are getting at in 3rd bullet.

<Lauriat> Next Requirement: 3.2 Flexible maintenance and extensibility

Lauriat: Next requirement. Silver task force has gone through many of these. We have some things drafted.
... We can add to them.
... Flexible maintenance and extensibility.

<Lauriat> Create a maintenance and extensibility model for guidelines that can better meet the needs of people with disabilities using emerging technologies and interactions. The process of developing the guidance includes experts in the technology.

Lauriat: <reads>

<Lauriat> a. Pick at least 2 emerging technologies, take success criteria for those through the entire cycle,

Lauriat: We have a few tests in here <pasting...>

<Lauriat> b. Demonstrate our work product with SMEs in the area of that emerging technology.

<Lauriat> c. If our work product meets with their satisfaction, and we have more than one, we would call that a success.

Lauriat: I'll join those together, seems like the same thing.

Andrew: Emerging technologies and emerging interactions?

Lauriat: We have 2 tests for this req.
... Giving it a practice run through maintenance.
... Second is bringing in sme's for an emerging technolgoy and going through this with them to see if it works and is relevant.
... Any q or comments?

Katie: Is there some other area where we discuss that it's going to meet the needs of that technology specific area? Are we also taking about the user needs?

Lauriat: That would be part of the review with the sme's.
... If we go to subject matter experts for VR, and they say "this has nothing to do with our tech", then that's an indicator that the model is not sufficient.

Katie: We need different types of experts, technology purpose AND accessibility.

Lauriat: Should we break that into a 3rd thing?

Charles: We need to define "expertise".

Katie: Then once it works with the technology, then for C it's about the users using that new technology and the success is if they can.

Lauriat: Because certain emerging technologies will have a certain space that's built up, prior to working with that technology, how do we phrase that?

Katie: We want the sme's to have expertise in the technology, and then we need AT sme's to be sure that what we want to have happen works.

Charles: Pre-touch, senses before you touch the device, you need someone who understands arthritus.

Katie: Don't have a better way to say.

Charles: There may be no AT in the middle.

Andrew: Offered an alternative in irc <reads>

Katie: I want it clearly spread out.

Alastair: creation authoring vs usage.

Judy: Trying to grow them deliberately.

LW: be cognisant of using american lingo.

Lauriat: Making note on term so that we don't lose it, but don't have to define it now.
... For flexible maintenance req. first test is a test run, b and c are meeting with experts in the given technology and seeing if it meets with their satisfaction
... and c is getting experts on users and accessibility and if it meets their needs.
... B is creation of as well as platform.
... C is users.

Andrew: Your condition is on B and not C.

Lauriat: I'm inclined to remove the condition, we'll check with experts.
... Anything else we can use to test a conformance model against this requirement?
... Creating a maintenance and extensibility model.

Andrew: How are we testing maintenance? This tests extensibility.

Lauriat: Pick 2 emerging technolgies, take them through the cycle...
... Apply an emerging technolgy to all the guidance.
... If you are authoring content for that emerging technolgy, makes sense to see how you are doing.
... The guidance for that emerging technolgy may be hard to follow, we should consider that a failure.

Katie: Difference between theory and practice.

Charles: What does "through the entire cycle" mean?

Lauriat: Through the entire process until... any methods we need to create, tests, anything else we need in support.

Charles: We can't test conformance model against a test on a method that wasn't written. It CAN be done on the framework.

Lauriat: And we run through it showing HOW it can be done, not just that we believe it can be done.
... IF we find that it's difficult to write up guidance for ar and vr, and it's hugely more complex, that's a red flag.
... If it's too easy that raises other questions.

Andrew: How many sc should we run through this process for 2 emerging technologies. Presumably not all.

Lauriat: Good question. We can't wait for migration to finish. Can't be all. But we need SOME represantive sample.

Charles: Do we need to quantify it?

Andrew: I hope it's not all, but somewhere between 1 and all.

Janina: Finding the right set is the key.

Lauriat: Choosing things we think are easy, and also chosing things we think are hard.

Katie: What about the 4 that are part of the ... keyboard trap... non-interference ones.

Lauriat: ...ideally mid range...

Andrew: Getting back to prior question. What is "maintenance"? I think "we need to change things later on".

Lauriat: Specifically for adding things. WCAG adding guidance for touch.

Andrew: Flexible extensibility?

Lauriat: Maintenance is a part of that. There's implications...

Charles: Or regroup or adding edits. Or deprecating.

Lauriat: For emerging technologies. I think that most things involved in maintenance and extesibility... assistant voice for example.. how do we add guidance around that?
... Keyboard focus trap, you can have an anolog to that in a voice interface, don't get stuck. How does that actually work when we write the guidance.

LW: We want to lift it up to "don't get trapped".

Lauriat: That's the process we are going through.
... Some things don't apply to voice interface.

Charles: Pefect example from "no keyboard trap" to "no trap".

Lauriat: Language of page is another example, to language of environment. You could apply to AR and VR.
... I have 2 bullets <read>
... don't just include the easy things.

Katie: Non-interference SC's.

Andrew: Might be worth while calling out something that says part of this review is considering the editing or removal of existing sc's, making sure that's possible.

Alastiar: conformance model in terms of coming up with a final score for your site/page... minimum number to make that work? Possible answer is "we talk about that later".

Alastair: A full set of methods vs a handful of methods, will that be enough?

Lauriat: Good question. That's two parts to "we aren't quite there". First we aren't there as far as how scoring works.
... Other side of "we don't know" is more specific to cases of emerging technologies and some on the web, what the platform offers for things that you can meet.
... for example language of environment, applies to or should apply to VR. Because you want the AT to read the signs through the virtual hotel.

Katie: And you want sub-captions to be correct.

Lauriat: Say the vr platform has no ability for you to declare the language. We want to work out how to expose gaps between tech and AT.
... Maybe the environement supports setting the language, but maybe gets ignored.

