W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2003

Re: User agent support of SUMMARY attribute in tables

From: David Poehlman <poehlman1@comcast.net>
Date: Wed, 22 Jan 2003 15:29:17 -0500
To: Phill Jenkins <pjenkins@us.ibm.com>, w3c-wai-ig@w3.org
Message-id: <005701c2c254$f03ee3d0$6501a8c0@handsontech>

the problem with the stake holder list, is that some of them can change and
some of them cannot.  I suggest the checkers become more sophistocated.  I
do not know how this is done, but it seems to me that those of us who rely
on accurate rendering are being held hostage by the checkers in this case.
I disagree that summary should not be read automatically.  Where appropriate
and it is difficult to say where is appropriate when there are so many
inappropriate ways to use summary apparently, If I hear a data table with
the automatic reading of the summary, it makes perfect sense.  It even makes
more sense that not hearing it makes and not hearing it might imply that it
is not there so automatic reading of summary is in my estimation what should
be allowable but at the same time it should not be used as a stick to say
well, it should not be read automatically.
----- Original Message -----
From: "Phill Jenkins" <pjenkins@us.ibm.com>
To: <w3c-wai-ig@w3.org>
Sent: Wednesday, January 22, 2003 3:06 PM
Subject: RE: User agent support of SUMMARY attribute in tables

>Which would be preferable; no summary attribute, or a null summary

The null attribute and the missing attribute are hard to tell apart except
when in source reading mode.  I would recommend a reserved keyword such as
"layout" in the summary.  The screen readers can be programmed to ignore
"layout" just as easily as "quote quote" and the fact that the source still
has the "layout" keyword is useful to the author and source code reader.

I think it is important that we remember the requirements of all the stake
holders; including
      screen reader users
      sighted users
      checkers & repair tools

We could also lobby for a new semantic tag, such as <layoutable>, but
getting that added to XHTML would never happen with the existence of CSS,
and would still break in older browsers and screen readers.  Giving up on
using tables and just using CSS for positioning is also a problem for older
browser users.  So I still come back to my suggested convention of the
keyword approach <table summary="layout" yada yada yada..

Phill Jenkins,  IBM Research Division - Accessibility Center
Received on Wednesday, 22 January 2003 15:30:28 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:36:13 UTC