See also: IRC log
<scribe> scribe: Brian
presnet+ Brian
<fesch> https://lists.w3.org/Archives/Public/public-svg-a11y/2016Aug/0007.html
fesch: any news, or changes to the agenda?
AmeliaBR: we now have
cross-browser support for tabindex.
... as of last week update to edge we have support, as of
Firefox nightly. There is a 6 week cycle for nightly ot
stable
... Chrome and Webkit already have had tabindex support for a
while
fesch: any news from the SVG working group?
AmeliaBR: We're working on
getting SVG2 to candidate rec
... all normative changes are locked now.
... There is a couple required reviews. The
internationalization team needs to review. There also needs to
be an accessibility review.
... a request was sent out. Not quite sure the process there,
but there will need to be a formal sign off that this doesn't
do anything bad for accessibility. We may be the ones asked to
confirm
chaals-ordhord: Leonie will have a good idea of what happens next
LJWatson: My guess is that it will show up on tomorrow's agenda (APA)
shepazu: I think it's worth
nothing, for whoever who's on that meeting, though we didn't
seek explicit accessibiltiy review before, we have been working
with accessibility people to address accessibility needs in
advacne
... not only do we have tabindex
LJWatson: Fred and I are usually on the APA call, we'll make sure this gets picked up then
fesch: that'll jsut go through the normal channels. Rich, Leonie and I are all on APA. We don't anticipate any problems there
shepazu: During that call, I'd
appreciate, Rich as you point out, we waited to get the review.
But, we didn't wait to get the review. We've been working with
accessibility folks for several years now
... trying to resolve problems in advance. This isn't so much a
review, but a sign off to get to CR
... the other one that got snet out, by the way. The
internationalization, they said they didn't like seeing them
put off so late, but we didn't change much, and most of the
changes for language is with CSS
fesch: what about lang and titles and description
shepazu: I thouhgt we didn't do those changes
AmeliaBR: those are still in. The two changes are multi-title/desc and switch langauge and allow reorder
chaals-ordhord: is the multi-lingual title/desc real? Does it work anywhere?
AmeliaBR: No, that's why it's marked at risk
shepazu: I don't think it'll make it out of CR because we don't have implementations of it
AmeliaBR: I think we need to present it as an accessiibility feature
chaals-ordhord: I hate to say it, but I'd be pretty surprised if that worked
AmeliaBR: there's more work
people catching up on accessibility
... Microsoft fixed keyboard accessibility, and have been
waiting for us to stabilize everything else
fesch: anything else?
these are good to go. They're in the queue, I think behind dpub
AmeliaBR: For SVG2 I added
cross-links for the words for WhatWG spec for shadow dom
... the independent shadow dom spec has stalled, they're
looking to integrate into the core DOM and not have two specs.
The changes are in the WHATWG spec. This is a living
standard
... If you're ok with using that as the definitive reference
for terms, that seems to be the most up to date place
... I haven't pushed the clarifications to the AAM. I need to
check with Michael if I can push changes, now that he's
adjusted changes permissions to github
... It'll either be apull request from the repo, or a fork of
the repo. I can let you know the changes I have marked as todo
are to avoid the use of the term "shadow dom content", use more
precise language
... There's another paragraph I need to clean up, Brad was
saying it's a little confusing
... The main change is being precise in terminology
fesch: Rich, did you have questions baout shadow dom shadowtree based on interactivity concerns
Rich: I think in the last meeting that the events were the same, I don't think we need to differentiate
AmeliaBR: dealing with IDrefs, due to the nature of the use element, and how users don't have control over the stuff inside it I don't see it to be an issue. It's more a problem using the shadowdom and custom widgets.
Rich: When you have a click element inside a tree does it flow up to the parent element?
AmeliaBR: It does. If you have a mousein mouseout type event where it's going in and out of two parts inside the shadow tree, and with the use element you'd be going in and out of the same element it stops bubbling.
fesch: If you actually have a click inside a symbol referenced by a use, and you reference document, you're doing this programmatically, and your referencing document will it give you the parent document or the local document it'll refer you to
AmeliaBR: the document is the sort of containing document. I'd need to grab the spec to get all the details. But, yeah, there's a whole set of new additions about what's contained to the local tree or what's related to the document.
the shadow tree is related but not a separate document, like an iframe is a separate document
fesch: can you programmatically find those?
AmeliaBR: yes
fesch: Would JQuery get confused? Can you find things with it inside those shadow trees?
