<Helen> chair+
scribe+ dan_tripp
scribe+ Dan_Tripp
helen: done very little except planning for today.
giacomo: done a bunch of reviews.
today hoping to discuss process W3C regarding publishing
activity of rules, update of tools - but we'll need more people
here. opened an issue related to focus indicator visibility.
think it's an atomic rule. thinks it's too permissive. maybe
not today, but sooner or later - we need to discuss.
... aqa is now listed in implementors list.
sage: last week was on call (office hours?), just me and wilco, worked through PR stuff. just what we did on the call.
sasha: not much from me. wasn't able to do anything, been very busy.
shunguo: it was a very busy time last couple of weeks, haven't done much.
<giacomo-petri> scribe+
<giacomo-petri> Dan_Tripp: due to time contraints no much done
scribe+ dan_tripp
helen: change of direction being discussed. due to lack of participation.
helen: since 2022: haven't done
one yet. giacomo has. part of the issue has been: getting a
working example (code snippet). need a developer.
... think about how to improve the process. so we're not
waiting on someone who's busy.
<giacomo-petri> !s/contraints/constraints/
dan: trying to get netlify preview working on a PR so that examples actually work. part of the solution.
sasha: one of the ways, which will still require another person, could be an AI agent to give ideas for examples. would still need to be reviewed by a person.
giacomo: many times AI is doing very bad things (with regard to a11y)
sasha: it's not going to be
correct all the time. need to give a very good prompt.
... might confuse people more. just an idea.
giacomo: to publish a rule, we don't need a code example. but if the description of the example is clear, any developer can help. this is a good exercise to provide clear descriptions.
helen: lots of manual test rules
to be written. if anyone wants to take one on: see what doesn't
have coverage currently, and give it a go.
... and then we can tell wilco jean-yves carlos that the group
(manual test rules group) can stay.
... one of the reasons that the group was maybe discounted:
apart from Trusted Tester, we don't have many such users. so:
is there a want/need for this
?
giacomo: we have manual
methodology in our platform. also semi-automated.
... eg. "link is meaningful". automated tools can help eg.
"learn more". but ultimately you need human validation.
... when we have rules, happy to include our tool in the
list.
helen: intention is to help keep
guidelines true to the tech. and to tell people how to test.
because wcag understanding docs cover the high level and not
every variation.
... eg. accessible help. doesn't mention: is a find function a
help?
giacomo: we shouldn't define in
ACT what's manual vs. what's automated. because potentially,
with AI, something that is currently manual could become
automated.
... eg. "link is meaningful". AI might do this in the
future.
... eg. image without alt attribute. that's automatic. but
we're not saying that it /can't/ be manual - a user could
inspect it.
helen: I would like those
thoughts in an issue.
... that's a good forward-thinking point that we need to be
clear on.
shunguo: basically we have two
meetings. one is rules, one is task force.
... sometimes there's overlap. so if it's possible: merge these
two into one, so we a have consistent path. and
continuation.
helen: task force is for w3c
members. CG: no.
... I agree, but raises question about the paperwork that
daniel would have to do.
... we're doing just one meeting per week now. people are doing
the office hours.
shunguo: I see participation for both is pretty low. and attendance overlapping.
sage: would letting more of CG
people into task force be helpful.
... if I was laid off, I would have more time, and would do
more work on ACT.
... maybe we should reimagine who is allowed in.
helen: TF used to be more
fussy.
... economic climate now is: those who have work are working a
lot of hours.
helen: sharing screen now. github
"beginner issue / help wanted" act rules. does anyone want to
have a look at these?
... this github query that I've done here, is good for finding
something that isn't too time-consuming.
sasha: I can take a look at a
couple of those.
... being assigned something gives you responsibility of doing
the thing.
helen: some of these might not be
needed. you might look into it then figure out it's not
needed.
... here we have a deprecated rule. all the info is in here.
any volunteers? sasha just took 2 of the 5.
shunguo: I can.
helen: error message?
shunguo: again, I can.
helen: would anyone like to take
this last one?
... re: silent video.
... actually, this one is done. will mark it as done.
<giacomo-petri> https://github.com/act-rules/act-rules.github.io/pull/2158
giacomo: you can point to ^^ which is the PR that solves it.
helen: now creating issue on
github for rule classification (automated vs. manual)
... i.e. giacomo's idea from earlier on this call.
... created issue 2370
... due to time, we should not discuss stateful rules. but it's
worth a read.
... re: rule classification, we should make it "subjective" vs.
"non-subjective".
... let's have final thoughts now.
shunguo: we are moving slow. that includes me.
sasha: I hope this group is not going to go away.
shunguo: agree.
sage: agree.
<giacomo-petri> scribe+
<giacomo-petri> Dan_Tripp: hard problem to figure out how to optimize ACT work
<giacomo-petri> Dan_Tripp: might not be solvable because of many reasons: need to read things and re-read things, re-read a lot of things all the times; another time-consuming thing is not easy reading!
<giacomo-petri> Dan_Tripp: I don't know how to solve them, I tried re-wording things where I can
<giacomo-petri> Dan_Tripp: Good luck to us because it's not an easy problem to solve
scribe+ dan_tripp
giacomo: another problem is that we've been working the most on automated rules. so there's not so much to do on automation right now. we're refining. eg. I drafted something on aria required attributes recently. but there's overlap, which slowed down the progress.
<giacomo-petri> https://github.com/act-rules/act-rules.github.io/issues/2368
giacomo: ^^ please review this
before the next meeting.
... applicability "at least one device pixel..." which is very
broad. focus indicator is completely absent eg. "outline: 0".
but once you open it, you open a fly-out. so there's a change
in pixels in the page. but it's not associated with the
control.
... so all of them are passing right now.
... so the rule is too broad. we're allowing some scenarios to
pass even though they should fail.
... so please everyone review if you can.
... the pixel should change within the component.
... not something far away in a fly-out.
... we should say something that is "next" or "close" - not
sure how to word it.
helen: comments on 2.4.7. details lost.
This is scribe.perl Revision VERSION of 2020-12-31 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Default Present: Helen, Sage, Dan_Tripp, sashanichols, giacomo-petri Present: Helen, Sage, Dan_Tripp, sashanichols, giacomo-petri No ScribeNick specified. Guessing ScribeNick: Dan_Tripp Inferring Scribes: Dan_Tripp WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth 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: 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.)[End of scribe.perl diagnostic output]