Re: Deprecating FONT and CENTER

James Tauber (jtauber@library.uwa.edu.au)
Sat, 1 Jun 1996 11:45:18 +0800 (WST)


Date: Sat, 1 Jun 1996 11:45:18 +0800 (WST)
From: James Tauber <jtauber@library.uwa.edu.au>
To: "Marcus E. Hennecke" <marcush@crc.ricoh.com>
Cc: www-html@w3.org
Subject: Re: Deprecating FONT and CENTER
In-Reply-To: <199605311850.LAA00802@cougar.crc.ricoh.com>
Message-Id: <Pine.SUN.3.91.960601113119.24986A-100000@docker.library.uwa.edu.au>


On Fri, 31 May 1996, Marcus E. Hennecke wrote:
> I suggest the following:
> 1.  See if there is rough consensus on having a strict version of 3.2
> 2a. If yes, somebody with more experience in writing DTDs needs to
>     figure out what changes are necessary so that FONT and CENTER are
>     not in the strict version. Propose those changes to W3C and we're
>     done.
> 2b. If no, see if there is rough consensus on declaring FONT and/or
>     CENTER deprecated and goto step 3.
> 3.  If it was decided to deprecate FONT and/or CENTER, figure out what
>     changes are necessary and propose them to W3C.
> 
> My personal vote would be for having a strict version without FONT and
> CENTER *and* for deprecating FONT and CENTER (actually I am not sure if
> there is a difference).

If you deprecate FONT and CENTER by putting them in marked sections then 
you can get two DTDs, one with the marked section INCLUDEd and one with 
it IGNOREd. These two DTDs would give you the normal and strict versions 
respectively.

But it occurs to me that this isn't really what deprecation is. Doesn't
deprecation mean "allowable but not recommended"? In which case it isn't 
something that can be defined in the DTD but only in the explanatory 
text.

As I've suggested before, I'd like to see the strict version also do the 
following:

	* Allow only element markup (ie no PCDATA) in DIV.
	* Allow only element markup (ie no PCDATA) in BODY (like HTML 2.0 
	  Strict)

Ultimately I'd also like to see an H or HEADING element that is rendered 
according to the nesting of DIVs. But I would suggest that be in another 
DTD.

So, in other words:

	HTML 3.2 Strict
	^^^^^^^^^^^^^^^
	Same as HTML 3.2 but:

		- no FONT
		- no CENTER
		- no PCDATA in DIV
		- no PCDATA in BODY

	HTML 3.2 Structured
	^^^^^^^^^^^^^^^^^^^
	Same as HTML 3.2 Strict but:

		- H or HEADING element rendered according to DIV depth
		- no Hn elements
	
I think all versions of HTML 3.2 (including normal) should allow a CLASS 
or TYPE attribute on DIV elements.

James K. Tauber / jtauber@library.uwa.edu.au
University CWIS Coordination Officer
The University of Western Australia