On Sat, Aug 18, 2007 at 09:22:18AM +0200, Petter Reinholdtsen wrote: > [Martin J. Carter] > > But then again, I'm aware changing what runlevel reports may break > > other things. Feel free to reassign this to portmap. > > I suspect this is related to the fact that the string returned from > /bin/runlevel get its value from /var/run/utmp, and it is updated when > init switches runlevel and is not quite correct the entire rcS.d run > because it is read-only when the rcS.d starts. No idea how to avoid > it.
Local context: /usr/local/bin is on our site NFS-mounted from a central server to all client systems, the Vulcan nerve pinch is disabled, and the default runlevel is left at 2. An inadvertent ^D or "exit" in single-user mode on a client now means having to push the hard-reset button. I've now RTFM on init(8) under etch: not only is /bin/runlevel performing as advertised, but I've also been stepping into the blockhole mentioned in the new paragraph in the WARNINGS section. The only way round all this I can see would be to special-case S -> 2 as (S then 1) -> 2 in /etc/init.d/rc, but that way madness lies. More later, as and when I get time to investigate further. -- Dr Martin J Carter Computer System Administrator Astrophysics, University of Oxford x 73291 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]