Just a reminder:

Nearly everyone on these lists has looked at identd and decided it
does nothhing for *them*. But there is a nonzero population of NetBSD
users, in sites where the {network,system} administrators have looked
at identd, decided that it *is* useful in their environment,
and run it (and expect it to be running).

Even if it's just enough to deter a significant fraction of the
undergrads who start forging email.  Stanford email servers use (or
used) identd pervasively.  The extra latency for emial, sent from
SMTP client hosts not running an identd, is quite noticeable.

(Yes, Stanford does have networks with physically-secured,
network-accesible, multi-user machines, and where most/many
single-user clients can be pinpointed to a specific physical location
and responsible human.)

Identd does have uses.  They may not be at the forefront of our
collective minds. They may not be what strong proponents of identd
claim; but they're quite real. I won't argue with the people who do
decide to deploy identd, and expect it from their clients, unless I
have to sleep in their shoes (and carry their pagers).