See also: IRC log
<trackbot> Date: 25 July 2013
<janina> Meeting: HTML-A11Y Task Force Teleconference
<janina> Meeting: HTML-A11Y Task Force Teleconference
<janina> Chair: Janina_Sajka
<scribe> scribe: MarkS
JS: flurry of email following LCWD and some discussion. MathML was brought into the discussion. Would like to have James here, but he will be busy with Elijah.
JS: I have asked chaals to make a word change, which he obliged to. No other comments.
<paulc> Will the ask to the HTML WG occur this week?
JS: we should now seek approval from the HTML WG and PFWG per Plan 2014
PC: when will the WG see it?
JS: probably next week.
JS: touch base on Canvas 2D.
... we should prioritize this evaluation and provide feedback in
clusters rather than one response at the end of the CfC
<richardschwerdtfeger> http://lists.w3.org/Archives/Public/public-html-a11y/2013Jul/0062.html
PC: please let us know if there are no comments on particular sections as well.
RS: I met with Dominc yesterday.
Never really specified what a focus ring is supposed to look
like.
... use system focus indicator. in windows, dotted ring
... author can override default indicator.
... this wouldn't force us to go back to CR
... any objection to go back to HTML to put this note in?
PC: file a bug?
RS: yes.
PC: mention that you have one implementation
RS: Dominic said the code went in yesterday
JS: what are our concerns with testing canvas 2d? is Path in jeopardy?
RS: focus is not in jeopardy. You can
tab though fallback content on at least 3 browsers.
... showing the visible location on the canvas drawing surface, one
implementation in Blink. Hopefully IE and Mozilla next.
TO: I think James is a good person to ask RE apple representative.
RS: Path and hit testing
... hit testing is not going to get into 5.0. Identified as at
risk, implementors want to do something different with Shape
... Path, synchronize canvas closer to SVG. I think Path is at
risk.
PC: yes, and hit regions
RS: If we don't have Path, then we may not have all the support we need.
<paulc> Path Object and Hit Regions are at risk: http://www.w3.org/TR/2012/CR-2dcontext-20121217/
JB: looking for clarification on how canvas can be accessible without path or hit regions.
RS: want to tell a magnifier where
object is. Mags usually follow focus. focusring allows us to do
that.
... the functionality that gets lost is the ability to zoom to
arbitrary information on the drawing surface.
... everything on a web page has a location, mags can zoom to
it.
... not everything in canvas has a location.
... what we can do, is put things in the tab order to control
focus. not the best, but meets bare minimum.
... would tell authors to indicate tab index
JB: sounds as though the hit testing approach is not popular, but there is one that is more palatable? would it fill gap in the same approach?
RS: Developers have been want to use different object, Shape vs Path. Bind that to hit testing. Doesn't exist yet.
JB: would this work on multiple platforms?
RS: this is being discussed by
Mozilla and WebKit and Blink. I don't know about MS
... this would allow us to not have to put everything in the
tabbing sequence and allow us to zoom to important objects that are
represented in fallback content.
... this may not have to wait until 5.1. there may be earlier
implementations.
PC: if this is something that is well spec'ed, you could do an extension spec.
JB: Support the extension spec idea. lets capture this work and move it forward.
RS: i haven't seen shape object yet. I can talk to Rik Cabanier who was talking to Hixie for writing it up.
JS: I think I hear you saying that we should support Shape over Path
RS: I think we should support
whatever will give us the location of objects.
... let me follow up with dbolter on where they are.
JB: if we are trying to insure that
this graphics format is accessible, we can't support both
approaches.
... lets be thoughtful about the timing on this.
RS: in the focusring drawing, two of the api's use the path object. they would drop off without path.
JS: so we would probably wan to keep path then...
RS: I think an extension spec for
shape is great.
... i know that fallback content works well. I have a team that has
demonstrated that now.
PC: does the TF have a schedule to review these two docs given that July is almost behind us.
JS: we need Rich to look through the canvas 2d spec to spot any accessibility snafus that may be in there.
MS: i am willing to help you out with that.
RS: are blink and webkit the same at this point?
PC: [clarifies the definition of unique implementations]
<paulc> 'Each implementation must be developed by a different party and cannot share, reuse, or derive from code used by another qualifying implementation. Sections of code that have no bearing on the implementation of this specification are exempt from this requirement."
<paulc> http://dev.w3.org/html5/decision-policy/public-permissive-exit-criteria.html
TO: there is at least one case in CSS where an Adobe developer has been working on code in both webkit and gecko and we have determined that this are not unique.
JF: google has announced that they will be supporting Longdesc. No indication from Safari. Evidence that these browsers are forking.
PC: same code, one
implementation.
... would like to have a regular agenda item on Canvas and HTML
testing.
<janina> scribe: janina
ms: Has been working with spec
editors on plan to test using issues we found
... Also identified several items we reported as not requiring
anything more than filing bugs
... Will report to the TF
... Expect this next week
http://www.w3.org/Bugs/Public/index.cgi?GoAheadAndLogIn=1
<MarkS> JS: steveF has sent a note to the list that all changes from the standalone document have been ported to 5.1 and 5.0 heartbeat.
<MarkS> ...need everyone to take a really close look to verify that this is what we want in the spec
<MarkS> ...SteveF wants us to file bugs if you have any issues.
<MarkS> ...I think we need to move on this and close it.
<MarkS> ...there is one section, near the beginning RE: a logo, but the alt text does not mention that.
<MarkS> JF: we have a standard or convention that we preface the description with the type followed by a colon. more of a convention.
<MarkS> http://lists.w3.org/Archives/Public/public-html-admin/2013Jul/0058.html
<MarkS> JS: next step, lets return to this topic next week to see how we should handle this.
<MarkS> JF: do we have a deadline on this, or can it wait until Steve's return.
<MarkS> JS: I think 2-3 weeks won't be a problem.