See also: IRC log
https://mit.webex.com/mit/e.php?MTID=mfbccb8d6db27f5aa089c12da792afa6d
<scribe> scribe: AmeliaBR
Fred: How many more meetings before end of the year?
Rich: No more than 2. Who's going to be available next week?
Fred: I would like to meet next week to follow up on the testable statements work in the main WG.
Rich: I'd also like to get started on the test suite. That gives people things to work on while others are on holiday.
Chaals: I'm available next 2 weeks, but then will be in Australia which doesn't work with this time.
<chaals> [I'll be OK for next 2 weeks, then in Australia, where this meeting is not at a reasonable time. I'll be probably back on 22nd Jan.]
RESOLUTION: The Task Force will meet Dec 11 & 18, won't meet Dec 25/Jan 1, then will reconvene Jan 8.
Fred: Since the TF is split between different WG, it's hard to use the tracker for Actions. Michael Cooper has offered to set up a specific Tracker for the task force.
(no objections/comments)
RESOLUTION: Ask Michael to set up an issue & action tracker for the task force.
Fred: Rich, are you still planning to set up a poll to consider a new telcon time? Rich: I'll do that in January
<fesch> amelia: issue with role test - will it add extra functionality? Unlikely new functionality get into SVG 2.
<fesch> amelia: support for general idea but not likely to happen for SVG 2.
<fesch> amelia: a couple TODOs on mapping spec
<fesch> amelia: need to go through and do research on current behavior, before we make proposals for changes
<fesch> https://www.w3.org/wiki/SVG_Accessibility/Testing
Fred: The ARIA WG is moving ahead with setting up new tests for ARIA specs, we will need to coordinate
<fesch> https://github.com/SVG-access-W3CG/svg-a11y-tests
Fred: the wiki list is a set of
testable statements I put together
... Chaals and Leonie have been putting together test case SVG
examples on GitHub
... I would like someone else to go through my testable
statements list & identify gaps. For starters, I don't have
anything related to roles.
Chaals: It looks like a good list, but I'd like to see them turned into actual tests. Do you plan to do that?
Fred: I will, but I'm waiting on the ARIA WG to decide on a template or format, so I can work with that.
Leonie: If the list could be copied to the GitHub repo, that would make it easier to review, update, and file issues.
Chaals: Even if there are missing things, we can start making tests, and update or delete as needs be. Start with what we've got & then add more.
Fred: So I should start with the format you've already got?
Leonie: If you clone the repo, there's already a template file in there.
Chaals: You may also want to talk
to Michael, I know they have some JavaScript magic to put the
proper test structure.
... but you can start by converting the prose into actual SVG
code.
Fred: Does someone want to review to be sure my tests are correct?
Amelia: I will eventually, but it may need to be an iterative process.
Chaals: I can look at it this week.
<chaals> ACTION: Chaals to review Fred's test proposals as far as possible in a week. [recorded in http://www.w3.org/2015/12/04-svg-a11y-minutes.html#action01]
<trackbot> Created ACTION-1757 - Review fred's test proposals as far as possible in a week. [on Charles McCathie Nevile - due 2015-12-11].
Fred: I also need to look at the roles, I only looked at name and description
<scribe> ACTION: Fred to expand testable statements to include roles. [recorded in http://www.w3.org/2015/12/04-svg-a11y-minutes.html#action02]
<trackbot> Created ACTION-1758 - Expand testable statements to include roles. [on Fred Esch - due 2015-12-11].
Fred: Looking forward, how will the tests actually be done?
Rich: There hasn't been the bandwidth to set up an automated process yet.
Chaals: So far, it's been manual, testing and then writing results into an HTML file. If we could at least set up a system where someone can just click yes/no and have it submit the data, that would be helpful.
Fred: Chaals, you've been working on this.
Chaals: We have put together a
collection of images; I've been starting with ~5 of them,
trying to make them as accessible as I can.
