URC defns

Daniel LaLiberte (liberte@ncsa.uiuc.edu)
Fri, 19 May 95 08:56:54 CDT

Date: Fri, 19 May 95 08:56:54 CDT
From: liberte@ncsa.uiuc.edu (Daniel LaLiberte)
Message-Id: <9505191356.AA18314@void.ncsa.uiuc.edu>
To: Renato Iannella <ren@dstc.edu.au>
Cc: uri@bunyip.com
Subject: URC defns
In-Reply-To: <Pine.OSF.3.91.950519120240.11013M-100000@foxtail.dstc.edu.au>

Renato Iannella writes:
 > Question - will there be a series of URC syntaxes?

Multiple URC syntaxes are very likely to be formalized and continue to
exist.  There are already several metadata syntaxes in widespread use
that we will have to learn to live with in some way.  Furthermore,
there are multiple incompatible metadata semantics with much more
difficult problems.

 > If so, will this be confusing for the browser to decide,
 > since the URN will point to an unknown number of possible
 > URC specifications? - Or does the URC identify itself (typeing)
 > so the the user/browser can then decide to handle it or just 
 > return the URC spec.

The URC ought to identify itself using MIME types, at least when sent
via HTTP.  The MIME type might be something like "urc/<urc-syntax>",
where <urc-syntax> might be urc0, sgml, tei, iafa, rfc822, rfc1357,
soif, prdm, usmarc, grs-1, sutrs, asn.1, dif, etc.  Presummably, each
format would have a way to indicate the type of semantic content if

 > Question - is the URA proposal just another URC?

I'm fuzzy on this one, but I believe URAs are targeted at a
representation for code as opposed to metadata.  The initial URA
proposal was perhaps geared to information discovery.

 > Question - we would like to create a URC spec for query-able
 > information sources - should it be a URC or should we define a
 > URQ (that is specific for information sources that provide
 > a query interface - as opposed to a "traditional" URC that seems
 > to favour static documents)

I would think that a general enough URC format ought to be able to
describe collections of resources and/or the interface for interacting
with them, such as a query interface.  The boundary between
description of interface and code is fairly blurry, though, so I
suspect we need to describe interfaces to metadata as well as (or
perhaps instead of) particular data representations.

We need to repeatedly remind ourselves that "resources" may be static
or dynamic objects or services.

Daniel LaLiberte (liberte@ncsa.uiuc.edu)
National Center for Supercomputing Applications