W3C home > Mailing lists > Public > www-svg@w3.org > October 2007

Re: ARIA support in HTML & XML-based MLs: a call for consensus

From: Gregory J. Rosmaita <oedipus@hicom.net>
Date: Wed, 24 Oct 2007 04:02:22 +0100
To: "T.V Raman" <raman@google.com>
Cc: public-html@w3.org, public-xhtml2@w3.org, www-svg@w3.org, wai-xtech@w3.org
Message-Id: <20071024025825.M85137@hicom.net>

aloha, raman!

1) ARIA 1.0 is merely the beginning -- it is not a panacea, but a triage
effort to keep those dependent upon third-party assisstive software from 
falling even further behind the developmental curve of web 2.0 
applications...  ARIA has 2 tracks:

a) a solution to today's problems (inaccessible widgets) today

b) it provides a framework for maturation and extensibility (due, in no 
small part to its being in sync with XHTML Roles Module (with which, i 
know, you are intimately acquainted)

2) ARIA 1.0 is an intensive triage effort, that not only binds the gaping
holes in scripting languages, but which provides a means for annotating 
and enhancing all users' experience of specialized knowledge domain 
markup languages -- politeness levels are as applicable to any current 
ML as they are to any future ML; and there will always be a need for 
them in order to communicate with out-of-the-box assisstive technologies 
or plug-ins -- the kind most commonly found in the workplace, at 
educational institutions, in libraries and other public-access points 
where a dedicated assisstive technology is either not available or 
out-of-date and limited in its capacities, which is why i put such high 
stock in the work that charles chen has performed in not only supplying 
a self-voicing extension to FireFox, but in supporting ARIA whilst doing 
so...

3) whether you like it, i like it or not, there is and will remain an 
outstanding need for ARIA markup to be integratable with the widest 
variety of markup languages possible -- ARIA is not the "sole" solution,
but it is a standardized approach that fills a gap between native support
(or lack thereof) for role, state, and property metadata that, in 
conjunction with an expert handler communicating with either ATK/AT-SPI
or IAccessible2 to make a specialized markup language usable by someone 
using a third-party assisstive hardware and/or software, and the capacity 
for generic, non-specialized, out-of-the-box assisstive technologies to 
make such interaction feasible

ARIA is framework, not a flagstaff or flogging poll, and it does no one
a disservice to point out that it addresses today's problems today, and 
has the framework in place to address tomorrow's problems...
  
raman, you know as well as i (actually, far better than i) the gap 
between specification and common usage, and the languages you cited as 
having built-in accessibility -- whilst specifying means of providing 
universal access -- are not very widely supported, and not used widely 
in the education of students who depend upon assisstive technologies or 
those who attempting to become experts in a given field...

this is not the forum in which to debate why this dismal situation 
persists, (that would be at the bar at TPAC) but it is a practical fact 
of life online that:

A) languages with strong native support for accessibility have VERY weak
third party support; and

B) if a specification falls in the woods, complete with native 
accessibility features, and no one is there to feel/see/hear it, can 
one really claim that accessibility has been achieved?

4) there is a migratory process from generalized content markup languages
to specialized content markup languages, and a standard framework for 
interacting with and exposing the content of specialized markup languages
is, and will remain, an intractable problem with multiple partial 
solutions, rather than a standardized, comprehensive approach to 
providing full support for specialized content markup which is applicable 
cross-platform, cross-UA, and cross-AT -- that is a true injustice: 
access for the select few, rather than the many...

accessibility is in the eye, ear, fingertip or whatever is available to 
an individual user, and ARIA provides the type of support for 
customization of what is exposed, when, and how; no one has all of the 
answers, which is why ARIA 1.0 is only the first step, but the first step 
is the most important step, for without it, one is stuck in a static 
situation with no relief or recourse...

gregory.
-------------------------------------------------------------
SELF-EVIDENT, adj.  Evident to one's self and to nobody else.
                    -- Ambrose Bierce, The Devil's Dictionary
-------------------------------------------------------------
    Gregory J. Rosmaita: oedipus@hicom.net
         Camera Obscura: http://www.hicom.net/~oedipus/
Oedipus' Online Complex: http://my.opera.com/oedipus
-------------------------------------------------------------

---------- Original Message -----------
From: "T.V Raman" <raman@google.com>
To: david.poehlman@handsontechnologeyes.com
Cc: raman@google.com, oedipus@hicom.net, ian@hixie.ch, public-
html@w3.org, public-xhtml2@w3.org, www-svg@w3.org, wai-xtech@w3.org
Sent: Thu, 18 Oct 2007 05:17:50 -0700
Subject: Re: ARIA support in HTML & XML-based MLs: a call for consensus

> So  how  exactly do you expect ARIA to help with  the following,
> each of which had accessibility support built in from the ground
> up -- but were enumerated by Gregory in his message?  I can only
> think in terms of concrete examples, and I get the sense that
> ARIA as a nonocker is being turned into a generic flag to wave
> around --- or a flagstaff to hit people on the head with ---
> which would do it and everyone a serious injustice.
> 
> Examples from Gregory's message:
> 
> MathML ...
> XForms (my own addition of course)
> 
> -- 
> Best Regards,
> --raman
> 
> Title:  Research Scientist      
> Email:  raman@google.com
> WWW:    http://emacspeak.sf.net/raman/
> Google: tv+raman 
> GTalk:  raman@google.com, tv.raman.tv@gmail.com
> PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
------- End of Original Message -------
Received on Wednesday, 24 October 2007 03:02:40 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:37 GMT