Meeting minutes
<JF> GM< I can, but I am currently sitting on hold trying to get a vaccine appointment set up
<JF> I'm currently #4 in line...
<JF> so I will be 'late' to join this call
<Lionel_Wolberger> tells JF: good luck getting your "V"!
<LisaSeemanKest> more important!
Becki drafted a response to i18n issue #144 (no response back) https://github.com/w3c/personalization-semantics/issues/144
sharon: aPhilips has not responded yet.
… , this is similar 21 days ago now, we added a note
<JF> Close issue and move on?
janina: I think we are Ok at this point.
… , I think we can close it
LisaSeemanKest: do we need to add a new note in the issue?
janina: No we can just close it and they can open it if they want to reopen it.
Resolution: ok to close #144
sharon: JF & Becky to draft a response and add a note in it and close it
becky: No info yet, we did not meet.
sharon: ok we will wait then.
Zakim: next item
sharon: deal with purpose?
Matthew_Atkinson: I was thinking how these could be simplified. biggest question, could we combine action, destination, etc. but I don't want to delay us from our next release.
<Matthew_Atkinson> Demo page: http://
Matthew_Atkinson: , made an example page static page here to a browser extension if any pages actually implemented our spec. it implements our spec now. and have a set of tests. and there is alternative approaches where we get the things from the role not the attribute.
Becky: this is awesome, going from 3 to 1 is a big change for authors.
JF: I talked with mathew, ease of authoring, we chose action: destination:purpose mirror button, link and input, we can specify the role using aria, at that point why do we need to also specify this? only downside, tools supporting our spec must also support AIRA but this may not be a bad thing. Mathew thanks for building out a test suite. no one is really authoring right now.
<LisaSeemanKest> https://
Charles: agrees this is a good idea we should consider
LisaSeemanKest: I did a hackathon and saw where we saw mistakes, people don't really read the spec, similar names developers would do it potentially incorrect. Implementers making the extension if they haven't done it correct, both sides implementor and authors. AI gets it wrong too often. That problem may go away with time. we changed it the names are less ambiguous now. I don't think we can go down to 1 item.
… , other things which may require a more complicated areas where we experimented with more complicated with more value pairs. I think we need to reopen topics has something has change. something that might have changed names are less confusable, it won't get confused. if something invalid.
<Zakim> janina, you wanted to comment on relying on ARIA
janina: I love the idea of simplifier for the author, I am worried on depending on other specs.
… , we need to distinction clean, take up to APA maybe up to Judy. this is the first normative spec for accessibility since ARIA etc.
… , AWAG may disappearing. we need to consider interdependency so that when you do markup we don't just reply on ARIA for VI users, but this is really interesting and thanks Mathew, I leave this to Lisa and Roy etc.
<Zakim> JF, you wanted to note that all 3 are essentially "purpose" - purpose of link, purpose of button, purpose of form input.
JF: Mathew is not wrong purpose of the link, the form, the button etc. knowing if its a button, link, input, and we need to know the role, but again role is not specifically with ARIA, but also HTML5.
Becky: If people is going to be confused when to use destination or action, Do we thing data-purpose and data-destination on the same element?
… , looking at the number of things and may be confusing for authors on which to use.
Matthew_Atkinson: This is exactly what I was hoping for.
… , I am not wed to any approach.
… , just trying to understand thing better since I don't have this context. Wondering if more helpful look at the example page I created and come back to this after we have had a look at it. some of the things we discussed are examples. Do we want to pause this? and come back to it?
janina: JF please confirm HTML input. we can rely on HTML.
… , people implementing the spec will read it more than authors of web pages.
… , Thanks Matthew_Atkinson
Lionel_Wolberger: +1 to Matthew's work and insights. If we can prove we need all 3 attributes then Matthews approach may not work.
becky: someone using the anchor tag wrong destination if they don't understand. the extension or user agents. we may go back to the minutes 1+years what we may have forgotten.
<Matthew_Atkinson> +1 to going back over the minutes (would we know when that would be?)
