[69] in bugtraq
Re: Fingerd Summary
daemon@ATHENA.MIT.EDU (Adam Shostack)
Fri Oct 21 17:55:23 1994
From: Adam Shostack <adam@bwh.harvard.edu>
To: gildea@x.org (Stephen Gildea)
Date: Fri, 21 Oct 94 15:20:46 EDT
Cc: bugtraq@crimelab.com
In-Reply-To: <9410211820.AA05498@alex.x.org>; from "Stephen Gildea" at Oct 21, 94 2:20 pm
| Here's another feature we'd like to see: allowing fingering specific
| people, but not a generic finger of a machine, which could be used to
| learn login ids. Which of the fingerds you looked at had that feature?
rfingerd could be made to do that with a little input
filtering. Its perl, so the input filter should be pretty simple--if
anyone writes one, please pass it on to me. No need to reinvent the
wheel. :)
fingerd-1.0 has the ability built in.
sfinger provides it by only allowing certain finger targets,
defined by the existance of a file in its directory (which is the
output of finger.)
Both rfingerd & fingerd-1.0 call /usr/ucb finger. If you
replace that with gnu finger, gnu finger allows you to define
arbitrary actions, including a 'default' target if no argument is
given. If this default target is site info, you get the effect of not
allowing 'generic' fingers.
Also, don't forget to turn off rusers. :)
Adam
