NOTE: This charter is a snapshot of that in effect at the time of the 38th IETF Meeting in Memphis, Tennessee. It may now be out-of-date.
Patrik Faltstrom <paf@swip.net>
Roland Hedberg <Roland.Hedberg@umdac.umu.se>
Applications Area Director(s):
Keith Moore <moore+iesg@cs.utk.edu>
Harald Alvestrand <Harald.T.Alvestrand@uninett.no>
Harald Alvestrand <Harald.T.Alvestrand@uninett.no>
General Discussion:find@bunyip.com
To Subscribe: majordomo@bunyip.com
In Body: in body: subscribe find
Archive: ftp://ftp.bunyip.com/pub/mailing-lists/find
On the Internet, several more or less localized directory services have evolved over the last couple of years. Also two global directory services have been deployed, X.500 and Whois++. To be able to find something or someone, one needs to know what service to use, and what server to query.
One step towards the solution of this problem is to define one and only one common indexing protocol which all directory services can use when passing indexing information. When a user queries one server it should be possible for that user to get a referral to another server and even another service, if the two servers have exchanged index information.
For this to work, one common protocol must be developed. The idea is to expand on the Centroid ideas used by Whois++, to allow it to be used for other services than Whois++. At the very least, a localized service should be able to be polled by an indexing server using the Common Indexing Protocol (CIP). To be specific, three specifications are to be presented:
· An interface spec, where one says how you present a query and what the referrals you get back look like · A server interface spec, where one says that the CIP will be able to include information presented in this format · An engine spec, which specifies that this is how one support the functionality using Centroids a la Whois++.
The task for this working group is to create the Common Indexing Protocol so it is (1) usable for other distributed directory services such as X.500, (2) allows the use of non-distributed directory services such as CCSO in the distributed directory service, and (3) addresses needs such as replication to make the protocol itself more stable.
Just because the Common Indexing Protocol is already in use by Whois++, but not published, the first task of this group is to publish version 1 of the Common Indexing Protocol as is. After that, the protocol must be extended according to the specification below. This will result in version 2 of the protocol.
Other topics to be addressed potentially include:
· Incremental updates of indices · Support for the UTF-FSS encoding of UNICODE · Guidelines for building an effective mesh of indexing servers · Administrative protocols and procedures such as server registration · Security between directory services
The working group will work in very close cooperation with the working groups ASID and IDS in the IETF.
The working group will use the following Internet-Drafts as input:
Architecture of the Whois++ Index Service, Chris Weider <draft-ietf-wnils-whois-03.txt How to interact with a Whois++ mesh, Patrik Faltstrom <draft-ietf-wnils-whois-mesh-01.txtGoals and Milestones:
· The Common Indexing Protocol (CIP)
· A Tagged Index Object for use in the Common Indexing Protocol
· A CIP-based Centroid Exchange for LDAP
· CIP Index Object Format for SOIF Objects