Meeting minutes
<JF> Previous minutes: https://
<JF> Previous minutes: https://
<ShawnT> Google document: Protocol Proposal: https://
JF: Good morning, happy new year.
JF: Time to dive in and bring this to a point where we have a proposal to take to larger group.
<JF> https://
JF: URL's that are important... <pasted>
JF: Minutes and google doc. To summarize...
JF: Our last call I got excited, inspired by Jaunita, mentioned decision tree. Characteristic of a protocol, provides step by step guidance.
JF: Useful model. One other thing to introduce was collection of AUR's out of WAI domain.
JF: We need to clearly define what we think a protocol looks like, our last conversation got us close.
JF: Any comments or recollections from past calls before we proceed?
Jaunita: We should also mention difference between protocol maintained by somebody else, and is there a portion we would maintain?
Jaunita: Defines whether or not the protocol was followed and sufficiently meets guideline?
JF: Important data point that we need to resolve. Protocols, do they live only at W3C? Do we accept 3rd party protocols? I'm on the fence.
JF: I thought allowing 3rd parties introducing would be useful addition to idea, but until we have a clear definition... "must contain these things". I don't know that 3rd party protocols would fit that model.
JF: I think we need to clearly define first, and reserve comment. Don't lose the idea, but have the discussion later after we define protocol.
Jaunita: Might make sense to have W3C approve.
JF: Yes, introduces another discussion. What is the vetting process? I think that to make extensible, we should allow for 3rd party, but they need to meet a structure.
JF: If 3rd party presents, W3C will need checklist.
JF: We are still at point of defining structure. We started talking about a guided tour. A walkthrough like decision tree.
JF: Does everybody agree that this is part of what a protocol would consist of? A decision tree process? Or too specific?
JF: Open conversation.
Jeanne: I think a decision tree process seems to exclude some protocols we wanted to include. I'm concerned about defining a protocol as a decision tree process.
Jeanne: It's fine to have it as a protocol, but limiting it to decision trees is too restrictive.
JF: Broadly speaking in my mind. Is the goal of a protocol to provide expert guidance to content creators on how to meet accessibility needs? Do we agree that is what a protocol will be?
Jaunita: Yes, along with some form or addendum you can measure against. If you do x, y, and z, you meet the requirement.
JF: I embrace the idea of some form of repeatability. Alt text decision tree has repeatability.
JF: Jeanne, Jarred and Mike might look at an image, and all 3 provide good alt text, but all very different.
JF: By virtue of experience they go through in their head. At end of the day that this is the process they used.
JF: Measurability piece is ... is it obvious that they went through the process?
Jaunita: Measurement is, is it sufficient?
JF: But that is arbitrary. Measuring sufficient may be problematic. Very opinion.
JF: If they follow the process in the protocol, they will land in the circle of "sufficient".
Jaunita: I'm thinking that someone follows protocol, but doesn't lead to a sufficient result.
Jaunita: I'd like some repeatability in WCAG 3, leveraging latest research. If you don't, and if you are relying on an arbitrator, and there isn't a line we can draw, it would be difficult to adopt and scale.
Jeanne: I think this is why we can't put protocols at bronze level. We can't require them. If they are required, then they have to be "guidelines". We are already doing guidelines. I'm excited about the fact that it gave us the things that do not have measurable testable repeatable outcomes, look at best practices, that do have some measurement, but don't have to be repeatable.
Jeanne: We can have measurements on inputs. We could say for content usable, you need to hire editors, have a style guide, editorial process. We can do those measurements in advance.
Jeanne: We can put it at a higher level that is not regulatory. Bronze can be regulatory, and companies can have silver and gold levels. Let's stop trying to make it a traditional guideline.
Jaunita: Earlier discussions is that silver would be equiv to WCAG 2.0. Now we are saying bronze. If bronze only includes stuff that can be automated...
Jeanne: To be clear that this is not what I'm proposing.
JF: I've been talking about a different scoring mechanism. TF and WG has not yet selected a scoring mechanism. All options are on the table. My proposal is your total score lines up with the level, bronze, silver or gold.
JF: Not even W3C can claim conformance to WCAG 2.1 A and AA.
JF: In my thought process for scoring, you will be rewarded for public commitment.
JF: Companies will be accountable.
