W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 1998

Re: Tables - my thoughts

From: Jon Gunderson <jongund@staff.uiuc.edu>
Date: Fri, 24 Jul 1998 10:42:13 -0500
Message-Id: <199807241542.KAA05128@staff2.cso.uiuc.edu>
To: Jason White <jasonw@ariel.ucs.unimelb.EDU.AU>, WAI Markup Guidelines <w3c-wai-gl@w3.org>
JRGs response to Jason White.  
I am not advocating the missuse of tables for positioning.   I am saying if
someone is using a tables for positioning or for data display that they can
label how they are using the table to improve accessibility by a user
agent.  I don't think WAI or the W3C can stem the tide of people using
tables for positioning.  I am sure that if you ask most Web authors if they
thought positioning was a missuse of tables they would say no, since most
authoring books, workshops and authoring tools tell them to do it that way.
 I have concerns about waiting for majority of the HTML authors to switch
to CSS positioning techniques for pages to become more accessible.
Remember that most HTML authors have not even heard of the W3C and most
view HTML as a presentation language, not a structural language.  While
this is unfortunate, it is also reality.  I have done presentations here at
UIUC on using CSS for positioning and the feedback I get in why people will
not use CSS is the incompatibility of the current browsers that support CSS
positioning and that authors want their pages to look how they intended on
older 2.x and 3.x browsers.     

At 11:28 AM 7/24/98 +1000, Jason White wrote:
>The most important consideration here is that the use of tables purely for
>purposes of layout is contrary to the intention of the HTML specification
>and ought to be strongly discouraged in favour of style sheet positioning,
>as soon as the latter is supported by user agents. This is why the
>guidelines have been carefully expressed to make it clear that "layout
>tables" are only permitted on an interim basis, pending the implementation
>of CSS 2 positioning. A table prescribes a relationship between a set of
>header cells and a set of data cells, thereby grouping related data into
>categories. This is the assumption which underlies the attributes that
>were added to HTML 4.0 so as to provide for speech output and better data
>base integration (HEADERS, SCOPE, AXIS, SUMMARY, etc.).
>For this reason I would not regard it as appropriate to offer any further
>support within user agents or otherwise, for the misuse of table
>constructs. Authoring tool developers in particular need to ensure that
>their products generate appropriate CSS positioning rules. I think the
>user agents will support positioning first, with the authoring tools
>following closely behind.
>Perhaps the guidelines should make it clearer that the misuse of tables
>for layout purposes amounts to poor design practice and is problematic
>from the standpoint of accessibility, when considered in relation to
>today's HTML user agents. On the other hand, given the ability to
>manipulate and selectively deactivate style sheets at the client end, the
>proper use of positioning would allow the layout to be adjusted (the text
>would be linearized and formatted according to the HTML elements used).
>Perhaps thiese issues could be more fully addressed in the Techniques
Jon Gunderson, Ph.D., ATP
Coordinator of Assistive Communication and Information Technology
Division of Rehabilitation - Education Services
University of Illinois at Urbana/Champaign
1207 S. Oak Street
Champaign, IL 61820

Voice: 217-244-5870
Fax: 217-333-0248
E-mail: jongund@uiuc.edu
WWW:	http://www.staff.uiuc.edu/~jongund
Received on Friday, 24 July 1998 11:42:49 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:27 UTC