PRE and TABLE - a proposal for changing their relationship

Benjamin Franz (snowhare@netimages.com)
Fri, 8 Sep 1995 07:27:23 -0700 (PDT)


Date: Fri, 8 Sep 1995 07:27:23 -0700 (PDT)
From: Benjamin Franz <snowhare@netimages.com>
To: www-html@w3.org
Subject: PRE and TABLE - a proposal for changing their relationship
Message-Id: <Pine.LNX.3.91.950908070629.6316A-100000@ns.viet.net>

Something that has bothered me for a long time now is the decision to 
not allow <PRE><TABLE>...</TABLE></PRE>. I tried looking in the archives 
but didn't immediately find out why this was done. It appears clear to me 
that in the real world of badly broken content negotiation and
a general unwillingness to maintain two documents whith the sole
difference being the use of HTML 3 TABLE, <PRE><TABLE>...</TABLE></PRE> 
would provide a much more robust transition strategy from HTML 2 to HTML 3.

By specifying a line break above and below <TABLE>...</TABLE> and that 
within the TABLE a wrapper <PRE> </PRE> is ignored, better than 90% of 
all tables I have seen could be easily maintained for both HTML 2 and HTML 
3 - without requiring content negotiation to actually work.

Any arguments about potentially causing older browsers to break with 
illegally included markup in the table is pretty moot - <TABLE>...</TABLE> 
does a pretty good job of breaking old browsers right now. Almost no 
pages now using tables are even comprehensible in a browser that doesn't 
understand <TABLE>...</TABLE> - and most of those that are comprehensible 
are because of using the (currently illegal) method of  
<PRE><TABLE>...</TABLE></PRE>. The cases where a legal HTML 3 table would 
be illegal HTML 2 are far rarer than tables than can not be viewed *at all* 
right now in many browsers.

What compelling arguments are there against changing the allowed 
context of tables to allow <PRE><TABLE>...</TABLE></PRE>?

-- 
Benjamin Franz