So, you presented an alternate proposal that violated one of my
constraints without describing what it offers that my proposal did not
offer.

In particular, you propose changing the initiator.  I don't like
changing the initiator to support this use case because it's possible
for an initiator to mis-implement the new functionality or if it is
optional to choose not to implement it at all.  Then you get initiators
that work against some environments but not against high-infrastructure
environments, which tend to be sensitive to working with as many
initiators as possible.

Also, I don't think the set of RP-side intermediates is the initiator's
business at all.  We don't tell the initiator what AAA proxies we use,
so why should the initiator be involved in this?

Finally, host-to-realm in clients has generally been regarded as a bad
idea in Kerberos.  If we had it to do over I would not support
host-to-realm functionality in the clients.  I don't support having that
functionality in clients in a new protocol.
_______________________________________________
abfab mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/abfab

Reply via email to