Alastair: cover or stick a pin, if you have a smaller set of methods you can use for a technology, does adding one unbalance the conformance model?

Lauriat: Yes, and we want to look into that.
... Two things then, one is for emerging technolgoy, and also for emerging interactions and user needs.

Alastair: How easy is it to add something.

Charles: May be covered in a different requirement. We are looking at "can it be done".

Andrew: May be "regulatory environment".

Lauriat: Will probably overalp with several.
... One thing in our favor is that for wcag if you add it and fail it...

John: Q on terms. Languages haptic feedback or other forms of comunication besides written/spoken communication. Are we broadly defining or keeping it narrow.

Lauriat: Table that until we migrate THAT specific guidelie.

LW: Chinese, french, etc. Morse code could be counted.
... Some identifiable thing.

John: That exists today.

LW: Not necessarily, a light pattern.

Lauriat: We can hold off adding Klingon.
... Any other aspect not covered?

Jake: One q. We talked about how many sc to include. I'm not sure about the mapping. Half is not 25, we also have AAA.
... AAA are the ones that sometimes may help but may be more difficult to implement for other people. Should we focus on AAA because they are the hard ones?

Lauriat: We are including ALL sc's, including AAA.
... Captions for pre recorded, live content for example. live content is AAA.

Jake: We talk about some other ones.
... Want to make sure coga and low vision are included. Should that be more the focus?

Lauriat: We have noted for what to include... difficulty.. does that cover your concerns?

Jake: We can learn from the ones that didn't make it or ended up in AAA.
... We might learn more giving them more focus.

Lauriat: We have that called out for the previous requirement.
... Is there something that didn't make it in that we can use?

Jake: Multiple ways is more a theory. AAA may be more practical to learn from.
... Just thinking we may learn more from those.

Katie: Tech is beginning to support AAA. Office with reading tools.
... I get what you say. This is for testing the conformance model and seeing if it's relevant.
... Maybe good to add at least one AAA.

Lauriat: I've added it to the ones we choose.

Katie: Which means includes AAA.

Jake: Need to get use to whole plane language.

Lauriat: Anything else that we don't have covered with our bullets?

Katie: I'll add something at 3AM.

Lauriat: Next requirement

<Lauriat> Next requirement: 3.4 Technology Neutral

<Lauriat> Guidance should be expressed in generic terms so that they may apply to more than one platform or technology. The intent of technology-neutral wording is to provide the opportunity to apply the core guidelines to current and emerging technology, even if specific technical advice doesn't yet exist.

Lauriat: At least some of maintenance aspect we just talked about would help with assessing technology neutral.
... We have two tests in here.

<Lauriat> Test a. Take 4 existing success criteria and 2 new guidance proposals from COGA or Low VIsion Task Force through the content development process including Guideline and supporting Methods and tests.

Lauriat: Sounds like what we just talked about.

<Lauriat> Test b. Demonstrate our work product with AGWG, Low Vision, and Cognitive Task Forces. If they give approval of our work to date around technology neutral, that will be considered a success.

Charles: That's also in the next one. That may be the number we are looking at.
... ...for existing... for and to is quantifiable benchmark for each of these requirements.

Lauriat: We may want to discuss in context of each of these. for maintenance and flexibility, I want to do more than 6.
... have a larger body to work with to see the effect of adding more. but taking something and seeing what it looks like in the conformance model, that's a good start.

Janina: Does it include a conversational interface tech?

Lauriat: One of our goals is that we don't have to know. It CAN apply to different things, but there's some guidance that doesn't apply to a specific tech.

John: Why wouldn't we use the term "communcation" instead of language.

Lauriat: Language is not in this requirement.

Charles: Language of page is one we used.

John: Future proofing... communication is better word concept.

Lauriat: Something good to talk about when we migrate that one.

Charles: Terminology overall.

Lauriat: We can remove the clause...
... Demonstrate with stake holders ... those and we are the folks that have to maintain this going forward. So bring it around to different members of those groups.

Jake: Did you consider adding the mobile task force? We had a lot of possible sc which didn't make it.

Lauriat: Nothing against adding mobile.

<AWK> Scribe: AWK

<Chuck> awk: What ways to we antiicpate... current conformance model doesn't touch on technology neutrality. will that be an issue for silver?

AWK: Does technology neutral impact the conformance model?

SL: We need to assess how well the model does or does not impact meeting this.

<Chuck> AWK: Fine to assess, I hope easier piece to assess.

Jake: we now also have biometric proposals may be interesting to add

SL: changed to "4 existing success criteria and 3+ new guidance proposals"
... Have two existing tests for the Tech neutral req - anything else?
... OK, moving on to Readability/Usability

<Lauriat> Next Requirement: 3.5 Readability/Usability

<Lauriat> The core guidelines are understandable by a non-technical audience. Text and presentation are usable and understandable through the use of plain language, structure, and design.

SL: JSp added a note that this is more about the content

Charles: this says "core guidelines" - how are we evaluating the conformance model against the core guidelines

SL: mostly about not painting ourselves into a corner
... first is take existing guidance proposals and try it out
... second is to check with COGA

JR: How to assess if an audience understands?

Charles: The style guide is the measurement proxy for understanding. The Style guide needs to be developed to ensure that it represents understandable practice.

<Joshue108_> I like the comment 'plain language is a proxy for understanding'.

<Chuck> AWK: Not sure tracking your q Charles.

<Chuck> AWK: We've tested that things that follow the style guide are more understandable?

<KimD> Does 4. 3.5 address whether something is easily translatable?

<KimD> And should it?

SL: so one test for success is related to the style guide and the second is checking with the plain language experts

JR: It will be impractical to assess actual understandability with people
... we may need some measure (e.g. some education level rating) to use as a proxy

MU: Need to make sure that it is understandable to non-english speakers too

<KimD> +1 to Makoto

JB: One challenge we have had for years is that the education level for content readability varies between countries. Do we know how to deal with that.
... per Makoto, agree that this is a challenge and I am working with Shawn Henry to try to identify best practices to help WG's with that.