... It's fairly straightforward for tree diagrams, chord
charts, and basic bar charts.
... I haven't touched anything interactive.
... I also haven't really used any of the new roles.
... I'm mostly focusing on trying to make things work now, with
keyboard nav, which means not relying much on ARIA.
... Existing implementations are so bad, that we can't rely on
it.
... Going forward, we want to create best practices examples.
Which means making code that sometimes seems like stupid
choices to an author, to work around the rubbish browser/AT
implementations.
... But the other thing is to make examples of what *should*
work, based on the specs, so that we can take them to the
browsers & ATs and tell them to fix it.
Fred: One thing, is how to address interactive issues. Does W3C have interactive tests?
Amelia: Yes, you just need a clear script of what the tester should do and what the outcome should be.
Chaals: We definitely need to test interactivity, because navigation and many other functions are inherently interactive.
Fred: Are there interactive SVG tests? Should we follow their format?
Amelia: There are in existing test suite, hasn't been updated in a while. SVG may be switching to the test system used by CSS, which is more automated.
Chaals: For interactivity and other features, we need to write up these ideal examples.
Amelia: I would like to create a system of paired examples: For each example, write it up once with the ideal, minimal markup, and once with the current best-practice redundant markup. The redundant markup would be current advice for authors, but it would also help emphasize to browsers/AT why they need to support the simpler declarative/semantic syntax.
Rich: We need to sync up with SVG as far as test suites. E.g., tabindex is SVG 2. Has any work been done in SVG WG about getting test suites for that?
Amelia: Not that I've heard. The current focus is getting the spec finalized, tests will come next.
Fred: Should we even address that in our tests, then, if it's part of the main SVG 2?
Amelia: I think we should, so long as we make sure we're coordinated and consistent.
Rich: We may also have features that we're more concerned with that we want to emphasize.
Chaals: It would be useful if we can include our tests within their test suite, so that they are clearly visible when someone is testing. But if that's not possible, so long as we push forward and make sure people are reminded of the results when tests are being failed, that's fine. It's just more work for us to emphasize it.
<fesch> http://www.w3.org/TR/gamepad/
Fred: I've been working on the gamepad spec for APA. They use SVG as <embed> in an example. I was hoping to work with Jason to help mark that up ideally.
Amelia: If there are examples of how SVG should behave in other accessibility specs, it makes sense for us to review for best practices.
Chaals: There a lots of SVG figures in W3C specs, and most of them are garbage markup from authoring tools.
Amelia: That might be over-reaching to try to fix every figure.
Fred: But if I see similar examples and figures, can I grab them or bring them to peoples' attention?
Chaals: I have been doing that, putting them in the repo. Please use the repo for any similar example.
Rich: Right now, for name and
description, this is a core issue, need to review the proposals
Amelia made.
... I also need to write up a mapping spec for the new mapping
spec for the new graphics roles. This is separate from the SVG
mapping, as these roles may be used with other languages. But
it should be a short spec.
... I will also follow up with putting pressure on browsers to
get support for key things, such as tabindex.
(Discussion on current sorry state of tabindex support in SVG)
trackbot, end telcon
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) Succeeded: s/Resolution/RESOLUTION/ Succeeded: s/time?/time? Rich: I'll do that in January/ Succeeded: s/TOPIC: Name and Description/TOPIC: Specifications/ Succeeded: s/Right now/Right now, for name and description,/ Found Scribe: AmeliaBR Inferring ScribeNick: AmeliaBR Default Present: Fred, AmeliaBR, chaals, Rich_Schwerdtfeger, shepazu, LJWatson, Jason, fesch, Doug Present: Fred AmeliaBR chaals Rich_Schwerdtfeger shepazu LJWatson Jason fesch Doug Rich Found Date: 04 Dec 2015 Guessing minutes URL: http://www.w3.org/2015/12/04-svg-a11y-minutes.html People with action items: chaals fred[End of scribe.perl diagnostic output]