[165] in Zephyr_Comments
Re: Why Kerberos tickets?
daemon@ATHENA.MIT.EDU (dkk@ATHENA.MIT.EDU)
Thu Feb 23 19:51:15 1989
From: <dkk@ATHENA.MIT.EDU>
Date: Thu, 23 Feb 89 19:49:35 EST
To: <bjaspan@ATHENA.MIT.EDU>
Cc: zephyr-comments@ATHENA.MIT.EDU, watchmakers@ATHENA.MIT.EDU
> f dzoller@eagle.wesleyan.edu will tell me if my friend David is
> currently logged in, and Wesleyan does not require anything like
> Kerberos tickets for the service. If Athena requires that anyone
> on the Internet must have tickets to determine if I am logged in, that
> is (gasp!) fascist; more so than other institutions.
> Barr3y
I've had this argument with Jerry before. His point was, just because
Unix "finger" gives certain information to other users doesn't mean
that it *should*. If "finger" breaks MIT's privacy rules, he argued,
fingerd shouldn't necessarily be taken off the packs, but we also
shouldn't make it even more of a privacy problem by adding that
functionality ... uhh, I mean bug ... to zephyr.
I see Jerry's point, thought I don't agree with the tickets-only
policy of zephyr. I object to being unable to make myself visible to
a non-athena-user without sending out a packet (a write message, for
example) every time I log in, and periodically thereafter, to *every*
person I want to be visible to!
I can just see it now, a "fingerable" daemon which pipes an
announcement of login to every user in my .announce file when I start
it up and every 10 minutes thereafter... Or better yet, just modify
the code REAGAN.AI.MIT.EDU runs to log users, so that it accepts
writes from athena machines...
-Dave
