Can we really deprecate tables?

Hi
    I attended the WAI meeting at the RNIB in Peterborough last week.  I had
lots of questions which various WAI wg chairs suggested I raise on the
mailing list.  I hope I'm not asking a FAQ (I had a quick scan of the
archive and couldn't find anything) and that I'm using the correct list.

I'm asking my questions in my role as an adviser to UK Universities on web
issues.

I've been promoting the use of style sheets for some time.  Architecturally
they're clearly the way forward.  In a number of my HTML files I use CSS for
simple things such as indenting the left margin (to make use of white space)
or outdenting the right margin (so that actions in minutes stand out) and
drawing boxes around paragraphs to emphasise decisions.  Although  I have
had no problems using Internet Explorer my colleagues who use Netscape have
lots of problems with paragraphs not being printed correctly if at all.

These are the type of simple but valuable features which people who start to
use CSS will begin with.  The printing problems with these simple features
on one of the most widely used browsers will surely put people off CSS - and
potentially put people of the GL guidelines.

I had hoped at the WAI meeting at the RNIB to hear of some magic behind the
scenes which would overcome this problem (i.e. an Apache module which would
implement some form of CSS content negotiation to strip out troublesome
CSS).  However I did not hear any encouraging news about things such as
transparent content / feature negotiation.

Instead of a technological fix the GL Guidelines see the solution to
backwards compatibility problems to be the (rapid?) demise of old browsers.
Various others have made similar predictions (XML, XSL, RDF, ... will be so
good that everyone will upgrade just as everyone through away their Gopher
clients when they first saw Mosaic).  I'm rather dubious about this.  Jakob
Neilson in his March 22 Alertbox column described "The Increasing
Conservatism of Web
Users" (see http://www.useit.com/alertbox/980322.html)

He argued that user's will only upgrade their browsers very slowly and that
new technologies need to be deployed by making the servers smarter.

The W3C Core Style Gallery seems to be doing this by using browser-sniffing
to send "safe" CSS.

To sum up.  I'm worried that the premise of the GL guidelines and the
culture associated with it (CSS good, tables bad), while architecturally
sound, is not currently deployable and attempting to do this could result in
a backlash.

I hope someone can convince me that I'm wrong!

Thanks

Brian Kelly

------------------------------------------------------
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

Received on Tuesday, 4 August 1998 10:05:25 UTC