TABLE and IFRAME

nir dagan (dagan@upf.es)
Mon, 2 Mar 98 6:02:36 MET


From: nir dagan <dagan@upf.es>
Message-Id: <199803020502.GAA23795@sahara.upf.es>
To: www-html@w3.org
Date: Mon, 2 Mar 98 6:02:36 MET
Cc: dagan@upf.es
Subject: TABLE and IFRAME 

Some problems with the Spec of tables and inline frames.

Tables (Chapter 11...)
----------------------

1. TFOOT must come before TBODY. 
 Motivation: to make rendering faster (so far so good)
 Problem: Backward imcompatible with HTML 3.2; the rows of 
 the TFOOT will be rendered above those of the TBODY.

2. Optionality of end tags of TFOOT and THEAD. Why adotp a 
 definition that cannot be written in the DTD, when they 
 could have been made REQUIRED. (this is philosophical, since the spec
 is recommendation now). 

3. The (deprecated) align attribute in TABLE element.
 It seems from the spec (and also of 3.2) that it is 
 like in H1...H6 and P, i.e. like text-align in CSS, rather than
 floating (floating like in <IMG src="foo.gif align="right" alt="Foo">)

 However it seems that most graphical browsers I now of, float tables
 when marked align="left" of align="right".

 Did I get it wrong or did Netscape? (Probably the answer is 
 that tables were implemented before they got in the specs.)


In-line frames (Chapter 16)
---------------------------

1. Why have a longdesc attribute when IFRAME is a container, whose 
content should do the job (like in OBJECT)?

2. Minor idea: add a standby attribute like in OBJECT in future
 specs. Indeed rendreing HTML pages is faster than "heavy" objects, 
 but slowness has no limit strictly above zero.

Best regards,
Nir Dagan.

email: dagan@upf.es
URL: http://www.econ.upf.es/%7Edagan