<Jemma> Meeting: ARIA-APG
<scribe> scribe: MarkMccarthy
[chit chat, waiting for participants]
<Jemma> Subsequent meetings: December 15 and Jan 12.
<Jemma> No meeting December 29
mck: meetings are on Dec 1 and 15, may revisit January meetings on the 15th meeting
mck: thinking ahead to first half
of 2021. a lot of our work is reactive (responding to PRs,
questions, issues, etc)
... some of that is just because of the way the community works
and talks to each other
<Jemma> next priorities include ARIA 1.3 support
<Jemma> Redesign project
<Jemma> Navigation systems
<Jemma> Coverage
<Jemma> Code quality
<Jemma> Mobile support
mck: but i'm thinking of how to
be more focused on key areas, overarching themes in issues, or
chatter outside of communities that could be related to
APG
... what does everyone think of that? or other input on
priorities or reactivity vs. proactivity? thoughts on the list
[above, pasted by Jemma]?
Jemma: so the redesign project is what we talked about with EO right?
mck: yes, moving onto WAI website
Jemma: i'm interested in mobile support. we've made significant progress on quality with our examples, but what are we doing for mobile?
mck: i think we're at the beginning of answering questions around navigation - not quite done ---
Jemma: what else needs to be done?
mck: mostly becuase we have lots
of issues outstanding; lots of examples lacking cohesion; I
don't think we've got good guidance at the moment
... things will have to stay in draft until we have some
consistent examples
... the updated menubar from Jon and then Sarah's work...once
we get those in we're off to a really good start
carmacleod: that's probably when we should put that new section in. that is, once those are done, just put em in
mck: i'd love to, hopefully by the new year. but time is short
carmacleod: indeed, lots to do. if we have focus on those three things getting polished, and finishing the draft of the new section, it's _only_ four things
mck: those are all listed in issue 89
<Jemma> https://github.com/w3c/aria-practices/issues/89
mck: mobile is probably the biggest untouched thing, and we're getting good feedback about compatibility
Jemma: maybe we can set micro and
macro priorities. redesign project will take a long time, but
nav systems could be done in 6 months let's say
... so i think we'd benefit from setting short and long term
goals
mck: one thing about redesign:
it's kind of standalone since it's not quite the core of what
we do, it's improvements to delivery. so i'm comfortable having
vendors etc. work on that
... whereas mobile support and nav systems are the core of what
we do
Jemma: +1
... and, i think some limitations of our examples (like
responsive design) will be solved during the redesign
mck: yes and, the mobile item on
the list here is about providing patterns for folx who want to
implement patterns for small-screens and/or touch devices
... maybe we should clarify "mobile support." for some
patterns, there might not be enough support on the mobile
browsers. so maybe _those_ need warnings or something
similar
... in other words, mobile support may mean saying it's _not_
supported
<Jemma> +1 for clarification of "mobile support" to distinguish it from delivery platform issue, APG redesign project.
carmacleod: i wonder if we should
look around and try to recruit more people. we're taking on a
lot of work and some of us, at least myself, have no experience
working on mobile devices
... it's a big area, and i personally feel i can't contribute
enough. but there _has_ to be folx out there that are
wizards
mck: YES. to do it well, we certainly may need more people
carmacleod: jamesn, maybe the
slack channel would be a good avenue to post?
... or maybe we could all discuss who some of the right people
would be
jamesn: when i come across those that seem receptive to our work, I do try to recruit them. but open ended things like that tend to garner negative responses or the wrong people
Jemma: so we're looking to figure out what the major problems are with touch devices and other small screens
mck: eventually, we want the work
we do to support the whole community, not just us. in some of
these cases, multiple flavors of an example might be good
... for instance, things might be done differently if using
touch or small screens
... some of what mobile support entails is workarounds, and we
don't always specialise in workarounds. we demonstrate
ARIA.
... as we expand our scope, we might want to have _some_
examples that _do_ have workarounds to get information out
there, but then streamline them later as we continue
working
Jemma: so our first priority next quarter is working nav systems, this other stuff running concurrently?
mck: yes
... if someone wants to help lead this mobile space, recruit,
etc., that'd be awesome
Jemma: i can help! but would love some other hands too
mck: so maybe a recurring agendum
once a month, progress reports etc.
