W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > October to December 2004

Re: WAI comments on SVG 1.2 last call working draft

From: Chris Lilley <chris@w3.org>
Date: Mon, 22 Nov 2004 12:07:33 +0100
Message-ID: <1412101138.20041122120733@w3.org>
To: Jon Gunderson <jongund@uiuc.edu>
Cc: Al Gilman <Alfred.S.Gilman@IEEE.org>, w3c-wai-ua@w3.org, w3c-wai-pf@w3.org

On Monday, November 22, 2004, 7:45:42 AM, Jon wrote:


JG> Al,

JG> Here are comments from the user agent working group on the
JG> last call working draft of svg 1.2 document.

JG> http://www.w3.org/WAI/UA/2004-11-wai-comments-svg12.html

JG> These comments need to be sent to the svg working group by
JG> Wednesday the 24th.

But if you need a couple more days, just say so.

JG>   If anyone from ua or pf has comments
JG> please get them to Al Gilman or myself before Wednesday.

I have a couple of informal comments having read the draft:

a) It would be helpful to indicate ways in which you feel the specification
does not meet your comments. I am thinking in particular of

 Color information must be made available through accessibility
 APIs (UAAG Checkpoint 6.4 )

Since it *is* already available via the DOM, what else are you asking
for? Similarly

    Allow users to define a compositing operation be applied to
    the text element of an SVG document
    Allow users to define a compositing operation be applied to
    the entire SVG document

text {filter:url(#mycustomfilter) }
svg {filter:url(#mycustomfilter) }

in the user style sheet?

B) It would also be helpful to propose an algorithm, especially when an new
requirement seems to contradict the laws of physics and information
theory:

Gray Scale Viewing Option
    Colors are converted to a grayscale values
    Colors with the same luminosity need to adjusted to make them
    distinguishable as gray scale

The first one is easy. The second one is a research project at best.
Consider for example:

"Text in mixed case must be converted to all lower case. However,
letters which were of different cases must be distinguished". Its rather
like that, except more so. Information has been put back. While the grey
values can be adjusted in various ways, while mapping form a much larger
set to a much smaller one there will always be values that were distinct
in the larger set but not distinct in the smaller set. That's simple
information theory. So, could you clarify that one a little so its
physically possible?

C) The comments on accessible examples of flowing text (and how it could be
more accessible to use this feature) are well made, thats a very good
idea.

D) The points about captions are well made, we are not sure either. Some
media include video, audio in multiple languages, captions in multiple
languages; its not clear how the SMIL model indexes into these (althouh
we have asked them that, for the case where audio is to be controlled
separately from video, so its the same thing here).

E) A bunch of the separate control, start, stop, slow functionality is in
SVG 1.2 and was not in SVG 1.1, because of time containers. This shoud
have been more explicitly pointed out.

F) It does seem that we should have a 1.2 accessibility appendix - the 1.1
appendix also applies, but for the new features there are new
accessibility opportunities that need to be pointed out. Would anyone o
this group be interested in working with SVG WG to create such an
appendix?

JG> Jon Gunderson, Ph.D., ATP
JG> Coordinator of Assistive Communication and Information Technology
JG> Division of Rehabilitation - Education Services
JG> MC-574
JG> College of Applied Life Studies
JG> University of Illinois at Urbana/Champaign
JG> 1207 S. Oak Street, Champaign, IL  61820

JG> Voice: (217) 244-5870
JG> Fax: (217) 333-0248

JG> E-mail: jongund@uiuc.edu

JG> WWW: http://cita.rehab.uiuc.edu/
JG> WWW: http://www.staff.uiuc.edu/~jongund






-- 
 Chris Lilley                    mailto:chris@w3.org
 Chair, W3C SVG Working Group
 Member, W3C Technical Architecture Group
Received on Monday, 22 November 2004 11:07:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 October 2009 06:51:27 GMT