> Interesting -- DOS format magtape. 6.3 filenames...
> Some of the command files on that tape look like RSX ones, and some
> are RSTS ones. Makes sense; compilers and things like that built for
> RSX generally run just fine on RSTS (in RSX emulation).
Which reminds me... students at Groningen University wrote a multi-tasking
system for the PDP that was loosely based on EMAS (big mainframe system
from Edinburgh, sort of like Multics but better). I remember one feature
of the OS was that it used an RT-11 format disk structure for raw storage
even though the user view of files was quite different. For example
I think RT-11 had version numbers (in the style of VMS), but GUTS
showed only the top-level file, and had a "pop" command which deleted
the most recent file and made the version below it visible.
The source is lost (though we do have a disk pack which *may* contain
the OS, if it is still readable) but a scan of the three-book report
is available online:
http://history.dcs.ed.ac.uk/archive/scans/guts/ -
Orange is the users guide; Red - system design; Yellow - selected
source listings.
Edinburgh was also fairly big on PDP11's, having written a few operating
systems for it. The british networking operations ran on PDP11's
for years, on an O/S called DEIMOS. (Binary disk images online
at
http://history.dcs.ed.ac.uk/archive/os/deimos/ ; sources also
at
http://history.dcs.ed.ac.uk/archive/os/emas/emas2/deimos/ - the
binaries ought to work in an emulator; I haven't tried them myself)
Deimos was also used for the front-end terminal servers at Edinburgh,
and as the basis of the Edinburgh Remote Terminal Emulator (ERTE)
which submitted jobs as if they were typed by remote users, to
do accurate performance monitoring without the Heisenberg interaction
of measuring the performance on the system itself.
This has no relation to the 'fuzzball' arpanet routers, written
by David Mills (of NTP fame) who was also at Edinburgh at the time.
Another PDP/11 O/S was "MUSS". There's a manual page for it
here:
http://history.dcs.ed.ac.uk/archive/os/emas/users/ercm09/emas-2900/docum_mmanual.txt
I think the sources of MUSS along with the FEP's and ERTE are in
that directory somewhere too. DEIMOS, MUSS, ERTE and the FEPs
were written by Brian Gilmore, now head of the Computing Service.
Ian Young wrote an operating system for the PDP-11 called "rats"
as a student project - I believe he has a paper listing which we
might have scanned some day. That was the first implementation
of lightweight threads that I remember seeing.
All the Edinburgh software was written in Imp, except for a few
assembler parts, but we had our own assembler for it as well:
http://history.dcs.ed.ac.uk/archive/os/emas/users/ercm09/emas-2900/asst11j.txt
There were two compilers for the system; one written by the ERCC
based on a simpler bootstrapping compiler called "SKIMP" -
http://history.dcs.ed.ac.uk/archive/languages/imp11-tmp/imp11/sk11v.imp
... and a later one written by the Computer Science department for the
Imp77 language:
http://history.dcs.ed.ac.uk/archive/os/emas/users/ercm11/emas-a/imp119s/
(and a very early version of the above, we think:
http://history.dcs.ed.ac.uk/archive/languages/imp11-tmp/earlyimp11/ )
There's a good writeup of the pdp-11 imp compiler optimisations
here:
http://www.gtoal.com/athome/edinburgh/imp/PDP11IMP.htm
Graham
Received on Fri Oct 22 2004 - 22:12:38 BST