25 Oct
2000
25 Oct
'00
7:46 a.m.
Date: Wed, 25 Oct 2000 10:39:51 -0400
From: Douglas Quebbeman <dhquebbeman(a)theestopinalgroup.com>
Subject: RE: classiccmp-digest V1 #407
Hello, Douglas:
A guy named Roger Roach who is or was at MIT
supposedly has magtapes containing the CTSS
d ocumentation. If he does still have them,
I doubt he would release them from his possession,
and I don't think he personally has the time
to try to recover the contents.
:-(
He might permit a researcher access to the
tapes at a facility local to MIT or wherever
he now happens to be.
Maybe in my next vacations. :->
No, seriously, maybe a goal that this tapes could be
translated to Pc Format like Paul Pierce made with
IBSYS and PR150, two OS's for the 709 (if I remember
good...).
I do not know Roger; I got this information
from a third party.
Well, if would be possible to contact via this
third party...
Most of Multics is written in PL/I; unfortunately,
PL/I compilers are hard to come by.
I have two PC PL/1 compilers and some other from
CP/M and know somwthing of the language. I am
waiting one book from IBM that I suppose will
complete my stuff to begin.
I, too, have had Java recommended to me as a language
for implementing emulators. I have concerns about its
performance.
I needed to work with Java from two months from now to
develop one financial application to my job. It offers
all what I need to develop this emulators but specially
to simulate a time-sharing system.
Teorically all the papers about Multics that have now
maybe sufficient to simulate something like the
pagination mechanism. and the virtual memory. The
specifications about the first Multics let me do
a virtual disk from 4 Millions of words of 36 bits.
There is an study about the modules that compounds
the kernel and the recommendations to reduce its size.
I have some source code in PL/1 of some Multics utilities
and a list of 314 commands like appears in
http://www.multicians.org/multics-commands.html ...
I only needs the messages of the commands to simulate them
but logically I should agree the source code :-) And, in
other way, the mechanism of console or terminal session
that Multics used. That's all for now. I prefers at this
instance to made an hipotetical machine that runs the
System before than a real General Electric 635/645. Even
I'm not interested by now in the security mechanism that
were developed later.
Well, that's all for now. Thanks for your kindly answer,
Douglas. Greetings from Spain.
----------
Sergio Pedraja
Administracion de Sistemas
Division de Tecnologia
Caja Cantabria
Spain
----------
mailto: spedraja(a)mail.ono.es, spedraja(a)casyc.es
----------