... I'm feeling that we'll need to work on having measures of
success in order to keep money flowing and having vendors help
us out
... so "Reach" is one kind of success, and give us some hard
data about things. so blog posts or MDN resources, etc.
Anything that can share our content
... if anyone has any ideas on how we can do this, would love
to hear!
carmacleod: i think the CodePen buttons are huge for reach
jamesn: as long as people don't share their modified version of the APG example as the de-facto
mck: definitely, that seems to have been very helpful
jamesn: i'm wondering if we can get metrics from the website; page views etc.
mck: that'd be good, boucoup can look into that, as can Boaz
jamesn: MichaelC might know, but he may not
mck: Boaz is pretty good at that
Jemma: Shawn shared the analytics
data for the links, so maybe she has some data too?
... jamesn, did you mean getting feedback from CodePen
clickthroughs?
jamesn: no - saying that people will save the modified codepen as a reference, rather than the "official" example
Jemma: gotcha
mck: does anyone have ideas of solid success measurement?
Jemma: not quite success meausrsement, but I'm thinking about possibility of integrating ARIA-AT etc. on CanIUse
mck: definitely investigating
it
... if anyone has ideas, let us know! it's nice to go out and
talk about what you've done besides just listing off a pattern
you worked oon
<Jemma> https://a11ysupport.io/
mck: nav treeview example looks to be coming together well
<Jemma> https://github.com/w3c/aria-practices/pull/1558
mck: carmacleod, you just have to look at functionality?
Jemma: i can do more work on the a11y review but the design review is done
mck: there was some discussion on
aria-current; is there a common visual convention we should be
emulating?
... i.e. what's the most common visual indicator of what the
current page is?
Jemma: usually a border, font, or color change. haven't seen arrows other than something like an accordion
carmacleod: haven't seen an arrow in a while
jamesn: --it's an old thing
carmacleod: maybe flipping the colors, or bordering
mck: but that's not WCAG compliant?
jamesn: it could be; as long as
the luminosity changes, it's not relying on color alone
... luminosity _AND HUE_
Jemma: visual design, I think i'm confident on what i've done
mck: great. did you find anything else that needed to be addressed?
Jemma: i'm not sure how we can deal with touch target size, since it's not responsive
mck: they're too small?
Jemma: yes
... color contrast review is good, that should help out Jon a
little
mck: so re: touch target, if those things are too small, is it difficult to resize or do we have room? it's a AAA requirement though?
Jemma: yes, i'll have to double check
jamesn: so, is it just me or is the focus movement really weird?
carmacleod: yes
jamesn: it's fine for a keyboard user, but if a mouse user clicks on an item in the tree, it shouldn't move _their_ focus
mck: okay, but where's their focus beforehand?
jamesn: fair enough - but i shouldn't move it away when i just click in there. what if I wanted to use the arrow key?
mck: true! if you were using click to bring focus in the tree...
jamesn: yep, which isn't uncommon
mck: so that focus change is likely happening based on the link getting activated
jamesn: yep
mck: it's _really_ helpful for a screen reader user, especially in a SPA, which i suppose is our envisioned ideal audience
jamesn: but a visual keyboard user _won't_ want their focus to move when they want to expand the tree
MarkMccarthy: if shift-tab even _goes_ back to the tree
jamesn: yeah. i think focus should stay where it is, and it's up to the screen reader user to navigate away in the way they want
carmacleod: it seems really nice to me, arrows and all
jamesn: as a keyboard user, i go
to the apply page "oh the info I wanted wasn't on that page, it
was on tuition page" and my focus is gone and I have to get
back to the tree
... an _application_ is different; on a _webpage_ you usually
want to read stuff
mck: most of the time in
treeviews, the whole thing gets reloaded so you're back at the
top
... for an SPA, we've had good feedback at Facebook about
primary nav activations only refreshing part of the page
jamesn: Facebook is different
from typical webpage navigation on a webpage
... if you select something on Facebook nav, you probably are
going to spend time _in_ the page you selected. a typical
webpage, you might be reading/scanning through info. it seems
very different to me
mck: most of the time when you click links in typical nav, you start over. so you usually don't have that type of behavior anyway
jamesn: i don't think it's that unusual these days
Jemma: +1
jamesn: so I think if we're targeting correctly, that's okay. but this pattern seems to be optimized for a screen reader user, not a keyboard user
mck: something in the
accessibility feature section noting these difference might be
a good approach
... could folx also give feedback on the caution on top of
this?