Jake: Dutch is close to english but there were lots of problems

SL: Interested in the work with Shawn
... Agree with Makoto

<KimD> +1 to including non-native English speakers language in the document

JR: quickly respond - carefully used "lower secondary" as an example
... a danger of simplifying text is losing content but the simplification also makes it easier for non-english speakers

Jake: Readability and Usability - why the choice for blending those two?
... is the hope that better readability leads to improved usability?

SL: don't recall exactly how we got there

Charles: originally meant information design, not presentation design.

SL: leaving req text as is, but we can revisit if needed

KHS: Perhaps we should also ask usability experts in addition to plain language experts

<Chuck> Chuck IS the audio for the room.

SL: When we test content we will

<Lauriat> Next Requirement: 3.6. Regulatory Environment

SL: Onto "Regulatory Environment "

<Lauriat> The Guidelines provide broad support, including a. Structure, methodology, and content that facilitates adoption into law, regulation, or policy, and b. clear intent and transparency as to purpose and goals, to assist when there are questions or controversy.

SL: two tests

<Lauriat> 1. Demonstrate the point system proposal to a group of at least 5, but not more than 10, Regulatory stakeholders including at least 3 countries and 2 accessibility legal experts.

<Lauriat> 2. Measure success by general approval, with no objections. Not all suggestions for improvement need to be implemented, but there cannot be any “I can’t live with that” objections.

JB: one of the things that has been notable with WCAG 1.0 and 2.0 uptake
... that there are very different systems around the world
... think that we should test with >3 countries

SL: suggestion?

JB: suggest "jurisdictions" possibly
... the key is using different @@something@@
... three looks low to me
... say "governments or jurisdictions"

SL: replace "countries" with "regulatory systems"?

JB: sure

JSp: thinking about countries with different economic situations
... three came from "who do we know to approach?" to some degree

<Zakim> Judy, you wanted to talk about LMICs and to also comment about legal experts

JB: we are starting to look at use of our specs in low and middle income countries - so that is good to do
... understand that silver has been talking with legal experts and am a little concerned that they may be from the US and EU and need more diversity

<Chuck> AWK: I agree that 3 sounds low, but also its not the last time people will have the chance to look at this (we'll have public working drafts).

<Chuck> AWK: If a bunch of comments come in that the model is not sustainable, that will make it more challenging. I would feel better if the number is 5.

SL: Where do we have contacts?

JSa: US, Australia, Japan, China, Canada - that's five

<Zakim> bruce_bailey, you wanted to say that "Demonstrate the point system proposal" is not a high bar

BB: "demonstrate" is easier than "buy-in"

SL: we get scattered results, so will need to review contents

<bruce_bailey> Also, we should not expect "buy-in" (formally, from regulatory bodies) until TR

JSa: This is a fairly important one to get right

SL: At least 5 then?

JOC: Is "regulatory system" an american phrase?

SL: It isn't really a familiar phrase to me, so I don't think so

<bruce_bailey> i agree that "jurisdiction" is not a great term to use

JB: terms used are not consistent between countries so we will need to identify more in the glossary

SL: should we say that whenever we go to stakeholders we need to strive for the largest most divergent set that we can?

JB: wrt how w3c looks at conformance models it needs to be verifiable whether it has been done
... so "largest possible set" isn't verifiable
... for ATAG we defined 6 or so authoring tools and needed specific proof to show that it was met

SL: We need to write a note to define the tests before so do them, so we have that clarity for assessing completion

JB: in some governments they make the case that they don't have the resources to implement so we need to take differences between govs into account

MC: when we get to the right point we should precheck the plan with Ralph Swick

SL: We will share with AGWG also, need to get agreement in this room and in other rooms to make sure we don't miss things

MC: I NEED COFFEE

SL: ok, take a break, back in 15

<Ryladog_> <scribe>

<alastairc> scribe: Ryladog_

Shaun: anything more on this Demonstrate the point system proposal to a group of at least 5, but not more than 10, Regulatory stakeholders including at least 5 different >
... I want to reword this for bth sides of regulations. Should we add more numbers?

CharlesH: Judy suggested remove lawyers

AWK: I think it is fair to remove them as well

<KimD> +1 to keeping legal experts

Jeanne: Some of our contributing users were specifically attorneys

<bruce_bailey> legal experts not the same as people who need to follow the regulation

I agree with Jeanne

<KimD> I agree with Bruce - this is still not clear

Shaun: Do we need to define the stakeholders?

<KimD> I don't know what "each aspect" means

Shaun: trying to avoid to complications

AWK: I woukd writethe same thing twicefr each audenice to make it clearer

Jake: might want to get rid of column

Jeanne: remove regulatory systems from one line

SL: I would vote for including it

<JakeAbma> https://docs.google.com/document/d/13A8mGMnQujfEVqcw_LmAUYT8DDq_qW0TNcHxmCHd0io/edit#

SL: I suggest we remove bullet 3

AWK: I think it is good to respond to loud objections
... I think we should be measuring for the best possible consensus

MC: We want to challenge the reasons there ifthey are obscure

CharlesH: Change to no one strongly objected

SL: SL: I want to change it all to list what stakeholders were spoken to and reviewed.
... We have removed approved

<Lauriat> 3.7 Motivation

SL: We aregoing to finish the Conformance Requirmenet

<Lauriat> The Guidelines motivate organizations to go beyond minimal accessibility requirements by providing a scoring system that rewards organizations which demonstrate a greater effort to improve accessibility.

SL: For this motivation iteslf
... Find a broad set of implementors, how well does this support that

Jake: I have a hard time with undrstanding how do we reward them by doing more
... It needs a value - the scoring system creates a reward?

