Re: ROWSPAN issues with TFOOT

"G. James Berigan" wrote:
> 
> I've been trying to elicit comment about this issue in both www-html@w3.org
> and comp.infosystems.www.authoring.html, but no one seems to be willing to
> comment on it.  I hope you can comment on it or make some kind of official
> stand on the issue as it impacts a program I'm developing as well as the
> behavior of MSIE and the development of Mozilla.[1]

I'm sorry that your questions haven't met with replies earlier.
 
> The issue where TFOOT must precede TBODY has caused some problems with
> parsing ROWSPAN attributes.  Parsers today appear to parse the entire table
> and then move the TFOOT to the end of the table.  This becomes a problem
> when you have a ROWSPAN in a THEAD that extends beyond the bottom of the TR
> rows in the THEAD and ROWSPANs in the TFOOT that extend beyond the bottom
> of the TR rows in the TFOOT.  Current browser behavior has the spans past
> THEAD rows spanning into TFOOT rows and into TBODY rows, as well as spans
> beyond TFOOT rows into TBODY rows spanning into TBODY.  Then, when the
> TFOOT is moved to the end of the table, you have a mess.
> 
> I can only see one hard solution to this problem:  It needs to be stated
> that ROWSPANs that extend beyond rowgroup boundaries are illegal markup and
> that browsers must either stop rowspanning at such boundaries or extend the
> number of rows in the rowgroup to account for the excessive spanning and
> not span into other rowgroups below (or disjointedly above).  Either change
> would be in line with the new behavior of ROWSPAN=0 in HTML 4.01.

It seems to me obvious that if you specify a rowspan of greater than
one on a cell in the last row, that this is an error. (This is not
stated explicitly in section 11.2.6.1 [1], which only talks about
overlapping cells). It would also seem to me that this error condition
would also apply to the last row of any row group. The number of
rows in a table is given by the number of TR elements; refer to
section 11.2.1 ("Note that the number of rows in a table is equal to the
number of TR elements contained by the TABLE element."). Therefore,
I would also propose that it's an error to specify a rowspan > 1 in the
last
row of any row group.

 - Ian

[1]
http://www.w3.org/TR/1999/REC-html401-19991224/struct/tables.html#h-11.2.6.1
[2]
http://www.w3.org/TR/1999/REC-html401-19991224/struct/tables.html#h-11.2.1
> The issues can become extremely hairy when the page then attempts to hide
> rowgroups with CSS and ECMAscript, including ROWSPANning across TBODY
> boundaries, which is why I think this restriction needs to be applied to
> all rowgroups.  The CSS/ECMAscript issue may also be a problem for
> COLSPANning across COLGROUPs, but is not as dire an issue.
> 
> The sections are (row groups) 11.2.3, (colgroups) 11.2.4.1, and (spanning
> in general) 11.2.6.1.
> 
> [1] See <http://www.war-of-the-worlds.org/html/span-vs-group.html>

-- 
Ian Jacobs (jacobs@w3.org)   http://www.w3.org/People/Jacobs
Tel:                         +1 831 429-8586
Cell:                        +1 917 450-8783

Received on Friday, 24 March 2000 15:35:29 UTC