OS INFO POST

Harry Barman hjb%cl-steve.cam at UCL-CS.ARPA
Tue Mar 12 09:13:58 AEST 1985


Posting for a friend....

-------------- forwarded message ---------------
> From jpb at uk.ac.cam.cl Mon Mar 11 14:11:40 1985

Further to the discussion on operating system sizes readers may be interested
to know about the TRIPOS operating system, developed at Cambridge University
England in the late seventies, and in use at a number of British Laboratories.
It was intended to be a small, portable operating system for real time
applications. It is used experimentally, and there are a large number of
versions in use for different purposes, which makes the figures I give below
very approximate. It also does not provide some of the features found on large
operating systems, notably any form of memory protection. However it is still
in widespread use, to some extent for teaching, but mainly for research into
distibuted computing on high speed local area networks.

I shall give figures for two versions. The first runs on Computer Automation
LSI4's, in a disributed environment (there is only one I/O device, the LAN
interface). This comprises of the order of 10 000 lines of code (half BCPL,
half assembler), and produces about 40-50K of object code.

The second version is a standalone version which runs on a BBC micro (an 8 bit
6502 based micro, with 32K RAM, and OS and BCPL run-time system in ROM). This
although a very rudimentary system (it has a simple command line interpreter
and floppy disc device drver only), provides full multi-tasking facilities on
the micro (about 200 - 300 task switches per second). It comprises about 3500
lines of code (BCPL and Assembler), producing 12K of object code (but bear in
mind there is also the BCPL run-time system in ROM).

Readers may be interested in this example of a small operating system which is
still useful in 1985. Further information can be found in:

M. Richards, A.R. Aylward, P.B. Bond, R.D. Evans and B.J. Knight "TRIPOS - A
portable operating system for mini-computers", Software Practice and
Experience, 9, 513 - 526 (1979).
---------- end of forwarded message ------------



More information about the Comp.unix mailing list