Meeting minutes
Intros or Topics
alastairc: any newcomers or changes of affiliation?
WCAG 2.x issue batch https://lists.w3.org/Archives/Public/public-wcag2-issues/2024Dec/0005.html
alastairc: we have some WCAG 2 stuff to look at first
alastairc: *shares the board https://
<alastairc> w3c/
alastairc: this is on an update to Failure 2, it was highlighted the example wasn't really valid, this is to remove the example
alastairc: in the project board, we have a 'Sent for WG approval' column
alastairc: for these we have general support it seems, feel free to check them if you're curious
alastairc: in the For Discussion column… one of those is the “user inactivity” definition… we found it wasn't in alphabetical order and it was missing the 'note' class
alastairc: we had some support, but one thumbs down, no comments
alastairc: we also have one on target size, will come up in our discussion
alastairc: please do check the board if you can or are interested in WCAG 2 updates
WCAG 2 scoping discussion update https://github.com/w3c/wcag/discussions/4188#discussioncomment-11830612
<Wilco> The e-mail doesn't ask for comments, just a vote
alastairc: last week we discussed what things we felt comfortable to count as errate updates
<Laura_Carlson> I need to leave at the second hour.
alastairc: we had several examples of things we looked at last week
alastairc: the update since last week is whether and how we could make a differentiation for a new updated version of WCAG 2.2
alastairc: I posted a couple of option, including some kind of stronger date stamping (wouldn't affect URLs we'd just announce there's a new date stamped version with minor changes), the other was a minor dot release, eg WCAG 2.2.1, would be an indicator that it has been updated in a minor way
<mbgower> Wilco "The e-mail doesn't ask for comments, just a vote". If you're referring to the original proposed changes, it states:
<mbgower> Give a thumbs up on the pull request description to agree. We are looking for >4 +1 votes (activating the 👍 emoji) for substantive or errata changes
<mbgower> Or provide feedback in the Conversation to request changes
alastairc: another proposal, from Wilco, was to put it under a new URL… if I'mreading correctly, the title would be “WCAG 2.2 2025 revision”, would have a new URL, eg /TR/WCAG22-version-2025
alastairc: one thing nobody seemed to argue with… if we were making changes that update how people interpret conformance, that we should go through a public review process
alastairc: but that is a separate thing re URLs and names
alastairc: any Qs or comments?
GreggVan: if you're changing the conformance model, it really would have to get a new number, or you'd get endless confusion re what 2.2 is
<Glenda> +1 to what GreffVan is saying
GreggVan: eg if we publish something new with the same increment number that is not the same thing
<Zakim> alastairc, you wanted to comment on "refinement"
alastairc: not sure there's a difference in those?
alastairc: eg one of the cases I included was changing Timing Adjustable so that it is clear in that clause whether it is 10x original time or 10 opportunities to extend time. Some interpret this as 'you can do either'. I don't think that was the original intent
alastairc: in my mind that would be a refinement. But if people were interpreting as 'either', then clarifying = changing it
<Wilco> +1 what we meant isn't the requirement. What we said is
alastairc: (is the argument people are making)
GreggVan: what about conformance model?
alastairc: ah no it's about interpreting specific bits of criteria, not the conformance model
Glenda: if normative is changed, you can now pass a success criteria, because normative was changed to lower the requirement, and the version number did not change, then we have testing methodologies around the world where people didn't have patch that that happend, as they're not hanging on our every word
Glenda: eventhough W3C allows us to call it 'errata', any normative change must have a version number, only then it would be cristal clear something changed
<Zakim> kevin, you wanted to comment on process for patch versions or similar
kevin: if we go down the route of patch changes, it's effective going to trigger the whole process for public working drafts, incl AC review
<mbgower> This is covered under https://
GreggVan: I think we should make clear that we're not talking about changing conformance model, just about normative changes
GreggVan: if you're not changing what the intended meaning is but just the language to make it clearer, it's not normative to me
<dan_bjorge> -1. The working group's intent is basically irrelevant if the intent is in conflict with the language as written, if the language as written is legally required.
