RE: Discussion of DOM with Glen Gorden of Henter-Joyce (A) (A)

I'm not sure that the goal of cross-platform AT devices is even a reasonable
one.  For a DOM-like interface, or even MSAA, we are talking about code that
runs on the host platform.  It would theoretically be possible to write AT
software in Java, and have it run cross platform, but at the current state
of the art, even that is fairly platform specific.  This means, in practice,
that AT must be developed independently for each platform.

However, the concepts behind a generic AT interface are good ones.  What I
have problems with is the idea that HTML documents would use one generic
interface, word processors another, and spreadsheets still another.  This is
one step better than having each program use a completely different
interface, but not as good as the concept of a broad interface for AT that
would communicate with equal facility to browsers, word processors,
spreadsheets, and databases.

I agree with Chuck that asking Unix boxes to support MSAA would be fatuous.
On the other hand, the types of calls that are found necessary on any given
platform would probably be useful on another, so the API developed for MSAA,
for example, might well be a guideline for developing a UnixAA, or MacAA.

Something like 8 years ago, Randy Marsden of Madenta Communications proposed
an AT interface very similar in concept to MSAA for the Apple platform.
While others in the industry thought it would be a good idea, no one wanted
to invest the energy to make it happen.  Microsoft has, at least in part,
made it happen in Windows.  That work should be extensible to other
platforms, and make building AT more a user interface issue than a reverse
engineering of each application, be it browser or video game.

Denis Anson, MS, OTR
Assistant Professor
Computer Access Specialist
College Misericordia
301 Lake Street
Dallas, PA 18612

RESNA
The International Organization of Assistive Technology Professionals

Member since 1989



-----Original Message-----
From: w3c-wai-ua-request@w3.org [mailto:w3c-wai-ua-request@w3.org]On
Behalf Of Charles McCathieNevile
Sent: Monday, February 08, 1999 8:07 PM
To: Charles Oppermann
Cc: w3c-wai-ua@w3.org
Subject: RE: Discussion of DOM with Glen Gorden of Henter-Joyce (A) (A)


If we had a DOM which encompassed the scope of web documents (in which I
am including multimedia 'documents' and the range of other beasties that
are out there on the horizon) then the DOM would indeed be a very powerful
approach. It is also completely Open and platform-independent, which means
it is possible to develop applications which use it in standardised ways.
However, DOM level 1 is a long way from that promise. It seems that the
group needs to decide whether it wants to support DOM 1 and thereby
signpost its expectation that DOM2 and DOM3 and DOM4 and so on will
represent the best way forward, or whether the group would rather leave
the question of what interface should be used to User Agent developers.

Charles McCathieNevile

On Mon, 8 Feb 1999, Charles Oppermann wrote:

  Remember what DOM means - Document Object Model.  It's not an assistive
  technology interface, it's not even a user interface object model.  It's
an
  object model for documents - HTML documents to be specific.

  Text object model developers find DOM inadequate to represent higher end
  markup and layout.

  I caution the group not to put too much stock into DOM.  While I feel it's
  very useful to improve access to the web content - that is one small piece
  of a users experience with a computer.


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

Received on Tuesday, 9 February 1999 08:37:03 UTC