<Zakim> jeanne, you wanted to say that measurements can be done on inputs not necessarily outputs and to say that Jaunita is pointing out the flaws in the Single Currency proposal
Jeanne: I think that Jaunita is pointing out the flaws in the single currency proposal. If we have a single currency, we will have to have measurability at every level.
Jeanne: We should collect the ideas and arguments and present to AG.
Jeanne: I think we need to review a multiple currency proposal. Jaunita has pointed out flaws of single proposal.
MC: The idea of bronze being equivalent, that's a rule of thumb not a rule, let's not get too concerned. Needs exploration.
JF: Shawn, do you have some thoughts and comments? I want you to have opportunity.
Shawn: New to silver, trying to catch up. I think this is great. Trying to get something from Canada put into this list. Canada has been kind of behind the times.
Shawn: I like the ideas. I'll stay on the wall.
<Zakim> MichaelC, you wanted to note public commitment needs to be backed by accountability of procedures and to reinforce input, not outcome
MC: Talking about public commitments, there needs to be a measurability to those commitments. Some information about the procedures to ensure that they are meeting the commitment.
MC: There's a balance to company internal decisions.
MC: When we talk about protocols, that we are all thinking the same way. We need to focus on inputs: What is it you do to achieve greater accessibility? That's where we should focus.
<JF> https://
JF: I agree with most of that. There is some way of having some accountability. I've pointed to another document that meets my concept is the plain writing req from us fed govt.
JF: Spoke with kirkwood, it's a 2020 report from justice dept. A review of their adoption of plain writing. I think that to that public accountability piece, here's an example.
JF: Don't know if that meets the model we need, but it is an example of something that fits my idea, as well as the accountability.
JF: Q on table, is that sufficient for our needs?
<Zakim> MichaelC, you wanted to discuss conformance vs accountability
MC: We need to be careful about public accountability. Only a factor when making a public conformance claim. Most companies will try and meet without making a public statement.
MC: They may be required to make a conformance claim.
MC: Let's not mix up what's the point of conformance with what's the point of being a good citizen.
Jaunita: If we do this for each of the outcomes, may not be feasible. May put a target on companies. Whenever they make a claim you have drive by lawsuits.
Jaunita: Public declarations of how they meet will open up floodgates to critics.
Jaunita: If there is no way to say "we met because we've done x, y, z", and people can come back and say the result doesn't meet the standard, that could lead to problems for companies, especially if it becomes legally enforcable.
Jaunita: More of a rep issue at that point. Depends on how we are using this.
JF: I think you hit the nail on the head. I've been looking at a different angle, but close to agreement.
JF: difference between protocols is that we are not measuring outcomes, we are measuring inputs. We are measuring the process that the content creator used to arrive at the outcome.
JF: What is the decision tree used to determine that they followed the process?
JF: In terms of target on their back, that's what I'm shooting for. The idea that there is accountability, but not on outcome, instead on the process.
JF: If org adopted plain language, having it public is we are now asking the entity to point to the protocol or process that they used, so anybody can review.
JF: In case of drive by's, if it gets in front of judge, statements of here's the documents that was used.
<Zakim> MichaelC, you wanted to very controversially refer to user need mapping and to say protocols != outcomes and to s/public/targeted/ and to say this is all just for claims
MC: I don't see protocols as supporting outcomes. We have different kinds of outcomes.
<JF> +1 to another kind of statement, but I advocate that those other kinds of statements get rewarded for adoption
MC: Can sit along side the statements. This ties into the work that Jake has been leading with functional needs, identifying user needs and functions. We get a matrix and we see what are the outcomes needed to meet the outcomes.
MC: I'm wondering if protocols could follow same model.
MC: I'm not saying this is the proposal that should go forward, but for me it's a useful tool for thinking about this.
MC: We keep getting tangled up on "publicness" of claim. We should stay focused on A) conformance claims are not required B) conformance claims need not be made public.
Jaunita: only 2% go to court, most are settled.
Jaunita: When a company makes a declaration, and there are a lot of lawsuits, most will be settled out of court.
Jaunita: You don't have to prove anything to bring a lawsuit, just have a plausible claim.
Jaunita: I would bring that point to say when we say we want companies to do this, we need to understand consequences, and see if it's reasonable to ask them to make a claim without thresholds to meet the claim.
