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

Re: HTML Table Markup

From: Phill Jenkins <pjenkins@us.ibm.com>
Date: Mon, 13 Sep 2004 14:34:44 -0500
To: w3c-wai-ig@w3.org
Message-ID: <OF663375E8.BA627E72-ON86256F0E.0065F898-86256F0E.006B8D11@us.ibm.com>
On 13 Sep, Tina Holmboe wrote:

> And the difference between "We've added a null summary attribute
> 'cause this is a layout table", "We've added a null summary attribute
> 'cause we don't understand how to mark up a data table but don't wanna
> get caught during an audit" and "We've added the summary for this data
> table but not filled it in yet" ?

I suppose it is the same difference between as for the alt attribute:
a. "We've added a null alt attribute 'cause this is a spacer image", 
b. "We've added a null alt attribute 'cause we don't understand how to 
mark up an image but don't wanna get caught during an audit", and 
c. "We've added the alt for this image but not filled it in yet".

It seems like you and a couple others are saying that because someone 
might misuse it we shouldn't allow it at all?

> I'd oppose that. A table lacking TH should be flagged as an error, as
> the headers are needed to make clear the relationship between the rows
>  and the columns.

I guess I got confused with the previous phrase "potential misuse" and 
your current phrase "should be flagged as an error" - I guess there is no 
more potential - it's just an error, period.  However, I too subscribe to 
the notion that all true tables (aka data tables) should have TH headers. 
It's just that there are many tables coded in web pages today that are not 
true data tables. 

> A layout table was always an oxymoron, and - I'll dare claim - never
> needed. Let's not keep this illusion up.

Well, I wouldn't choose "always" and "never".  HTML tables (HTML 2 Nov 
1995) did come out a few years before style sheet layout (CSS 2 April 
1998) was published and implemented even later.  Although I'm not trying 
to "keep the illusions up", I am trying to find a transition technique for 
all those layout tables that need to be changed to CSS styling. 

It's not as simple as declaring that all layout tables must now be 
changed.  With the null summary technique I can get a count of how many my 
site may have, which helps determine resources needed to convert them 
correctly to CSS.

Phill Jenkins
Received on Monday, 13 September 2004 19:35:21 UTC

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