SL: This allows conpanies to say we allow companies to say we have done more and this is how we are identifying

MC: maybe you are less likely to get sued. Andmore customers may give you more money

Josh: Good question. I am wondering about motivation. Is it aspirational? Another thing, I see how hard it is for small companies or low capacity organisations to conform.

<AWK> 3.7

Josh: How is silver going to capture that an organization attempt to do their best?

LS: this is arequirement for silver that encourages businesses go above and beyond
... We want to provide a method for orgs to go ahead and do that
... How do we get the testing to be more achivable, I want to talk to ACT about this

Alistair: Do we want to try to define what levels orgs should strive for, or do we waant to leave that up to laws and regulators?
... How likely is a PWD to encounter a barrier, we come out with a percentage score. Such as a button with keyboard blocks to get to it.

SL: We want to be able to say you meet the minimum reqs BUT the users has a really difficult time to do it. We want to encourage via motivation to make the usabilty good

<Joshue108> KHS: Section 508 calls for that also - that we meet usability reqs.

<KimD> Functional Performance Requirement

<bruce_bailey> The 508 requirement for useabilty is at a pretty high level

<Joshue108> And the Air Carrier Access Act

<Joshue108> https://en.wikipedia.org/wiki/Air_Carrier_Access_Act

ChalresH: I think we want reward, to outcome over effort

SL: I will add a note about that

Judy: Am I audible?
... LEADS model, I like that appoach. I had he opposite reaction to Jake. Is 3.7 stated correctly
... I think it is best of motivation to outcomes

<bruce_bailey> From 508 E203.1: PWD "...have access to and use of information and data that is comparable..."

<Joshue108> +1 to Judy, I'm not really liking the motivation thing but really used it to comment on issues with even minimal compliance.

<AWK> q

<Joshue108> I question if that is needed at all.

<Ryladog> CharlesH

<Ryladog> Jake: is there some kind of a growth idea

<Ryladog> SL: it depends on how the conformance models work out

<Ryladog> Jake: I have heard many times we do not need to shoot for Gold

<Zakim> janina, you wanted to ask what we mean by "organization?"

<Ryladog> Janina: the word organization is bothering me somewhat. How does that limit?

<Ryladog> SL: it may be a very large org, or and individual

<bruce_bailey> Here is the better cite for usability in 508, from E101.1 Purpose:

<Ryladog> Janina: Si there are orgs that some parts make the effort and others dont

<bruce_bailey> ...to ensure accessibility and usability by individuals with disabilities

<KimD> +1 to the intent of 3.7 - and - to working on the wording (later)

<Ryladog> JohnR: Different issue. Is this self report? Is this like validating its VPAT?

<Ryladog> SL: No. Through the systems, so far many of the things are self reporting.

<Ryladog> Leonie: Reward is just what we had in gaming for a very long time. Being the best that you can me. The reward comes in being better

<Ryladog> Leonie: I am concerned about setting different req is relevant to various types of orgs

<Ryladog> SL: My company has more orgs than I know exist

<alastairc> Would the best outcome be that an organization would want to do better compared to *themselves* using WCAG 2.x?

<Ryladog> Jeff: I was reading the goal, having a req of rewarding orgs, Do not make it a req. Maybe that is a consequence.

<Ryladog> Jeff: that comes back to more of outcomes than a requirement

<Ryladog> SL: Silver motivation orgs to go beyond. By providing a scoring system, so we will revisit that second part, it may be too much in the weeds

<Ryladog> AWK: Motivation is a problematic thing/word. We cant say what is going to motivate them. If we talk about more specific levels

<Ryladog> SL: Should we change this requirements?

<Ryladog> SL: I am getting from folks we need some kind of need to do more?

<Ryladog> AWK: Yes I think we can. My vote for not using the word motivation b/c it is untestable

<Ryladog> Judy: I agree with AWK, you want to way to be able to differentaite the level.

<Ryladog> Judy: Reflect what has been accomplished with a level

<Ryladog> Josh: Making a rod for your own back.. Plus + to Jeff and Judy. Reward causing competition. Motivations pertain to the org.

<Ryladog> Josh: I would suggest removing that I'm not sure you can gamify reward here.

<Zakim> alastairc, you wanted to say wouldn't the best outcome be that an organization would want to do better compared to *themselves* using WCAG 2.x?

<Ryladog> SL: I would disagree

<Ryladog> Alastair: Apart from wordsmithing. How do we actually answer this questions. Do organizations wish to meet outcomes based on WCAG 2?

<Ryladog> SL: I get questions about this all the time. I want to be able to hand them a paper and say, go do this....

<Ryladog> SL: Having some kind of method to say this is how we are testing

<Ryladog> Chuck: we are struggling how to say this

<Ryladog> SL: Alastair had brought up a meta goal

<Ryladog> Alastair: Apart from Josh wanting to remove that, should we drop that?

<alastairc> scribe:alastairc

AC: Can we move on?

SL: We do need to work on the interviews, but yes can move on.

Judy: Understand it is about going beyond minimum; concept of minimum is tricky, especially "bare min". NL tried an Orange model (min) and saw how that worked, and the businesses didn't go beyond that.
... Need to be careful about using that term.

SL: Have been using WCAG 2.x at AA as the 'minimum'.

<Joshue108> Minimum and Maximums are not inherently bad, they are relative terms relative to something. It is the thing they are relative to, that is important.

Rachael: Understand a lot of it comes down to equivalence, can we do usability testing to check a conformance model?

<AWK> +1 to Rachael as something to seriously consider in the conformance model

AC: Wasn't sure if it was usability testing to test the models, or to include in the models.

Jake: Wasn't sure if it was motivating people or organisations? Where is the focus, where I come from the motivation comes from the working floor (people), not the organisation.
... We motivate people with certificates & blog posts etc. That's for people, but different for organasation.