JF: I want to challenge MC on conformance claims. Today they are not required and not needed to be public. I'm trying to change those 2 statements to be true.
JF: We introduce concept of progamatic conformance claims. Employ WIIFM
JF: There must be a value in doing a public claim. We would need a framework in place.
JF: It is the introduction of public accountability, and company gets rewarded. That needs to be discussed, and needs to go back to AG. In my mind the way protocols contributes is that there is a value in adopting.
JF: Otherwise they are just best practices. And it doesn't matter how many AAA requirements we have, they don't get adopted because no value in adopting them.
<Zakim> MichaelC, you wanted to ask about lawsuit requirements vs claim and to differentiate offering and requiring
MC: My thoughts are a combo of the others.
MC: I don't see how we can require a conformance claim. We can't force.
MC: If we require a public claim, I think some companies will decide the guidance is not for them.
MC: For legal, I don't know that law requires public claims.
MC: Depends on jurisdiction. I don't know if law requires. We should differentiate presumptive vs active conformance claims.
MC: I hear you that if we don't provide accountability, it will be analgous to AAA. I think we want to provide tools where we recommend.
MC: I don't think we can force.
Jaunita: Let's say it was cyber security. And we required public claims. And company says "our system is unhackable", and one flaw is introduced by vendor, but wasn't there at time of claims...
Jaunita: And someone exploits, anybody can bring a lawsuit. I agree with MC that we shouldn't make required. If it becomes replacement for what is legally binding, it would be difficult for that to become standard.
<Zakim> JF, you wanted to point to the EU "Cookies" requirements
JF: I hear what you are saying all, but I'm going to reference in EU... that legal requirement around cookies.
JF: 3-4 years ago we started getting "accept cookie" popups. We have examples where the law says you need to have a privacy statement.
JF: It was a local thing that became world-wide. Having a public statement around cookies and retention is a good thing. The mechanism was there.
JF: They aren't saying you must accept this cookie policy, but you must have A policy and that must be available. We don't need to encorporate in our model.
JF: Already a thing today.
Jaunita: That requires that they have a retention process and guarantee rites for individuals. It requires mechanisms and not outcomes. Different if outcomes are required.
Jaunita: Problem is that no matter what a business does, they can't guarantee an outcome if there is no line in the sand.
JF: Are we measuring outcomes or inputs? You focus on outcomes, I focus on inputs.
JF: If all structure and framework is in place, we have some confidence that outcome lands in the region.
<Zakim> MichaelC, you wanted to not object to accountability when policy requires conformance
MC: I'm hearing "how we measure best practices". We need to separate conditions.
MC: We are not requiring conformance, we can have accountability as best practices when conformance is not an option. We need separation of required vs optional.
JF: I'm prepared to admit I'm wrong. We are agreeing on some points, disagreeing on others.
MC: I think we are agreeing on more than not. Close.
JF: I'd like to finish up with... maybe "can we get a bulleted list" on the things we agree on. If we can finish with that list, that would be significant progress.
MC: And we need list of what we don't agree on.
JF: Can MC articulate what we agree on?
MC: I think we all agree that protocols measures inputs and not outcomes.
<JF> +1
+1
<Jaunita_George> +1
<ShawnT> +1
Decision: Protocols measures inputs and not outcomes.
<Jaunita_George> +1
MC: I think we agree that accountability mechanisms are necessary.
<JF> +1
+.5
<ShawnT> +1
Decision: accountability mechanisms are necessary.
<MichaelC> MC: For measurement, accountability mechanism of protocols corresponds to measurement mechanism of outcome statement
MC: for outcomes, there's a measurement for how you will get scored. There would be same thing for protocol, the actual measurement would be different, but the definition would be similar.
<JF> (measurement of protocols = adoption)
understood Jeanne.
<JF> +1
+.9
<Jaunita_George> +.5
<ShawnT> +.9 also
<JF> -1
Decision: For measurement, accountability mechanism of protocols corresponds to measurement mechanism of outcome statement.
MC: Accountability of protocols has same visibility as conformance statement, or is part of conformance claim.
Jaunita: Is claim public or private?
MC: I say we don't require either way.
<Jaunita_George> +1
MC: We would probably recommend public.
<JF> -1 to us not requiring
<ShawnT> +1 to chuck
Undecided: Accountability of protocols has same visibility as conformance statement, or is part of conformance claim.