[inforoots] Operating Systems from CDC

[Peter Capek]

Let me try to clarify a bit....

CTSS was the system developed at the MIT Computation Center ca. 1963-4.  The
name stood (maybe still does...) for Compatible Time Sharing System.  What
it was compatible with was the FMS (Fortran Monitor System) interfaces which
were in previous use on the 7090 system.  FMS was a batch system provided by
IBM (but, I believe, based at least in part on earlier work done by SHARE,
the user group).   In true MIT fashion, the hackers at the AI lab, primarily
Greenblatt and Gosper, I believe, created ITS, the Incompatible Timesharing
System.   How incompatible?  Well, it ran on a PDP-6 and then a PDP-10, so
not even the word length was the same.    So CTSS and ITS have nothing to do
with the 6x00.

Chippewa, also sometimes written as COS (Chippewa Operating System) but
never in my hearing enunciated that way, was the system written in Chippewa
Falls that was the only real
system for the first couple of years of the 6000 era.   (This is excepting
Livermore, about which I know little; at some point they got their own
time-sharing system running, but I don't know when that happened, or whether
they ran COS at first.  I think this was called Octopus and/or LTSS.   Can
any others on the list clarify?)   CDC was advertising something called
SIPROS (Simultaneous Input Processing Output System), but it was never a
real system running work in a production shop.

Chippewa was written in absolute octal without benefit of an assembler.  The
"source code" consisted of files with two columns of 4-digit octal numbers,
each line containing the PP address and contents of  a 12-bit PP word.  I
know this because I translated a good piece of it
to assembler, once we had a real assembler.

The basic structure of Chippewa (by which I mean: a dedicated monitor PPU, a
dedicated display program PPU, dynamically allocated PPUs for all other
functions, control points, essentially the entire OS function in PPUs, and
the details surrounding that) survived for quite a while in a number of
guises, including MACE and SCOPE.   At some point, based on additions to the
CPU architecture, it became possible to migrate OS function into the CPU.
But before that, work was done to increase the number of control points
(roughly: jobs, although system functions like spooling input cards to an
entry in the job queue, or printing output both took a control point per
device, initially.)  Initially, COS supported 7 control points, as Richard
says, but that was increased independently by a number of installations.

Dave Cahlander had nothing to do with Calidoscope.  He did create MODIFY,
the 2nd generation of source-maintenance program, and a very nice assembler,
but I don't think it was the first assembler, which I recall having been
done at NCAR.  He did also work on MACE which in interactive form evolved
into Kronos.

If this conversation is boring to the uninvolved I apologize, and will take
a gentle hint...

          Peter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.computerhistory.org/pipermail/inforoots/attachments/20060317/268bfd98/attachment.html