[10305] in www-talk@info.cern.ch
Re: Extended URL for frames
daemon@ATHENA.MIT.EDU (Ted Hardie)
Fri Sep 13 18:11:20 1996
From: hardie@merlot.arc.nasa.gov (Ted Hardie)
To: burchard@CS.Princeton.EDU (Paul Burchard)
Date: Fri, 13 Sep 1996 14:56:46 -0700 (PDT)
Cc: davidp@earthlink.net, ilopez@fundesco.es, www-talk@w3.org
In-Reply-To: <199609131728.NAA24220@atomic.CS.Princeton.EDU> from "Paul Burchard" at Sep 13, 96 01:28:25 pm
Reply-To: hardie@nasa.gov
Resent-From: www-talk@w3.org
P. Buchard writes:
> Frames are hierarchical client-side presentation resources that should
> be addressable by URLs. However, these URLs shouldn't necessarily
> bear any relationship to the URLs from which the content was
> retrieved. What I have previously proposed is that Netscape's
In the current scheme of things frames can have content referenced by
a url within a "src =" construction; retrieving the source of
particular frames can thus be as easy as retrieving the URL named in
the "src =" construction. When you want to retrieve the information
on how that source is presented within the context of a particular
document, though, you are currently forced to retrieve the entire
document. His proposal made a relationship between the original
entire document and source for a frame with a URL which referenced
the source for a frame, something like this:
main.html##[][][doc3a.html#here#[][doc3a-2a.html#there#]]
I can certainly see why avoiding this construction is desirable, but
I'm not sure if having a URL which bore no relationship to the
original is that great an idea either. I think a better way would be
to name the frames and framesets so that the URL could be constructed
of the original document and the frame/frameset name.
something like:
http://www.gamelan.com/index.html##frameset=corner
Ted Hardie
