Bad platforms

Making a bad platform work is very difficult. It is generally not impossible,
but requires a lot of technical knowledge and a fair bit of hacking. The
cost/benefit analysis cannot be generalised, but should be done carefully -
sometimes it is appropriate to tweak the system, and sometimes it is
appropriate to throw it out, retrain, and rebuild from scratch.

In respect of a database-generated site the great advantage is that it is
fairly regular, and a proxy which reconfigures the output (which I understand
is the way Betsie works) may be relatively cost-effective to implement.

My personal preference is for systems which allow complete control over the
output - being able to control the final code sent across the wire is the
best way to guarantee that it will be as good as your design ability can make
it.

It is the goal of the Authoring Tool Working Group to produce guidelines for
accessible tools, which produce accessible web content. I would hope that
those guidelines can be used not only by developers of tools who are seeking
to produce a better piece of software, but also by prospective purchasers who
want to evaluate whether a particular piece of software will suit their
needs.

The working drafts of the document are all public, available from
http://www.w3.org/WAI/AU and comment to w3c-wai-au@w3.org is always welcome,
as well as being specifically requested from the IG at certain times.

Charles McCathieNevile

On Fri, 21 May 1999, Richard Caloggero wrote:
[snip]
  My second question is how can one avoid the issue of bad choice of 
  platform. In other words: a designer or accessibility tester is hired to 
  make an existing site accessible, but the site is based on inherently 
  inaccessible technology which cannot be changed due to financial or other 
  reasons.
  Thanx in advance for any suggestions or information.
  
  					Rich

Received on Friday, 21 May 1999 12:43:56 UTC