Re: <BODY>, <FRAMESET> and <NOFRAMES> in MSIE 3.0 and HTMLPRO

Peter Flynn (pflynn@curia.ucc.ie)
27 Jan 1997 22:40:40 +0000 (GMT)


Date: 27 Jan 1997 22:40:40 +0000 (GMT)
From: Peter Flynn <pflynn@curia.ucc.ie>
Subject: Re: <BODY>, <FRAMESET> and <NOFRAMES> in MSIE 3.0 and HTMLPRO
In-reply-to: <s2e7952b.012@wposmtp.nps.navy.mil> (message from Charles Peyton
To: CTaylor@wposmtp.nps.navy.mil
Cc: www-html@www10.w3.org
Message-id: <199701272240.WAA12586@curia.ucc.ie>

   Using the Corel Visual DTD thingie, I've been looking at 

I'm not familiar with that...do you have a name and details?

   This is the HTML Pro model of frames:
   <HTML>
	   <FRAMESET>
		   <FRAME>
	   </FRAMESET>
	   <NOFRAMES>
		   <BODY> (%body.content)
		   </BODY>
	   </NOFRAMES>
   </HTML>

   Notice that <NOFRAMES> contains <BODY> and is at an equal 
   depth as <FRAMESET>

That's correct. I added the BODY, but it could be made optional.
The concept of NOFRAMES _following_ FRAMESET is as Netscape originally
defined it (inasmuch as they "defined" anything at all).

   The IE 3.0 DTD seems to go this way:
   <HTML>
	   <FRAMESET>
		   <FRAME>
		   <NOFRAMES>(%body.content)
		   </NOFRAMES>
	   </FRAMESET>
   </HTML>

This is utterly nonsensical, as the objective of NOFRAMES is for it to
be a separate part of the document, outside any framing material. I
think MS has made a glaring error here by grossly misunderstanding the
nature of NOFRAMES. Either that, or they are just being deliberately
obtuse and trying to be different from Netscape.

   Furthermore, if one is trying to set an example for his 
   organization by writing validating web pages, should one 
   lean toward the MSIE model or the HTML Pro model?  In 
   other words, what's the Right Thing (other than forget 
   about frames altogether.)

The required course for HTML Pro is to allow you to choose, so I will
vary the model to permit both ways...that much is easy :-)  Thanks for
finding this one, I missed their way of nesting it.

///Peter