W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 2004

RE: Layout versus data tables proposal for null summary attribute

From: Phill Jenkins <pjenkins@us.ibm.com>
Date: Fri, 27 Aug 2004 11:56:23 -0500
To: <w3c-wai-ig@w3.org>
Message-ID: <OFAC2DB0C9.2CD239EA-ON86256EFD.005A834E-86256EFD.005D0EB6@us.ibm.com>
John Foliot wrote:
> ... need for more measurable mechanical checks, and 
> this to me is a perfect example of how we can possibly do this.
> ...
> 3) table foo has no summary attribute.  In all 3 instances, 
> a final decision MUST be made by the developer, QA person

Exactly, one of the benefits I poorly tried to identify.  And remember 
that the next time you run the automated checker, you may want to be able 
to ignore option 2 - all those tables that have a null summary.  In other 
words, you want to be able to leave a hint in your code so that the 
checker knows you've already examined that table and determined it is a 
layout table that doesn't need any more corrections.  The problem with 
Jesper's suggestion - to remove the summary attribute altogether - is that 
there is not difference between tables that have been checked and 
determined to be valid layout tables not needing a summary attribute and 
those data tables needing a summary attribute. 

There are even more sophisticated algorithms in the checkers today that 
are guessing wrong and inundating the author with tons of false errors. 
The 3 option approach; namely "data", "layout", and "undetermined" works 
well with the summary="this is a data table", summary="null", and the 
missing summary attribute indicating "undetermined". 

> ... any programmatic element or attribute which speeds the "human
> intervention" aspect of accessibility checking cannot be all bad in my
> books...
> Another 2 cents worth

Excellent point.  Now, if we could only get paid 2 cents for every table 
that was incorrectly flagged by automated checkers.

p.s. I've been told by a few that the monospaced text that is quoted in my 
e-mail (rich text not HTML) appears in some e-mail clients as "very small 
font".  My e-mail client Lotus Notes sends it as "default monospaced 10", 
Please check to see if your e-mail client's default font is set to some 
undesirably small font.  [by the way, this text is being sent as default 
san serif font 10, the quoted text above was sent as default monospaced 12 

Phill Jenkins
IBM Worldwide Accessibility Center
Received on Friday, 27 August 2004 16:57:01 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:29 UTC