Message-Id: <9310282017.AA11811@merit.edu>
To: mitra@path.net (Mitra)
From: "Fred Swartz" <fred.swartz@umich.edu>
Subject: Re: Thumbnail Sketch of the UR* Arena
In-Reply-To: Your message of "Wed, 27 Oct 1993 20:44:40 PST."
<9310272044.aa04957@pandora.sf.ca.us>
Date: Thu, 28 Oct 1993 16:17:47 -0400
> From: mitra@path.net (Mitra)
> To: peterd@bunyip.com, marca@ncsa.uiuc.edu
> On Oct 27, 3:47pm, Peter Deutsch wrote:
>} Just a sidebar - do we actually _need_ a "URN->URL
>} protocol"? Certainly, we need a format, but do we need a
>} query protocol and all that it implies? Seems that we can
>} content ourselves with the equivalent of a MIME type for
>} formatting responses and we're done.
>
>No way Peter, if a URN is going to be inside a document then, as a
>client designer, I need to know how to resolve it, that measn throwing
>it at a resolution service and gettign back a result. That involves a
>protocol. The protocol could be really simple, but its a protocol none
>the less. Its totally useless if their are some URN->URL resolvers
>talking whois++ and others talking modified DNS and others talking some
>variant of HTTP.
Peter, I'm amazed to that you even raise the possibility of complete
UNR chaos. I would like my URN server to be compatible with Rob's,
Mitra's, yours, etc. Otherwise, as Mitra says, URNs are totally
useless.
-- Fred (fred.swartz@merit.edu)