<trackbot> Date: 28 February 2011

http://lists.w3.org/Archives/Public/public-canvas-api/2011JanMar/0075.html

<scribe> scribe: Rich

<scribe> Meeting: HTML 5 Canvas Accessibility subteam meeting

Bounding Rectangles Fallback content

Rich: I have been thinking about coords

Chuck: I have been using CSS to give it a rectangle but it works rather well
... The value of CSS is that if canvas could not be drawn it could be used
... The SVG group is working on on accessibility as well
... The best thing would be if the AT could send a request to the scripting environment
... It sounds like focus ring would work but we are talking about access to things that do not have focus

Rich: Correct. That is what AISquared wanted

Chuck: I see what they want but I am trying to find an elegant way of doing it
... it really should be callback based
... There are ways for the AT to walk the DOM.

Rich: moving focus to get things messes up the user experience

Chuck: I think for zoomtext that there are features they can use to get what they want
... This does not quite fix how canvas works
... Instead of putting all the things in the DOM I add or remove things from the DOM.

Rich: If you pull things out of the DOM this could mess up review mode

Chuck: the situation where you treat all elements like a list it is more canvas like
... I can say I have all of these and I need to rotate them all
... SVG benefits from being bound to the drawing surface

Rich: HTML is the same way. We just are not bound.
... is there a way to create an object in canvas that could be bound to the corresponding fallback object.

Chuck: there is only one object for the canvas
... if you are looking at an entire haptic screen ...
... a Braille devices is more text based
... I could use some use cases.
... At this point it is mostly a touch based screen
... there is no real drawing object. There is just drawing primitives. There is no positioning information

Rich: I wonder what the value would be if we had a drawing object.

Chuck: Has a create path item in Canvas API 2.0

Rich: it might be a good thing to create a collection of drawing functions into an object and compute the bounding rectangle

Chuck: Essentially, the SVG path provides a similar function
... ARIA is rich enough. If you set an ARIA tag on an element then you do this ...
... right now I am going through my mac. etc.
... there is nothing in these ATs that require that yet
... The text finder in Zoomtext is the closest. It uses focus

Rich: I will look at SVG path

Chuck: OK
... it is perfectly valid to use the SVG path.

Caret Simulator

Chuck: I have been working on it in darkroom
... the drawfocus ring using divs does work
... I put a div in a div in contenteditable and got the caret blinking
... I set the text color to be transparent. I used absolute positioning.
... It worked

Outstanding defects 11328, 11329, 11342

Chuck: Nothing has happened yet
... no success yet

Rich: I will look at SVG paths

Chuck: I will modify darkroom demo to address moving your mouse over something to hear what is going on.

<Downchuck> might try to track down some bspin devs

<Downchuck> they weren't treated very well in mozilla for their use of canvas; but they have good input on things

This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

No ScribeNick specified.  Guessing ScribeNick: richardschwerdtfe
Found Scribe: Rich
Default Present: Rich, Chuck_Pritchard
Present: Rich Chuck_Pritchard

WARNING: Fewer than 3 people found for Present list!

Found Date: 28 Feb 2011
Guessing minutes URL: http://www.w3.org/2011/02/28-html-a11y-minutes.html
People with action items: