Passfunctions (was Re: Password security - Another idea)

[Randal L. Schwartz @ Stonehenge]

In article <900 at eta.unix.ETA.COM>, bstrand at woods (Brad Strand) writes:
| Some Disadvantages:
| 
| 1. People might tend to pick simple 'pass-functions' (N = C + 3)
|    because more difficult ones might be either harder to remember, or
|    harder to calculate in one's head.
| 
| 2. Would need to come up with a standard method of representing and/or
|    encrypting the pass-functions for use in the '/etc/passfunc' file.
|    This is probably non-trivial.

Actually, it could be pretty easy.  Just say that /etc/passfunc/merlyn
contains a root-exec-able program that reads an integer on stdin and
writes the result on stdout.  It does not have to read more than one
integer (successive trials would reinvoke the program).

Simple users could put:

#!/bin/awk
{ print $0 + 3 }

while sophisticated users would have a much more complicated thing
(including lookup tables, things depending on time-of-day, etc.)

Actually, /bin/login could be modified to run /etc/passfunc/merlyn
*after* the normal user/password handshaking as an additional security
mechanism for those users that desired it.  The simple users have only
a (simple :-) password, while the sophisticated users would put both a
simple or complex password and an /etc/passfunc routine.

This would then be upward-compatible and user-selectable.  Cool.
-- 
Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095
on contract to BiiN (for now :-), Hillsboro, Oregon, USA.
<merlyn at intelob.intel.com> or ...!tektronix!inteloa[!intelob]!merlyn
HEADER ADDRESS MAY BE UNREPLYABLE if it says merlyn at intelob.biin.com ...
Standard disclaimer: I *am* my employer!