SL: Depends on organisation, but it's a good point. It is intended to mean those who need motivation.

Jeff: If you have a scoring system, higher is better. Those that care will be motivated, perhaps we are spending to much time wordsmithing?
... But, what does it mean to be bronze/silver/gold compared to A/AA/AAA?

SL: Found confusion with the As, some people thought an 'A' was good.

Jeff: If Bronze is acceptable, is silver what is expected of certain organisations. Is there a requirement that there should be a level?

MichaelC: Understood the original intent of A/AA/AAA

<Zakim> AWK, you wanted to ask if we have investigated reasons for why people don't go beyond WCAG 2.x?

AWK: People asked Shawn to do more, Jake found people not going beyond. Have we investigated this?

<Judy> [JB: Notes that there were specific meanings associated with the levels in WCAG 1.0; not so much so with WCAG 2.0.]

Charles: Have some of that in the conformance research.

AWK: Some people report the level, and there is no partial reporting, that could be a dis-incentive to go further.

SL: Having something structured makes it easier to go above.
... Have a case of that internally, but hard to scale up without it being part of the guidelines.

Katie: Good to show that improved accessibility is better than WCAG 2.x AA. Want to make sure the lowest level is not rubbish.
... like the idea of motivating organisations, it's about people caring about others. People make differences every day, good to have a formal process to test for that.

<Zakim> Judy, you wanted to speak to the publicly assumed "meaning" of the name of a level, rather than the order of ranking.

Judy: Might be aware of this, but: With WCAG 1.0 we had roughly defined meanings. With 2.0 that wasn't the case, there were more factors for what went into each.
... for WAI staff, we haven't had a good answer to that. Jeff's comments that there was some meaning for the levels, and I hope you will have some sort of meaning behind it.

SL: For assessing a conformance model, we have stakeholder interviews of people who would be going through process. How about adding disability advocate organisations to ask how well it meets their user's requirements?

<KimD> +1 to adding "Stakeholder interviews across a range of disability advocacy organizations."

Judy: People making the case internally, it helps if 'this level gives you this'.

SL: Let's take that to advocacy organisations and ask if it holds water?

Judy: Disability advocacy orgs track things quite differently. The kind of informed answer may not be realisticaly provided by some orgs unless you do a well-structured info sharing.

SL: Should we consider that as a test for this?

Judy: Debatable, as a community, we don't do enough to educate orgs to provide the answer you want. Have lots of contacts, there's trust in what WAI does, but that isn't enough.
... We would want an answer about the impact on people, but not sure you'd get that.

<CharlesHall> note: the original Conformance Survey did not ask about any incentive or disincentive to achieve specific levels. report 03/2018: https://docs.google.com/document/d/1iOut3_i1JBQu5_16plZ8u7xCd66T4TbW2zPR1KmWWF4/edit?usp=sharing

Judy: Would be easy to ask questions that would not be easy for them to answer.
... needs advocates who know this community well, it's a complex situation that needs more discussion (sometime).

<bruce_bailey> https://www.w3.org/WAI/GL/task-forces/silver/wiki/2019_September_16-20_in_Fukuoka,_Japan#Agenda

Migration of existing WCAG content to Silver: functional user needs

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

<Lauriat> Working doc where we left off: https://docs.google.com/document/d/1aCRXrtmnSSTso-6S_IO9GQ3AKTB4FYt9k92eT_1PWX4/edit#heading=h.wybvadozdbds

<Lauriat> SC 1.2.1 Audio-only and Video-only (Prerecorded), SC 1.2.8 Media Alternative (Prerecorded), SC 1.2.9 Audio-only (Live)

<JohnRochford> test

<JohnRochford> hello world

<JohnRochford> Shawn: In functional needs, we are not listing disabilities.

<JohnRochford> Shawn: Users with limited vision or no vision.

<JohnRochford> Shawn: We are trying to outline the needs of a user.

<Lauriat> Style Notes on Functional Needs: https://docs.google.com/document/d/1aCRXrtmnSSTso-6S_IO9GQ3AKTB4FYt9k92eT_1PWX4/edit#heading=h.fd17rscqblnm

<JohnRochford> Shawn: URL is for style notes of functional needs.

<Lauriat> Functional needs should be expressed in terms of the function that can or cannot be performed. Avoid including the solution to the problem in the Functional Need statement. Avoid using the name of a disability, except as an example. See the patterns, below next.

<JohnRochford> Leoni: Are there examples of functional needs by the type of disability?

<AWK> Speaker masquerading as Seattle Space Needle: https://adobe.ly/30i6A83 | Actual Space Needle: https://upload.wikimedia.org/wikipedia/commons/thumb/2/23/Space_Needle_2011-07-04.jpg/500px-Space_Needle_2011-07-04.jpg

<JohnRochford> Charles: The user story format is an area I was arguing against.

<JohnRochford> Charles: It is easiest just to describe the need and not the user who has it.

<JohnRochford> Alistair: I agree with Charles.

<JohnRochford> Janina: The approach by example needs, not by users, makes sense to me.

<JohnRochford> Shawn: Charles, you came up with a list of drafted user functional needs. Would you please lead us through that?

<JohnRochford> Charles: I copied the benefits bullets from the understanding docs.

<Lauriat> Usage with limited vision or without vision

<JohnRochford> I then focused on needs.

<Lauriat> Usage with limited cognition or understanding

<Lauriat> Usage with limited hearing or without hearing

<Lauriat> People who need the ability to search for non-text content and to repurpose content in a variety of ways.

<Lauriat> Usage without vision and without hearing.

<Lauriat> Usage with limited vision or without vision and with limited hearing or without hearing.

<Lauriat> People who need a real time text alternative to live audio

<Lauriat> Usage without the ability to listen due to: any human factor (like ear pain); or any technical factor (like poor signal); or contextual factor (due to quiet environment)

<JohnRochford> Shawn is copying and pasting the text from the doc as Charles is reading it aloud.

