In some mail I received from Mike Long, sie wrote
> 
> >From: Darren Reed <darrenr@cyber.com.au>
> >Date: Thu, 26 Jun 1997 17:48:28 +1000 (EST)
> 
> >> PR 1323 highlights that inetd(8) and other programs are run from
> >> /etc/rc before securelevel is raised. Therefore, it is possible for
> >> users to get into a system before securelevel is raised.
> >> 
> >> This is a complicated issue, and probably deserves its own thread.
> >> No idea on how to solve it.
> >
> >How about adding a loop to inetd that checks if the securelevel is 0
> >and if so, waits until it becomes > 0 ?  This is based on the theory
> >that securelevel is only 0 during booting and the eventual multiuser
> >value for securelevel > 0 whilst allowing it to start if secureleve is -1.
> 
> This won't work.  If inetd dies after the system has gone multiuser
> and you restart it, the restarted inetd will never accept connections.

so I missed explaining a case where inetd should not wait if securelevel
is > 0 (thanks for pointing it out).

Darren