Proposed wording for SC 1.3.1

John Slatin has a proposal for the wording for SC 1.3.1, based on the
discussion generated at last week's teleconference. Since John is
off-line this week, I'm forwarding it for him.
Loretta

Proposed wording for SC 1.3.1:
<proposed>
Information and relationships conveyed through presentation can be
programmatically determined.
</proposed>

Definition of programmatically determined:
<proposed> 
Programmatically determined Recognized by assistive technology that
supports the technologies in the chosen baseline 
</proposed>

Current wording for SC 1.3.1:
<current>
Perceivable structures within the content can be programmatically
determined.
</current>

Current definition of programmatically determined:
<current>
Recognized by user agents, including assistive technology, that support
the technologies in the chosen baseline </current>

Rationale:
Guideline 1.3 encourages authors to "Ensure that information and
structure can be separated from presentation." The success criteria
(1.3.1-1.3.6) define what must be true in order to ensure that
information and structure can be separated from presentation.

We define "information" as 
(1) a message to be sent and received, and
(2) a collection of facts or data from which inferences may be drawn. 

As of 2 March 2006, we define "structure" as 
(1) The way the parts of an authored unit are organized in relation to
each other and 
(2) The way a collection of Web pages or other primary resources is
organized."  

We define "presentation" as "the rendering of the content and structure
in a form that can be perceived by the user."

Thus the proposed wording means that any messages to be sent and
received, any facts or data from which inferences may be drawn, and any
relationships among the parts of an authored unit, must be capable of
being recognized by assistive technology that supports technologies used
to encode the information and specified in the baseline. "Technology" is
defined in the glossary as markup language, programming language, style
sheet, data format, or API. The baseline is the set of technologies
which the author may assume are active in the user agent.

For HTML and other markup languages, the proposed wording does not
require markup for any information or relationship that is not conveyed
through presentation. (An example of such a relationship would be the
use of some literary symbol to give thematic coherence to a document or
collection of documents.) Again for HTML and other markup languages,
markup would be required to identify any information or any relationship
conveyed through presentation. (An example would be a phrase whose role
as a section heading is indicated by placing the phrase on a line by
itself above some grouping of information, or the logical relationships
among numbers arranged in rows and columns.)

When information and relationships are encoded using data formats that
do not allow for structural markup - such as images, audio files,
animations, and video - WCAG already requires text alternatives under GL
1.1. Where these types of non-text content convey information or present
relationships, the information and relationships are conveyed by the
text alternatives. That is, the proposed SC does not require (for
example) the use of SVG for all images. But where SVG is used and the
image conveys information or relationships through presentation, the
information and relationships must be capable of being recognized by AT.
And since that isn't possible now - even the relationship between an
image and its text alternative(s) can't be recognized by AT yet - GL 4.2
comes into play and an accessible alternative is required - an image
that can be programmatically associated with a text alternative. 


Loretta Guarino Reid
lguarino@adobe.com
Adobe Systems, Acrobat Engineering