<Lauriat> My question, did it mean this: Usage with (limited vision or without vision) and with (limited hearing or without hearing).

<JohnRochford> Shawn and Charles discussing "or" and "and" useages.

<JohnRochford> Shawn: How many instances of those can we cover because they seem infinite.

<JohnRochford> Janina: Depends upon primary disability. If your primary disabilities is hearing, you learned to finger spell, for example.

<JohnRochford> Charles: If we write the user need as any intersection of the user needs, how do we account for all of them.

<JohnRochford> ?

<JohnRochford> Charles: Combinations of user needs are unique user needs.

<JohnRochford> Charles: A physical disability and a cognitive one is one need rather than individual needs.

<JohnRochford> Katie: There are general categories that should work for all user needsd.

<JohnRochford> needs

<JohnRochford> Shawn: Example: Text converted to sign language.

<Lauriat> Assistive technology can read text alternatives aloud, present them visually, or convert them to braille.

<Lauriat> Live text caption service (by a human or by automation with a human-in-the-loop)

<Lauriat> Insertion of notes into live text caption on any non spoken audio which is essential to understanding

<JohnRochford> Shawn: For intersections of user needs, what are the methods of addressing them?

<JohnRochford> Shawn: We know the list of user needs is not comprehensive.

<JohnRochford> Charles: We need to describe more user needs than we ever have.

<JohnRochford> Andrew: User needs in the list may be redundant.

<JohnRochford> Jake: Would we like to build a different approach based upon user needs and the already-existant SC, or based upon user needs without new SC?

<JohnRochford> Jake: We might end up with the same guidelines for different SC.

<JohnRochford> Shawn: As we go through this process, we will likely come up with new groupings.

<JohnRochford> Jake: We may be limited by what we already know.

<Zakim> AWK, you wanted to go back on what I said, somewhat

<JohnRochford> Andrew: I like the generalization of user needs.

<JohnRochford> Charles: Not just generalization, but situational and contextual.

<JohnRochford> Shawn: We are trying to keep it at a high level.

<JohnRochford> Chuck: There is current research in preception science to expand into new SC.

<Zakim> alastairc, you wanted to ask if it is trying to be a complete listing, or a categorisation mechanism?

<JohnRochford> Chuck: (for example)

<JohnRochford> User needs not dependent upon human ability, such as no hearing or limited hearing, but perhaps are dependent upon situation or context.

<JohnRochford> Alistair: Wouldn't such examples (situational and contextual) apply everywhere?

AC: Suggesting that the last bullets with the examples would go under the general funcational needs. E.g. contextual factors like poor signal could come under the 'limited hearing' functional need.

<JohnRochford> Shawn: Functional user needs will not be exclusive to user groups or to other criteria.

<JohnRochford> Katie: We do want to allow for comorbidity of disabiliites.

<JohnRochford> Alistair: Would we want to give higher scores to techniques that address multiple needs?

<JohnRochford> Katie: It's important we make room to address multiple needs.

<JohnRochford> Jeanne: When there is a different technique that addresses multiple needs, that's different from techniques for individual needs.

<JohnRochford> Note to self: That last notation was yucky.

<JohnRochford> Note to Jeanne: Please fix.

<JohnRochford> Charles: Let's make sure the needs are captured.

<JohnRochford> Shawn: Addressing that would take infinite time due to all the possible combinations.

<scribe> scribe: alastairc

Charles: General point - avoid a rabbit hole
... the two specific inter-sectional needs came from the benefits.

Katie: Can we include a list of intersectional needs?

SL: Adding

<Zakim> AWK, you wanted to ask if there is a master list of user needs

AWK: is this what we're creating in aggregate.

SL: Sort of.

AWK: Is there a current master list? MarkH may know? There's IMS.

Charles: Not final yet. We did a mapping exercise with the EN list, which is in a spreadsheet. Not complete, more need to be added.

SL: Yes, but know there are gaps.

AWK: The EN one is broader than IMS, not sure if we need to define that.

SL: Would prefer not to make the 14th list.

Katie: Is that master list, with user/tech/context, with a publicly identified one, there is a value add from us to include what accomodation folk see as the top combinations as a piece for us to include.

<janina> d/msg ZoeBijl

<janina> sz w3c

Katie: Wondering about animation, which is similar needs.

SL: Still counts as timed content.

Janina: Asking CSS to deal with animation.

AWK: Wondering about the ability to search item.

AC: Seems to be solution oriented.

SL: Could go in other places.

AWK: Same for the next one, seems like a sub-set of the other ones.

Charles: Any bullet starting with 'people' was added from collaborative editing. This one came from urgency, the key was real-time, so the need was having to have it real time.

SL: Two ways of achieving same thing?

Charles: Trying to differentiate text, from text NOW.

AWK: Some equivalency built into these, e.g. pre-recorded is getting the same content, different for live.

SL: Similar thing with audio desc, does urgency need it's own functional need.

Katie: Might fall under context, as far as user-groups/context go.
... notifications could go in there as well, or could be separate.

Charles: Could be something like a tornado warning, people don't check those afterwards.

SL: Urgency would be part of an intersection with other things. Could put that in the doc?

Charles: There is a need 'plus',

SL: That inherently affects the worth of the method you're choosing. E.g. transcript the day after a tornado warning not of use. Way of judging meeting the user-need, not necessarily being the user-need.

AWK: Would there be separate methods?

SL: Methods for captions might include: I've include captions which are 70% accurate. Another might be using a real-time captioning algorithm, better. Then there is using a well training and domain-aware interpreter.
... Better ones score more.
... Inclined to delete this one as it is context not user-need.
... added a note for adding 'urgency' with the various needs.

