Re: URN resolvers

Mitra (mitra@path.net)
Fri, 29 Oct 1993 20:05:55 GMT

From: mitra@path.net (Mitra)
Subject: Re: URN resolvers
Date: Fri, 29 Oct 1993 20:05:55 GMT
Message-Id: <CFoD5w.6Gs@pandora.sf.ca.us>

clw@merid.edu sez
: If you look at the history of internet
: protcols, none are truly ubiquitous. But it all still works.

Chris - how about DNS. URN->URL resolution is closest to DNS than to
anything else. I cant imagine trying to build a dozen different
protocols, into my client just so I could find different machines.

The key here is that the protocol variations come in retrieving the
document. I havent heard anyone argue that we NEED more than one
protocol for looking up URNs and getting back URLs. Even from competing
protocol camps (HTTP etc) I havent heard anyone argue for competing
lookup protocols.

Where we will have variation is in the kinds of meta-information stored,
the speed of the services, comprehensiveness, and in the kinds of things
we want to do but dont NEED for resolution. For example I could see
whois++ going down the route of allowing searches on meta-information,
whereas another service might specialize in being able to retrieve and
format-convert any document it can point you at.

- Mitra