GreggVan: if it's not important I would not going to go through a whole process
GreggVan: not sure how to label merging in errata
alastairc: in the case of PR #4122 it doesn't seem that this is the type of language change we're talking about here
alastairc: so sunmarising, there's things that can count as errata, can republish without new URL or version number, and others cannot
alastairc: in other cases we'd need new URLs or version numbers
alastairc: in that case I'd say we'd consider a WCAG 2.3, which we'd highlight as a 'no new features version'
<Zakim> alastairc, you wanted to comment on one-way backwards compatibility
alastairc: like the 'snow leopard' version that Apple released for macOS some years ago
GreggVan: how important are these changes you're talking about
<Glenda> Would WCAG 2.3 have any substantive changes that would require a11y experts around the world to revise their manual testing methodologies?
GreggVan what are the changes? if it's just 2.2 but we've just updated some word, that doesn't seem sufficiently substantive
<Glenda> Is one of the substance changes “Align AAA criteria with newer criteria” For example, adjusting the inline-text exception of Target Size (Enhanced) with Target Size (min), e.g w3c/
<Zakim> alastairc, you wanted to comment on potential changes.
alastairc: could get a list of changes. At this time there are quite a few, but it seems none of the individual changes seem worth the process
<Zakim> mbgower, you wanted to say 1) the w3c process supports clarifications that may affect interpretation as an errata process 2) there is overall agreement that we want to improve the standard 3) the nub of the discussion appears to be around how we go about publishing those changes and the impact of any specific change (errata versus 'stamped' versions)
alastairc: seems more of a process to gather them up
mbgower: W3C process does support clariifcation on interpretation within the errata process
mbgower: have no problem going more tight than W3C process, but wanted to make that clear
mbgower: seems like there is overall agreement re improving the standard. More of question re how
mbgower: we probablhy don't want to fix one typo and call it WCAG 2.3, but maybe in the next while there'll be an number of issues that could could go into a 'WCAG 2.3 bucket'
<Glenda> +1 (want to improve the standard) but…I’m not cool with substance changes under errata (the types of changes that would require a11y experts to all adjust their manual testing method…or…require tool vendors to change the way an automated rule logic works). If we say we are testing to WCAG 2.2 (but there are 2 versions of WCAG 2.2)…which one do we really mean?
mbgower: can we figure out where the threshold is to go through this? glad we're having this convo.
<Glenda> I have no problem with typos being resolved as errata
sfaulkner: just wanted to ask why would we want to go further than the rules defined by the W3C errate process? I understand consensus, but if we find there are conflicts that could inform improving the W3C errata process?
sfaulkner: we shouldn't create extra work for ourselves or others?
dan_bjorge: I disagree with Steve'\s position. The rules that apply to errata make sense to most W3C recommendation, but what we're working on is, in practice, a legal standard, so to me, it makes complete sense for us to be more conservative
Glenda: was shocked to see that the W3C allows for substance changes that increase or decrease requirements… I sat down wiht Chaals and asked if this is how it was meant
Glenda: my understanding is that, just like Dan just said, this is more for other standards of W3C, and WCAG is a special case as it is in legislation worldwide and 1000s of humans around the word need to potentially re-learn what things mean
<Wilco> +1 we make it easier on ourselves by making it harder for everyone else
jon_avila: am not sure about the legal standard, but what I know is that versions of WCAG are incorporated into many regulations
jon_avila: they tend to be incorporated with specific version numbers, eg in Section 508 and European standards, have specific numbers
jon_avila: eg in ADA WCAG is more of a benchmark, in a court it would be more about if a person could perform a task and less about conforming to WCAG
jon_avila: so if we're talking about this as a legal standard, I understand the challenges. The group's opinions have changed over time, that's a problem as we have inconsistencies and changes over time