<Lauriat> SC 1.2.3 Audio Description or Media Alternative (Prerecorded), SC 1.2.5 Audio Description (Prerecorded), SC 1.2.7 Extended Audio Description (Prerecorded)

<Lauriat> https://docs.google.com/document/d/1aCRXrtmnSSTso-6S_IO9GQ3AKTB4FYt9k92eT_1PWX4/edit#heading=h.e0yd4dvaz8if

<Lauriat> Users who cannot see or have trouble seeing need extra narration to understand the content of movies and multimedia.

<Lauriat> Users who have trouble seeing may need extra narration to understand the content of movies and multimedia.

<Lauriat> Users who can see but need extra narration to understand the content of movies and multimedia.

Charles: Language needs to be made consistent.
... wondering about 'extra narration', seems like the solution.

Katie: They need to know what's going on.

SL: All three say that, shall we clean up first/

first?

<bruce_bailey> still awake

s/first\/first?

SL: Looking at the grouping, add in the standard needs, then merge in the variations.
... limited/no vision covers first two.

Bruce: Extra narration = audio desc but plain languagy.

Janina: Could be audio, or could be text sent to an interpreter. The second is cheaper.

AWK: Added the intersectional (no hearing/vision) even though both sides are not covered above. But you'd have a different method for that intersectional need.

SL: (answering Jake). Thinking about using tagging to cover this.

Jake: When we do the intersectional need, a lot of those needs require an understanding doc in themselves. Harder for others to understand.

SL: Don't need to provide this as a list, these are to help write the other documentation.

Charles: There would be some sort of glossary. As well as tagging, also thinking about the cognitive walkthrough, and that technique would be easier with this list of needs.
... Sometimes 'intersection needs' is used in a more political sense.

AWK: Combined needs.

Katie: yes!

<mhakkinen> Concurrent needs?

SL: Everyone happy with replacing the 3 prior bullets?

[no objection]

<Chuck> scribe: Chuck

<Lauriat> SC 1.3.1 Info and Relationships, SC 1.3.2 Meaningful Sequence, SC 1.3.5 Identify Input Purpose, SC 1.3.6 Identify Purpose, SC 3.3.2 Labels or Instructions

<Lauriat> https://docs.google.com/document/d/1aCRXrtmnSSTso-6S_IO9GQ3AKTB4FYt9k92eT_1PWX4/edit#heading=h.d3u3fvumcxvq

Lauriat: Just getting started, back into groupings, working up functional needs. Linked to the grouping in question, copied in IRC.
... 1.3.1 info and relationships, 1.3.2 meaningful sequence, 1.3.5 and 1.3.6, 3.3.2
... Some are broken out into different groupings. For this grouping - when we made this grouping.
... The symantic need for blind users and the cognitive needs and methods

<Lauriat> Programmatically convey any semantics visually conveyed

Jake: not a bulleted list.

Lauriat: Placeholder. The list of SC's is long and confusing.

<Lauriat> More granular as well, beyond HTML & ARIA. Example: users swapping out icon sets

<Lauriat> Programmatic order matches visual order

Lauriat: This is the conveying of symantic information visually.

Janina: Came from visualization task force.

Jake: Do you want to switch the first one?

<Lauriat> Programmatically declare the specific kind of data expected in a particular form field

Lauriat: We'll probably split them up and redo them. We just pulled into the placeholder.

<Lauriat> Labels or instructions are provided when content requires user input.

Lauriat: Specifically from input purposes, but another aspect of programatic declaration.
... Labels or instructions are in a couple of places. In particular around the symantic relationship and expression.

Charles: Do you want to start popping in the user needs?
... I've got the first one.
... <pasted in the ones from 1.3.1>

<Lauriat> This Success Criterion helps people with different disabilities by allowing user agents to adapt content according to the needs of individual users.

<Lauriat> Users who are blind (using a screen reader) benefit when information conveyed through color is also available in text (including text alternatives for images that use color to convey information).

<Lauriat> Users who are deaf-blind using braille (text) refreshable displays may be unable to access color-dependent information.

Charles: Working on 1.3.2.

Lauriat: I'll separate these so we know where they are coming from.
... I'll review 1.3.1 as you work on the rest.

Katie: This sounds like it's coming from use of color.

Lauriat: It does sound like that.

Janina: text alternatives?

Lauriat: The first one sounded like 1.3.1. Adapting content... The second and third sounds like use of color conveying information.

<Lauriat> 1.3.1 Understanding: https://www.w3.org/WAI/WCAG21/Understanding/info-and-relationships.html

Lauriat: Directly from the benefits of 1.3.1.

Katie: Interesting.

<bruce_bailey> Seems so much more like 1.4.1, as Katie says

Katie: Andrew is that a bug?

AWK: Color is part of the presentation.

Lauriat: Just a very specific use case.

AWK: References 1.3 beneith it.

Jake: I can see if you have a link which separates itself by color, but that's...

<bruce_bailey> 1.4.1 is color is not the only VISUAL means of conveying information

Lauriat: That's different from info and relationships.

Jake: Blue text in a paragraph, it's a link.

<bruce_bailey> so 1.3.1 being about alternative to color could be okay

Jake: instead of only an onclick event.

Lauriat: Trying to understand the info in the understanding docs.

Katie: Maybe a mistake?

Alastair: For example if you have a table with red and green, you can have images with alt text or some kind of text.

Jake: That's 1.4.1.

Lauriat: They both apply there. Even if you do color and pattern, that more aligns with 1.3.1, and you need to programatically describe the meaning.

Alastair: Text works for that.

Lauriat: Aria works as well.

AWK: Second example is same situation. It's confusing. Why 1.3.1 needs to be broken into other pieces.

Lauriat: And we added five others!
... Sensory characteristics is the next grouping.

<Lauriat> The next SC grouping: SC 1.3.3 Sensory Characteristics, SC 1.4.1 Use of Color

Lauriat: That's a separate bucket of stuff. We finished. User needs for info and relationships, are all symatic expressions of meaning and relationships.
... The bit with label and instructions is the programatic instruction, not the quality of label.