carmacleod: needs something to stand out, a yellow box
jamesn: yup
mck: could someone make a <span> style for me to apply?
carmacleod: also missing the
green box we usually have
... oops, nevermind. the box is there. so same sort of style
except yellow for caution
mck: yep, having a caution style would be good
Jemma: so we're adding a note about use cases for the tree view, that approaches may have to be different.
mck: i'm fine with, for mouse users, focus not moving, or moving it to the tree
jamesn: 100%
carmacleod: that's not quite how the disclosure works though
mck: james, what do you think about that?
jamesn: the menu is different. once it closes, focus _has_ to leave it.
mck: is there a strong argument for making them different then? sometimes you want a pattern because you _do_ have a different kind of experience
jamesn: my general rule is, if the thing focus is on is still on the page, focus should stay with it.
carmacleod: they're same page links though
jamesn: yes, but in this case they don't look like that, so someone might not expect it
CurtBellew: there are other links on the page, outside the tree (or a link _inside_ tuition) that look like links. i see james' points - conceptually they're navigation to me, not links
jamesn: i'm against this going forward as is
CurtBellew: +1 - our answer is usually leave focus where it was activated
MarkMccarthy: me too - +1
carmacleod: unless it's a same page link
mck: still kind of debating the subtleties between menu, tree, disclosure. it's okay if we can clearly explain the reason for difference
carmacleod: it's the arrow keys
mck: that exists in menubar and
disclosure too
... the new disclosure pattern has a link and a button. my
assumption is the link moves focus, the button expands the
view
... i'm kind of wondering if this is a visual styling issue
jamesn: i'm also wondering if
this fails 3.2.2
... --change of context.
CurtBellew: that's what i was thinking too
<Jemma> https://www.w3.org/TR/UNDERSTANDING-WCAG20/consistent-behavior-unpredictable-change.html
mck: so it's not an input in that sense, but there's a little bit of ambiguity here
jamesn: [reading the success criterion definitions]. in my reading of that, if it was tabbing through it wouldn't be a single function
mck: you could easily have a button in a toolbar that moves focus
jamesn: a button is there and
seen as being okay (it's expected to change). it is ambiguous,
yes.
... these "same page links" are being exposed as
treeitems
... and look like that to visual users
<Jemma> thanks for exteneded scribing, Mark
mck: even though it's a link in the code, we have to remove the link behavior, unless someone is using a tree to navigate. by the logic here, it'd be illegal to use a tree if it loaded a new page
jamesn: if it does that, it's not technically a tree. yes, it's ambiguous. as is, it's very confusing
mck: is it clearer if we change the click behavior?
jamesn: clearer, but doesn't help a sight dependent keyboard user
MarkMccarthy: we are over time
CurtBellew: this is an interesting discussion!
mck: i don't think there's one right answer. i think there's more conversation to be had
jamesn: there are some places
where you know the user is going to want to interact with the
tihing on the right hand side
... it's almost an exception to move focus rather than a
rule
CurtBellew: thinking of a document on the right side, nav links on the left (onedrive online or similar)
jamesn: file explorers work that way too
carmacleod: you stay in the tree on the left until you tab to the right side in Windows' File Explorer
MarkMccarthy: yep
mck: we'll have to sort things out, seems like this is difficult
CurtBellew: the roots run deep
[playful jeering]
<Jemma> LOL - roots run deep
mck: thanks everyone, sounds like we could use some more deep dive discussions
carmacleod: VS Code could be worth some exploring too. clicking keeps focus in the tree, enter focuses into the code
jamesn: that's not what _I_ have with VS Code (on the Mac)
carmacleod: I'm on Windows.
[scribe has to drop off]
This is scribe.perl Revision of Date Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: s/present +// Succeeded: s/anda/and/ Succeeded: s/definitly/definitely/ Succeeded: s/?// Succeeded: s/AND HUE/_AND HUE_/ Succeeded: s/focus'/focus/ Default Present: mck, Jemma, carmacleod, sarah_higley, jongund, Jamesn, MarkMccarthy, Bryan, CurtBellew Present: CurtBellew Jamesn Jemma MarkMccarthy carmacleod mck Regrets: Jon jongund sarah_higley Found Scribe: MarkMccarthy Inferring ScribeNick: MarkMccarthy 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: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. 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]