Re: stupid multiple definitions for the TYPE attribute

Joel N. Weber II (devnull@gnu.ai.mit.edu)
Sun, 7 Sep 1997 21:36:11 -0400 (EDT)


Date: Sun, 7 Sep 1997 21:36:11 -0400 (EDT)
Message-Id: <199709080136.VAA00218@melange.gnu.ai.mit.edu>
From: "Joel N. Weber II" <devnull@gnu.ai.mit.edu>
To: Harald.T.Alvestrand@uninett.no
CC: estephen@emf.net, www-html@w3.org
In-reply-to: <1653.873627569@dale.uninett.no> (Harald.T.Alvestrand@uninett.no)
Subject: Re: stupid multiple definitions for the TYPE attribute

   Does the HTML group think that HTTP should have a header called
   "Content-Style-Type"?

Probably not IMHO.  I think that the content-type header is reasonable,
though.  I do think that the window-target HTTP header (am I remembering the
name correctly?) really violates the layering.

   If not, I regard <META http-equiv="Content-Style-Type" content="text/css">
   used as a way of indicating information about the document that is to be
   used in any context, for documents that may never be transferred over
   any mechanism called HTTP, specifying something that HTTP protocol
   machinery should not need to care about, as a layering violation; the
   "name=" should be used whenever the HTTP header is not an intended
   consequence.

I think I agree with you on this.

A cleaner way would probably be to have a content-type attribute on
every element, and have it inherit (s othat you can set it on the HTML or
BODY tag).  But I suspect that it's too late to change this.