FYI, there is discussion (off IETF lists for now) about using the URI RR type for Kerberos KDC discovery. Presumably that would yield an update to RFC4120. Such an update would have to be on the Standards track, natch. If the URI RR I-D ends up on the Informational track then we'll either: need a downref (feasible but uncomfortable), update the URI RR to be on the Standards track, or give up. Clearly it would be better if the URI RR lands on the Standards track. It will take more work, but it will be worthwhile. If the authors lack the energy for it, perhaps we can move this discussion to SAAG and find volunteers to write and debate the necessary text. Nico --