(lack of) security of client workstations

[Paul Gluckauf Haahr]

Karl Kleinpaste writes:
> > All the software security features in the world won't stop me from
> > hitting L1-A and twiddling memory from the PROM monitor. 

rbj at dsys.icst.nbs.gov writes:
> Who says your abort sequence has to be L1-A? Read man 5 kbd. I haven't
> tried it, but TFM indicates that the two key sequence can be changed.  It
> would take an awful lot of pounding to discover the new sequence.

by the same token, what's to stop someone from opening up /dev/kbd and
setting the sequence back to l1-a?  and chmod 600 /dev/kbd is a bad idea
because suntools, x, etc, would have to be made suid.

what is needed is some way to disable the continue command.

does anybody remember when you needed a skate key to get into console mode
on a pdp-11 or vax?  as i remember, all the keys were the same, but it was
some protection.

-- 
paul haahr
princeton!haahr   haahr at princeton.edu   haahr at pucc.bitnet