[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: paper tapes: B-5500 CANDE Extended ALGOL program

> Better
>work on timesharing/security was being done elsewhere DEC's PDP-6 & PDP-8
>vitual machines and the Multics project.

When I first started working on OS security in 1984, Multics was the
standard by which EVERYTHING was measured.  For Secure XENIX we even
developed mathematical models that were based on work originally done for
Multics.  There were some people during the late 70's and early 80's that
were crazy enough to try to mathematically prove OS security.  Some folks at
Ford Aerospace (ironically) developed a system called PSOS (the Provably
Secure OS).  I don't think they ever completed the project.

My first manager at IBM had worked at Burroughs for a while, and he often
used the B6700/B5500 as an example of how NOT to build a secure OS.  I used
to argue to him that security could be built into the language instead of
the hardware if done right, and I used to argue that RSTS V4 on the 11/20
was reasonably secure without hardware memory protection, but he wouldn't
buy it.  It's interesting that Sun is now building security into the
language with Java.  Everything comes around again.