Re: Debunking the need for "text-only" parallel sites

I don't think maintenance is much of an issue in situations where production
is sufficiently automated. The issue is about creating seperate classes of
site - although text-and-audio only serves blind users, and text only serves
deaf blind users, they do not address the needs of other groups such as
people with cognitive impairments. It is indeed true that the situation is
one of creating a site which has multiple formats, as Greg pointed out. The
problem is that removal of some accessibility barriers in one format, and
some in another, can often allow some barriers to remain unnoticed in both
formats. In addition there is the very real possibility of creating a
separate but inferior site for some groups of people, to a much greater
extent than their needs (or state of the technology) dictate. 

The idea of an accessible design is that it can be rendered as text only, but
need not be. Some of the newer features of HTML - the OBJECT element which
allows for a cascade of different types of object, CSS positioning and
formatting which gives the author great power in crafting the rendering of a
website while allowing the user to ensure that their needs are met, are
examples. [* My prediction...*] As XML applications which have been designed
from the ground up to be accessible become more common, along with browsers
which can render them appropriately for users, the idea of a "text-only"
version of a site will seem more and more like an evolutionary dead end which
denies the richest possible experience available to users, and we will see
much more accessible sites which are written to take advantage of the various
methods for user selection of suitable content types.

With regards to a hot-key to switch between versions, which is only
accessible to newer browsers (i assume it is an HTML Accesskey), why not add
a link and use display:none to hide it, for compatibility with older
browsers? (Until Lynx stops being the browser of last resort for people who
are blind, for example. This is just an idle thought - I appreciate that
there is also value in pushing people towards adopting better technology
where possible.)

And to answer Brain's question, the general mission of WAI is to provide
accessibility for people with disabilities. Although there are side benefits
for all kinds of other areas (robots, mobile devices, internationalisation,
etc.) work which is directed at those other areas is beyond the scope of the
current guidelines activities as I understand them.

my 2c worth

Charles McCathieNevile

On Fri, 21 May 1999, Brian Kelly wrote:  
  > I agree that a text-only site would be a disservice to the majority of
  > users, and less accessible to many, if it replaced the fully-formatted
  site,
  > or did not contain all the information, or was not kept up-to-date.
  However,
  > it seems to me that a parallel text-only site adds considerable value when
  > those pitfalls are avoided.
  
  I agree that a text-only representation of a web site shouldn't be
  automatically dismissed.  A paper at last year's WWW conference gave an
  example of providing a text-only version of resources which are intended to
  be indexed by robots.
  
  Maintenance is an issue, but if this is done on-the-fly or by backend batch
  processing (rather than manually - which we know is unlikely to be
  maintained) this shouldn't be a major problem.
  
  Architecturally it's a bit of a fudge, but it may be a useful pragmatic
  solution.
  
  BTW is accessibility for robot software covered by the User-agent
  guidelines?
  
  Brian
  ------------------------------------------------------
  Brian Kelly, UK Web Focus
  UKOLN, University of Bath, BATH, England, BA2 7AY
  Email:  b.kelly@ukoln.ac.uk     URL:    http://www.ukoln.ac.uk/
  Homepage: http://www.ukoln.ac.uk/ukoln/staff/b.kelly.html
  Phone:  01225 323943            FAX:   01225 826838
  

--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 Friday, 21 May 1999 12:09:01 UTC