<AmeliaBR> https://dom.spec.whatwg.org/
AmeliaBR: Not by default,
selectors only look inside one tree.
... This is not naything specific to SVG and use elements.
They're part of the DOM standard, which browsers are making
their implementations match
... We're trying to coordinate as much as possible. the
distinction being use element shadow trees are read-only
fesch: how would you show focus on them?
AmeliaBR: styles are still
independent. That's one important difference
... SVG using the shaodw tree model is different for SVG 1.1.
Things are really messy in browsers in focus and hover
styles
... The shadow tree model requires these are indepdent elements
with indepdenent styles. They are mostly identical, but the
interactive styles only apply to the element you're interacting
with
fesch: The gist of it is, there are still changes to be made in the SVG-aam spec. But the SVG spec is good?
AmeliaBR: Yes, the SG2 spec should be good unless people find problems with it. SVG-AAM has changes.
fesch: any other topics? We can talk about post-aria 1.1
Rich: I can't think of that right now. There's a lot to do still
AmeliaBR: you've held up on updating the test suites because of the graphics roles. Do we now have mappings for the main APIs?
Rich: We definitely have those
for Mac. We need to review the other platforms.
... we'll get there. Mac is there.
<AmeliaBR> https://rawgit.com/w3c/aria/master/graphics-aam/graphics-aam.html#mapping_role_table
Rich: if you expand all and look at mac. We don't get to call it a graphics-document it's just a document. All the differentiation we'd asked for is basically gone
fesch: One question, we have MSAA on the table, should we drop this?
Rich: no, why would be do
that?
... Iaccessble2 is an extension of MSAA, object attributes are
IA2. Oh, there's a typo there
AmeliaBR: as far as the tests,
Fred, are your scripts able to easily update the testing
outcomes based on these mappings?
... if these mappings change in the future, update
accordingly
fesch: I need to put these into
the wiki. Once they're in the wiki, they can convert them into
the form
... What form that is, is an ongoing issue in the ARIA testing
subgroup
<fesch> ack
fesch: I've tweaked the perl
script one time. The other thing that's kind of interesting,
becasue it's what I heard, we were going to try to have
automated tests similar to what Joni did with the webkit layout
tests, so they just sort of run when people build.
... I'm not seeing this happening, or what the general test
harness will look like either. Once we know the form, it won't
be hard to get the tests from the wiki to the proper form
... the people that are building the testable statements for
ARIA 1.1 agreed they'll be using a wiki to test their
statements as well. It's easier than manually creating JSON.
The form is unknown, but the testable statements will be
created in wikis
<fesch> ack
AmeliaBR: that covers a lot. Once I finalize the use elements, there will need to be updates to the testable statements accordingly. Probably some new tests to look at contents inside a use element that needs to be exposed
fesch: I'd suggest everyone look at the wiki. I don't think the wiki is specific to a document. If people think we should break out stuff that should only apply to the graphics document. However we want to handle them that's ok
<AmeliaBR> Root page for testing wiki: https://www.w3.org/wiki/SVG_Accessibility/Testing
fesch: the testable statements are broken out to two pages. One is title/desc with multilingual support. Every other testable statements are on the other one
<fesch> https://www.w3.org/wiki/SVG_Accessibility/Testing/Test_Assertions_with_Tables
fesch: We've sort of had an informal review as we've went along. Joanie did some review. If you want to look really quick, the last row is NOTE. That's Joanie's notes to herself about what test she's doing. If you're editing, just don't change Joanie's note.
AmeliaBR: it'll only change with respect to use elements
fesch: anyone want to go through them and check that they're up to date? They're broken into little groups. Testable statements, separate ones for role none and presentation
AmeliaBR: I will try to go through them. My first priority is getting these edits done. I will hopefully have more time to work on the a11y side now that I'm done with the main SVG spec
fesch: any changes we can take to the ARIA testing group
trackbot, make minutes public
<trackbot> Sorry, Brian, I don't understand 'trackbot, make minutes public'. Please refer to <http://www.w3.org/2005/06/tracker/irc> for help.
RRSAgent: make minutes
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: Brian Inferring ScribeNick: Brian Present: LJWatson fesch Brian AmeliaBR Rich_Schwerdtfeger Found Date: 09 Aug 2016 Guessing minutes URL: http://www.w3.org/2016/08/09-svg-a11y-minutes.html People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]