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

RE: The problem with tables

From: Arthur R. Murphy <arthur.murphy@arch.gatech.edu>
Date: Mon, 27 Jul 1998 12:50:32 -0400
Message-Id: <199807271651.MAA23477@crtsun.crt.gatech.edu>
To: David Bolnick <davebo@MICROSOFT.com>, WAI <w3c-wai-ig@w3.org>
W3C folks,

From my perspective, the issue of table use is 
three separate questions: What formats should be 
supported in Mark Up Language (evolving specification); 
How should information be coded into Web pages 
given the current state of HTML; How should 
browsers display information coded as a table.

The long-term solution requires us to be more 
explicit about content versus format.  That is, to 
define the various structures (arrangements of 
information) which will be supported, and to define 
the coding of each one.

The goal should be a recognition that, despite 
increasing bandwidth, there will always exits text-
only browsers, and despite higher density screens, 
there will be a need for speech synthesis of the 
display.  Thus, design and evaluation should 
involve Lynx-type browsers and speech synthesis of 
visual browsers.

Three specific formats are now overloaded into the 
Table Code: Tabular Data, (e.g. name and phone 
number lists), best understood as rows of 
information; columns, best understood as a linear 
structure which wraps from lower left to upper 
right; a Table-of-Contents block, best understood 
as meta-information for the page.

My experience has been that Lynx does a reasonable 
job at resolving all three forms by displaying the 
contents of each cell on a row-by-row basis.  This 
works for tabular data, and is acceptable if the 
column is a single row, and may work for a Table-
of-Contents block.    The shortcomings of this 
technique are not the result of Lynx restrictions, 
but the incompleteness of the full Markup language 
specification combined with the longstanding 
tradition of not using the column / row 

My recommendation is that we agree on a complete 
list of display structures, (e.g. rows of 
information, columns of text, etc.), generate a 
sample set of these coded in current HTML 
techniques. We might try some Style Sheets that 
support text only browsers, Style Sheets that 
support visual browsers for low vision access, 
Style Sheets that support visual browsers with 
synthetic speech.

This effort may inform the specific needs of Web 
designers, and also clarify the range of data 
structures supported for all Internet users.

I have taken the liberty of expanding slightly the 
example of David Bolnick.


Could there be a W3C site, which grows definitions 
and samples?

Arthur R. Murphy,  Research Scientist
Center for Rehabilitation Technology, Georgia Institute of Technology
490 10th Street,  Atlanta, GA 30332-0156
Voice: 404.894.0562  Fax: 404.894.9320  arthur.murphy@arch.gatech.edu
Received on Monday, 27 July 1998 12:49:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:01 UTC