Analysis, impact of externalized baseline decision on techniques

Ben, Becky, and I had the action to analyze the techniques to determine if there were issues caused by the externalization of baseline from the guidelines. We started by reviewing the HTML, CSS, and Script techniques against three hypothetical baselines:

1) A low-level baseline in which few assumptions are made about user agent support;
2) One appropriate to present-day mainstream browsers with CSS and Script supported and enabled;
3) Hypothetical user agents of the future in which the accessibility features of most technologies, including scripts and plugins, are supported.

For each technique in each baseline, we assigned the technique the following status:

a) The technique is sufficient to meet the success criteria in this baseline, possibly in combination with other techniques;
b) The technique is optional in this baseline - it may not do much good but it does not do harm;
c) Not recommended in this baseline - following the technique actually creates problems of some kind.

The raw data with notes is in the attached files.

In analyzing our results, it seemed that, with a few issues to resolve here and there, it was possible to categorize the techniques into the baselines and create sets of techniques within each baseline. Also, importantly, it appears possible to create a set of techniques for the low-support baseline, a different set for the high-support present day baseline, and the two sets could be combined into a recommended baseline that results in accessibility in nearly every user agent today. [Sorry, I may have to explain this paragraph better on the phone.]

There was, however, a major caveat to our analysis. The script techniques currently available were written to target the lowest baseline - that is, they assume scripts are not supported. We could not complete our analysis until we have examples of techniques addressing direct accessibility of script, to determine if things would break down with those techniques.

We have taken on an action item to propose - in very rough form - some script techniques in order to complete our analysis. The highest priority techniques would be for script features that cannot be replicated in HTML. We defined "cannot be replicated in HTML" as meaning "without requiring roundtrips to the server".

Our plan is to have this draft of techniques by next Tuesday, April 12.

--- Signature ---

Michael Cooper
Accessibility Product Manager, Watchfire
1 Hines Rd Suite 200, Kanata, ON  K2K 3C7  Canada
Tel: +1 (613) 599-3888 x4019
Fax: +1 (613) 599-4661
Email: michaelc@watchfire.com
Web: http://www.watchfire.com/