LisaSeemanKest: issue was how user agents handle it if its a link vs. a field purpose, and lets say generating tool tips put your first name here for your ID you can show a picture. there are may have a help pannel explains what to do if we merge it we put the burden on the 3rd party tools they have to do more now based on the role, etc. might get it wrong, we could test, actions are on buttons etc. enables authors who be abl
e to test these strange ambiguities, and UA must guess
<JF> @Janina - aria roles related to forms: role="checkbox", role="form", role="radio", role="search", role="textbox"
JF: Lisa still has a argument here, but there are higher role=search role=form. there is some semantic bindings there. we want implied semantics or defer to the host language. When I do button, I am implying role=button since its native in HTML. Lisa's argument we are guaranty we wont leave anything behind.
LisaSeemanKest: we have a 2 code bases from the Easy Reader project.
… , we have a few test pages as well.
charles: need to consider if the author puts this on something that isn't a button, form or link what happens.
Matthew_Atkinson: More test pages and real world examples so thanks for that Lisa.
<becky> Looks like the comparison table had changes in august-October 2018 timeframe. Note that single attribute pros and cons was listed early in the discussion
Matthew_Atkinson: , look at the example page I created and love to get your input.
becky: I just looked and there were 83 updates to that page, single attribute was listed Aug/Sept of 2018 if we want to scan the minutes.
<JF> +1 let this percolate for a week, and revisit next week
sharon: talk about this on the mailing list then? talk more about this, we need to understand the history.
janina: I think waiting a week
Lionel_Wolberger: CR may mean we need to answer this.
JF: both Matthew_Atkinson and LisaSeemanKest have great reasons, so I am on the fence. If we can get down to 1 attribute. Lets go take a look at it and come back to it next week.
sharon: I will look back in the minutes on this topic.
review open issues - https://github.com/w3c/personalization-semantics/issues
<JF> https://
JF: I did make some changes I have. PR open.
sharon: I did merge that yesterday.
… , look at some of the open issues.
… , we covered the i18n's
… , two from Matthew_Atkinson on the explainer
… , look at the editorial issues #177
becky: sounds like these are just editorial.
Matthew_Atkinson: we did discuss this briefly, and split out the non editorial ones which I did.
… , #177 are content related I can split it up into several issues. or we can start going through.
LisaSeemanKest: do we need the explainer for CR?
Matthew_Atkinson: I don't think the issues I have would block us from CR. where did the list of actions/destinations come from? I didn't see a couple I would expect to see which was missing some of this was work from Coga.
janina: we do care about substantive changes. Explainer is not essential for CR but it will help with implementations explaining what it needs to do and be clear.
<LisaSeemanKest> we just went though common elements, - but it wasyears ago. some proposals got rejected
<JF> https://
JF: process Question the explainer right now says its a WD for over a year now. is this a Note or a Normative…
janina: its a note.
JF: we should finalize this into a Note, seems to me as we move to CR and get the Explainer as a Note,
janina: we should wait, but bring it up to the current spec but lets not finalize it yet. updating it is a good idea.
JF: work on Matthew_Atkinson changes or we defer them but we do something with the explainer.
janina: priority to make substantive changes first and parallel the editorial changes.
JF: lets work on editorial changes, I can do that and do a PullRequest
sharon: #177 and #178?
Matthew_Atkinson: The documents are great these are minor things so they are all editorial both issues.
Action: JF to look at any editorial changes proposed by Issue #177 and respond next week.
<trackbot> Created ACTION-80 - Look at any editorial changes proposed by issue #177 and respond next week. [on John Foliot - due 2021-04-26].
Lionel_Wolberger: do you want me to ping Addison?
Action: JF to look at any editorial changes proposed by Issue #178 and respond next week.
<trackbot> Created ACTION-81 - Look at any editorial changes proposed by issue #178 and respond next week. [on John Foliot - due 2021-04-26].
janina: they are busy its fine for them to reopen. you can ping him but it doesn't hurt.