AWK: 3.3.2 is the existance of the label, not it's programatic determination.
... People would argue that the label and the field are 4.1.2 and... some of that's where we could get some clarity.

Lauriat: I'm inclined to flag in the comment so that we can move on and discuss the rest.

Jake: You sure? In Android we have the heading element. We want to be agnostic. It's a global convention to add heading on top.

Katie: Deque convention?

Jake: A recommendation.

Katie: Who recommends? Deque?

Jake: Yes.
... We are talking more size and color.

Katie: The point is this is kind of misleading.
... This doesn't seem to be the primary. It's a distractor here.

Lauriat: We'll leave it, make a comment, and move on.
... 1.3.2... <pasting>

<Lauriat> From 1.3.2: This Success Criterion may help people who rely on assistive technologies that read content aloud. The meaning evident in the sequencing of the information in the default presentation will be the same when the content is presented in spoken form.

Lauriat: I would say that there's other...

AWK: Are we running the same types of things?

Lauriat: We are trying to understand from the understanding docs.
... We don't want to make up a list of user needs from our personal understanding. We want to build off the existing work.
... Charles pasted it in 3 minutes ago.
... We have other things that can come from 1.3.2.
... Very specific to user needs around... with limited or no vision and using AT like screen readers.
... From 1.3.5

<Lauriat> 1. People with language and memory related disabilities or disabilities that affects executive function and decision-making benefit from the browser auto-filling personal information (such as name or address) when the autocomplete attribute is used to meet this Success Criteria, which means information does not need to be remembered by the user.

<Lauriat> 2. People with cerebral palsy, stroke, head injury, motor neuron disease or learning disability sometimes prefer images for communication. They can employ assistive technology which adds icons to input fields to communicate the purpose of the fields visually.

<Lauriat> 3. People with motor impairments also benefit from reducing the need for manual input when filling out forms.

Jake: I mentioned before, we use the auto suggest, I even call it hijacking. It's not just about identifying input purpose or identifying purpose. Really great side effect.
... get some stuff filled in for free. Wasn't the intent from the beginning (more personalization).
... Belongs to something else which we don't have right now.

AWK: This is a hepful benefit that these controls are identified. Just like another helpful benefit would be to swap out a label for an icon. All things that can happen as a result.

Jake: Autosuggest was not in play at the beginning.
... Most people think sc is about autocomplete.

Lauriat: We should come back to this one and note the different needs that are helped by this.

Katie: Automation is a great way to put it.
... That's personaization.

Lauriat: Primarily around a user need of understanding.

Janine: Do we want to id the type of IT?

Lauriat: Don't need to at this point. When we look at the different technologies.

Janina: Fair enough.

Lauriat: Last up...
... <reads>

Katie: Same automation thing. Automation for ease of use.

Lauriat: From 1.3.6.

<Lauriat> People who benefit have many different cognitive disabilities including:

<Lauriat> Memory: 1. Focus and attention 2. Language-related 3. Executive function and decision making.

<Lauriat> Meeting this Success Criterion helps users who need extra support or a familiar interface, including the need for:

<Lauriat> 1. Symbols and graphics with which users are familiar 2. Fewer features and less cognitive overload 3. Keyboard shortcuts

Katie: Might be useful to look at the matrix Coga has.

Lauriat: Coga user needs mapping.

Katie: A need is a simplified interface.
... Keyboard shortcuts.
... Fewer features and less cognative load.

Lauriat: Under identify purpose. How does it figure in?

AWK: If you know what everything on the page is by context, like having an AT attached with those things. A mail client that has a compose button, you can attach a keyboard shortcut to that.

Lauriat: AT provided shortcuts.

Katie: Not necessarily.

Jake: Like to assign your own button?

AWK: Assuming everyone did this on the web, and everytime someone had a home link, home link was identified programtically as a home link. If you want to go right to the home page...
... AT could be configured such that alt-h always takes me to that place.

Jake: Maybe should be assigning keyboard shortcuts.

Lauriat: It's customizing the user interactions or user interface.

Katie: Like having a user style sheet.

AWK: I think that what this is ... it's symatics on steroids.
... People in low vision task force when there's an H2 I want two dots in the margin so I can identify it... this is conceptually the same thing. Similar to 1.3.1.
... With undefined benefits.

Katie: A preferred rendering.

Lauriat: I wanted to understand that because the keyboard shortcuts here are custom shortcuts.

Jake: Assign your own.

AWK: May be a browser feature at some point.

Alastair: 3.3.2 doesn't make sense in this group because it's not about the meta data or structure.

AWK: I agree, it's about the content.

Lauriat: We included it because of the symatic associations. The robust level.

Katie: Is it visually associated?

Lauriat: Not this group.
... The only aspect is you have labels are they associated?

Alastiar: That's under 1.3.1

Lauriat: We have 3.3.2 in a different grouping. We'll remove, makes sense.

<AWK> Scribe: AWK

<JakeAbma> scribe: JakeAbma

Katie: why is 4.1.2 not part of this?

SL: It may be, we deal with it at the Robust section and see how it overlaps

<Lauriat> Comment I just left: Look into merging completely or in part with: SC 1.3.1 Info and Relationships, SC 1.3.2 Meaningful Sequence, SC 1.3.5 Identify Input Purpose, SC 1.3.6 Identify Purpose

AWK: some of the Brainstormed needs seems useful
... maybe add "bypass blocks"

<Ryladog> IMS https://www.imsglobal.org/accessibility/accpnpv2p0/spec/ISO_ACCPNPinfoModelv2p0.html

<Lauriat> trackbot, end meeting

- DRAFT -

Silver Community Group Teleconference

15 Sep 2019

Attendees

Contents

Migration of existing WCAG content to Silver: functional user needs

Summary of Action Items

Summary of Resolutions

Scribe.perl diagnostic output