Changes to G117 to support merging 1.3.1 and 1.3.4

At the December 4 team C meeting we discussed changes to the 1.3.1 how to 
meet document [1] to support folding 1.3.4 into 1.3.1.   Andi expressed 
some concern about including G117 [2], "Using text to convey information 
that is conveyed by variations in presentation of text", as a sufficient 
technique because it does not require the variations in presentation of 
text to use semantic markup as is required by 1.3.1. 

This brings up the age old question of what exactly is programmatically 
determined. Changes to the font via CSS can be programmatically 
determined. However, what that CSS styling change means (the information) 
is NOT conveyed programmatically. Thus, changes in presentation made via 
CSS  to convey information would fail 1.3.1 (and this is covered in 
Failure F2: Failure of SC 1.3.1 and 1.3.4 due to using CSS to create 
variations in presentation of text that conveys information without also 
using the appropriate markup or text).   I think the question is,  in the 
case where the user is not conveying the INFORMATION in a programmatically 
determined manner, is using text to convey the information a sufficient 
technique for 1.3.1? 


Here is how I would modify G117 to cover that: 
Description
The objective of this technique is to ensure that information conveyed 
through variations in the formatting of text it is also conveyed in text, 
unless the information conveyed can also be programmatically determined. 
When the visual appearance of text is varied to convey information and the 
information is not programmatically determinable, state the information 
explicitly via text. Variations in the visual appearance can be made by 
changes in styling or via markup elements such as HTML strong, em, cite 
and others. When variations are made via style changes only, the 
information conveyed can not be programmatically determined and that 
information needs to be available elsewhere in the content via text. When 
changes are made via markup elements which convey meaning such as strong 
and em, the information can be determined programmatically.  Including 
additional sections in the document or an in-line description where the 
variation in presentation of text occurs can be used to convey the 
information.
I think the examples can be used unchanged since they are examples where 
the information can not be programmatically determined and thus the 
information must also be provided in text. Perhaps provide an introduction 
to the examples:
Examples:  The following examples demonstrate situations where the 
information conveyed by the changes in presentation can not be 
programmatically determined and an additional text description is 
required. 
example 1 and example 2 remain unchanged. 
Add to the related techniques section:
G115: Using semantic elements to mark up structure 
Update the test section:
Procedure
1.      Find items where variations in presentation of text are used to 
convey information <ins>and the information can not be programmatically 
determined.</ins>
2.      For those items, check to determine if information conveyed 
<ins>only</ins> visually is also stated explicitly in text.
Expected Results
Check #2 is true.

[1] 
http://trace.wisc.edu/wcag_wiki/index.php?title=In_progress_update_of_HTM_1.3.1
[2] 
http://www.w3.org/WAI/GL/WCAG20/WD-WCAG20-TECHS-20060801/Overview.html#G117
[3] 
http://www.w3.org/WAI/GL/WCAG20/WD-WCAG20-TECHS-20060801/Overview.html#F2

Becky Gibson
Web Accessibility Architect
                                                       
IBM Emerging Internet Technologies
5 Technology Park Drive
Westford, MA 01886
Voice: 978 399-6101; t/l 333-6101
Email: gibsonb@us.ibm.com

Received on Monday, 11 December 2006 20:55:49 UTC