Amaya, part 1

Note: This is a brief overview, for two reasons: Irene took the detailed
notes of our discussion and plans, and is on vacation, and second because the
Amaya plan is member confidential, and to post details to a public group
which are not already available I need to clear it with the Amaya Team (as a
process requirement as well as a courtesy requirement). Having said that,
here is a brief summary of discussions on Amaya and guidelines 1-3.

In looking at what Amaya does and needs to do, I was generally pleased to
discover that the guidelines seemed implementable for a WYSIWYG authoring
tools such as Amaya. SOme things will be difficult, but overall it seemed
possible to implement a plan that would make Amaya triple-A compliant.
The timescale of that will depend on scarce resources - I would hope that by
the time we go to Proposed Recommendation it is level-A compliant, and that
it is double-A compliant by the end of this year.

Checkpoint 1.1 requires a fair amount of work, particularly in the windows
version. THis is not surprising given the background of the team - none of
its members has specific expertise in User Interface, nor in developing for
Windows. (Further testing includes using the solaris version with Ultrasonix
- the only widely available screen-reader for the X window system that can
compete with the best of the windows screen readers.) However there is
no magic in that, and the fact that the techniques pointed to a number of
reference documents as well as summarising the main points seemed very
helpful. More references would perhaps be a mixed blessing - having them
sorted was a helpful feature.

The rest of guideline 1 Amaya already does - using a local user stylesheet
for documents being edited, providing access to the structure, and quick
navigation thereby, etc. There are a few bugs, and a few features which are
waiting on better implementation of 1.1

Guideline 2 was pretty simple for Amaya, since it only produces correct HTML
4.0, XHTML, MathML.

Guideline 3 was an area where we discussed a number of specific, detailed
techniques - how should a longdesc be implemented in a user interface, what
elements required additional prompting (ABBR, the ability to put in null ALT,
etc). Some of this recurs in teh discussion of guideline 4, to which I will
return after a bit more sleep.

Talk to you all tomorrow

Charles McCN

--Charles McCathieNevile            mailto:charles@w3.org
phone: +1 617 258 0992   http://www.w3.org/People/Charles
W3C Web Accessibility Initiative    http://www.w3.org/WAI
MIT/LCS  -  545 Technology sq., Cambridge MA, 02139,  USA

Received on Tuesday, 27 July 1999 17:54:26 UTC