Minutes for KRB-WG at interim-2012-krb-wg-1
minutes-interim-2012-krb-wg-1-1

Attending:
  Greg Hudson
  Sam Hartman
  Jeffrey Hutzelman
  Simo Sorce
  Nico Williams
  Stephen Farrell
  Tom Yu
  Nalin Dahyabhai
  Thomas Hardjono
  Zhanna Tsitkova

Sam: this is an IETF meeting, subject to IETF property rules
  Participants should read / be familiar with
  http://www.ietf.org/about/note-well.html

Agenda refresher:
  1. purpose of the PAD and CAMMAC drafts (scoping)
  2. where it will fit(?) & quick review of how authdata works in general
  3. is this 1 draft or 2
  4. how about other types of data in authdata
  5. type number registration procedures
  have received an offline request to discuss CAMMAC before PAD

1. Purpose of the PAD and CAMMAC drafts (scoping)
 - Consensus on call is that the purpose of the PAD draft is to provide
   authorization attributes in support of authorization and account
   provisioning for access to application services on POSIX systems.
 - Consensus on call is that the primary purpose of the CAMMAC draft is
   to replace KDCIssued.

2. Where it will fit(?) & quick review of how authdata works in general
 - Agenda item 2 deemed not necessary.

3. Is this one draft or two?
 - Consensus is in favor of the split of the PAD and CAMMAC drafts.

4. Open issues on the CAMMAC draft:
 - Consensus is that public key signature (pubkey-signature) need not be
   specified at this time.
 - Consensus is that being able to add new types of authentication to
   the structure is worth discussion.
 - Discussion of the session id / AD-ID-ANCHOR will be moved to the
   list.
 - Partial list of concerns (discussion will move to list)
  - KDCissued uses the session key of the ticket, but the draft uses the
    service's long-term key
  - CAMMAC includes Checksum elements, but doesn't include kvno or
    enctype, which are probably needed.

4. Open issues on the PAD draft:
 - Containing arbitrary SAML attributes in a PAD is out of scope for
   this work, but specifying SAML authz data is in-scope for the WG.
 - Consensus is that the short name field will only be used to specify a
   string which can be combined with other identifiers to make them
   unique if they are not otherwise, validation discussion will be moved
   to the list.
 - Issue of central home directories and PAD-Posix-Homedir will be taken
   to the list.