W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2002

Re: definition of text question

From: Al Gilman <asgilman@iamdigex.net>
Date: Wed, 19 Jun 2002 10:00:15 -0400
Message-Id: <5.1.0.14.2.20020619093245.02534260@pop.iamdigex.net>
To: <w3c-wai-gl@w3.org>
Cc: wai-liaison@w3.org

At 05:19 PM 2002-06-18, Ben Caldwell wrote:

>In the June 13 telecon, an issue regarding the use of a standard
>character set was raised and it was suggested that a checkpoint or
>definition would be needed to more clearly define what we mean when we
>say text is provided. (see
>http://cgi.w3.org/ETA/issues.php3/wai/wcag/?issues_id=648 for full
>issue).

The actual accessibility requirement for 'text' in the context of a collection of sub-resources is for "accessible text" not just any old text.

Some of the requirements for "accessible text," such as marking up funny words (for which one needs help to either speak or to understand them) will be developed in the Web Content Accessibility Guidelines themselves.  But you have to start somewhere.  The current W3C answer for the 'somewhere' is the Character Model for the WWW, the Last Call draft of which is at

http://www.w3.org/TR/2002/WD-charmod-20020430/

The WCAG should baseline that "accessible text" is stuff that conforms to the Character Model plus additional constraints introduced here [here = in WCAG].

Otherwise we have a major disconnect between the way the WCAG talks about technology and the W3C technological platform on which we build accessible resources.

For more thoughts on accessible text, please see the preliminary requirements for a timed-text format

 Timed-text Format Requirements
 http://www.w3.org/AudioVideo/timetext.html

In addition, in the PF group we have an open issue as to the names of element types and attribute types in XML.  In order to ensure that an author tool can get assistive technology to represent a given element or attribute name in speech or Braille, for example, it may be desirable to restrict the Unicode characters that may appear in these names.  Control characters are a strong candidate to be forbidden at the XML level.  Dingbats and other symbols appearing in no natural language, for which there is not even a standard way to say a name for the character or transliterate it to a Braille character, are a problematical class.  Should these be disallowed in Web Content Guidelines, in Dialect-definition guidelines, or in XML itself.

We will be discussing this further today.  

Al


>While thinking about a possible definition, Gregg and I came up with th
>following:
>
>In this document, the word text refers to text where each character is
>represented by a number. 
>
>This definition would allow authors to employ multiple file types when
>providing accessible text (ex. HTML, plain text, XHTML, SMIL, etc.)
>while excluding uses of text that are rendered in image form (ex.
>captions that are streamed as a part of the image in a movie/animation
>file). 
>
>However, would it also mean that multimedia content with open captions
>could not meet checkpoint 1.2 at the minimum level?
>
>We weren't sure where to go with this from here, so thought we'd post it
>to the list for discussion/ideas/suggestions.
>
>Thanks,
>
>-Ben
>
>--
>Ben Caldwell | caldwell@trace.wisc.edu
>Trace Research and Development Center (http://trace.wisc.edu)   
Received on Wednesday, 19 June 2002 10:00:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:19 GMT