W3C home > Mailing lists > Public > www-ws@w3.org > January 2004

first there was HTTP: .... then there was FTP: ... and now FIPA:

From: Tim Finin <finin@cs.umbc.edu>
Date: Fri, 09 Jan 2004 11:43:03 -0500
Message-ID: <3FFEDA17.70800@cs.umbc.edu>
To: www-ws@w3.org, public-sws-ig@w3.org, www-rdf-logic@w3.org

Here's yet another chance to try to figure out how to marry the
world of the web with that of communicating software agents. We
may be better prepared for this now, given the ideas and tools
coming out of the semantic web and web services communities. Tim

-------- Original Message --------
Subject: [Chat] Call for proposals on an infrastructure to exploit the 'fipa' URN
Date: 	Wed, 07 Jan 2004 11:23:35 +0100
From: 	Bellifemine Fabio <Fabio.Bellifemine@TILAB.COM>
To: 	chat@fipa.org, jade-develop@sharon.cselt.it, ...

Dear colleagues,

IETF recently approved the 'fipa' URN, which is now part of the Official
IANA Registry of URN Namespaces (see _http://www.iana.org/assignments/urn-namespaces_),
also published as RFC 3616.

The FIPA Architecture Board, on behalf of FIPA, now calls for proposals
on how to implement an infrastructure for best exploitation of this URN.

In fact, through its specs, FIPA has defined a number of catalogues of
different classes of entities: communicative acts, content languages,
ontologies, interaction protocols, message transport protocols,
encodings, specifications, terms and definitions,  All these catalogues
are spread across the specification documents of FIPA and, in most
cases, they are not available in a format directly usable by software
agents.

The proposals should provide a technical solution for an infrastructure
that meets, at least, the following requirements:

- make available each catalogue and each element of the catalogue both
   to humans and to software (i.e. searchable and understandable by
   software agents)

- make these catalogues simple to maintain and to update by adding new
   entries/terms

   o each catalogue should have a version identifier and should be
     searchable by version/date

   o each entry in the catalogue should have a life-cycle slot (e.g.
     fipa-approved, fipa-pending, proposed by TCx, proposed by
     fipa-member X, ...)

   o the infrastructure must consider at least 4 different types of users:
     # a human (e.g. a sw developer) that needs to
       browse, search a term, understand its definition
     # a software agent that at run-time or at configuration-time
       needs to browse, search a term, and use it
     # a human (e.g. a FIPA member) that proposes a new entry
     # an administrator (e.g. the FIPA FAB) that controls the life-cycle of terms

Proposals should be sent to fab@fipa.org and they should arrive by 29th
of February 2004 at midnight and they should include the following
information, each one being clearly identifiable (e.g. a chapter) in the
proposal:

- author and affiliation
- description of the technical solution
- detailed description of how a human can access each catalogue (search,
   browse)
- detailed description of how a software agent can access each catalogue
   (search, browse) and use the content of the catalogue. Particular
   relevance should be given to a java software module, being java the most
   diffused language for FIPA-compliant agent platforms
- detailed description of how FIPA can maintain each catalogue, i.e. the
   catalogue life-cycle and the life-cycle of an entry in the catalogue
   (e.g. add/modify/delete entries)
- rough estimation of the cost needed in order for FIPA to implement
   such an infrastructure, including possible candidates wishing to apply
   for this job

At the next FIPA meeting, a session will be dedicated to the discussions
of these proposals. Participation to the next meeting for presenting the
proposals is desirable even if it is not strictly required.

Best regards,
FIPA Architecture Board.
Received on Friday, 9 January 2004 11:44:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:45 GMT