🚀 go-pugleaf

[A crossposting from comp.admin.policy - note the Followup]

I have criticized a statement in Princeton's user policy that
prohibits locking of terminals to enable other users to get access to
that machine.

In <BuBxzz.A21@rice.edu> jaw@owlnet.rice.edu writes:

>  In article <CKD.92Sep9143724@loiosh.eff.org>, ckd@eff.org (Christopher Davis) writes:
>  |>  Olaf> This means: 'If you must go to the restroom you HAVE to leave the
>  |>  Olaf> terminal open, so that other users can terminate your session if
>  |>  Olaf> they need the machine.' This is IMO simply unacceptable. Other
>  |>  Olaf> policies would REQUIRE to lock the terminal to prevent tampering
>  |>  Olaf> with your account.
>  |> 
>  |> MIT's Athena workstation policy is that you can lock it for no more than
>  |> 20 minutes before someone can feel free to reboot it or otherwise
>  |> "forcibly unlock" it.  The "standard" screenlocker has a displayed timer

Rebooting is not the way of choice, I think...

>  |> of "how long I've been locked".
>  |> 
>...  
>  At Rice, we have an X windows lockscreen program with a twenty minute
>  timer.  After twenty minutes, the lockscreen itself logs the user out.
>  The logout is pretty clean.  The only thing we haven't solved yet is
>  how to get the user's .logout file to be read on logout by the
>  lockscreen program.

What about the following?

A locking program that 
- displays the time it is running,
- after N minutes provides a button which will log the user out (with
N preferrably to be specified in a configuration file),
- perhaps communicates with other workstations in the cluster, to
enable the logout only when no other workstation is idle.

I leave the gory details of the implementation to the people who know
X (that excludes me, unfortunately ;-)

(Just an idea: At Karlsruhe we have a login process running under X
(whats the name? xdm? I'm not sure) that executes an ~/.xsession
script and logs the user out cleanly after the end of this script
(which may end in 'exec twm'); could it be implemented that the pid of
this process gets stored in a file and it simply killed for logout?
That should clean up everything.)
MfG,
        Olaf
-- 
    o     Olaf Titz - comp.sc.student - univ of karlsruhe - germany
 _ /<_    s_titz@iravcl.ira.uka.de - uknf@dkauni2.bitnet - praetorius@irc
(_)>(_)   +49-721-60439 - did i forget something?
Der gr"une Punkt ist halt genauso sinnvoll wie ne T"UV-Plakette
 auf nem Schrotthaufen. - Thomas Volkmar Worm