Trying to recall an OS I worked on that was called CPM but was not DR's CP/M

Question

I'm trying to trace a type of system I did a little work on in my first job, which was for a company which no longer exists and I am not in regular touch with most of my former colleagues.

It was a desktop-minicomputer style system, with a system unit and dumb terminals on (I think) RS-232.

The main system unit was a big box, and had an integral CRT. I.e. the host machine had an integral terminal, or was even built into a terminal. This unit had an 8" floppy drive, vertically mounted to the right of the CRT. Behind the CRT in the (large, densely-packed) case was a hard disk drive of bigger than 8" platter size -- in fact the biggest I've ever seen in anything I actually worked on. 12" platters maybe? The capacity was quite small, I think 5MB to 20MB on bigger ones.

It supported multiple users on dumb terminals.

The system unit had a lot of boards in it -- I have a feeling the processor may not have been a single chip, but might have been distributed across several boards. We only had a few customers with them, and they were already very old machines when I worked on them in 1988-1989. Only one guy in the company really knew much about operating them, and he was also our main Alpha Micro engineer/programmer.

I am fairly sure these were not Alpha Micro machines, though, and the OS was nothing like AMOS.

Applications software was I think mainly bespoke, accountancy or maybe stock-tracking type functions.

The OS was called something like CP/M-83 or something -- I think it was CP/M followed by a number but it might have been CPM* (CP/M Star) or something. I have a vague feeling it was not punctuated in the normal CP-slash-M place.

The CLI was not even remotely CP/M or DOS-like. In my very minimal contact with things like IBM S/36 and AS/400, it was more like that: commands were short abbreviations, sometimes just a letter or two, with arguments that were often numeric. It was more like interacting with a machine-code monitor than any other shell I've used (quite a few: DEC VAX-VMS, DOS, Netware 2/3/4/5, etc.) It was not a DEC machine, either -- quite a lot of my early career involved VAX, Alpha and briefly PDP-11.

We also supported Jarogate Sprite systems, running Concurrent CP/M -- it was not them, but it was an older system that was used in similar roles. (E.g. multiuser accounts systems.)

Answer 1

There are potentially many candidates for this question ("CP" for "Control Program" was used frequently in that era), so just to mention one:

The Xerox Sigma 9 had two Operating systems with a name similar to CP/M: CP-R "Control Program for Real-Time", and CP-V "Control Program Five". The latter included timesharing (multiuser).

You can find manuals at bitsavers and the Living Computer Museum, so please have a look if anything looks familiar.

There are also pictures at the museum website.

At least in this installation, the main cabinet does not have an integral terminal, but often this was somewhat configurable (though I don't know if this was the case for the Sigma 9).

Answer 2

(I'll leave the other answer there, because the linked system and OS may be interesting).

If the "big box" was a desktop unit comparable in in size and form-factor to a DEC VT-50", then I would assume it was not a discrete multi-board CPU, but a single IC CPU.

The only multi-user system similar to CP/M at this time I can think of is MP/M. As you can read there,

In the early 1980s Digital Research also developed a networking software named CP/NET used to connect an MP/M server with multiple CP/NET clients (named requesters) running CP/M. [...]

MP/NET was an MP/M system with networking allowing the MP/M system to function as both requester and server with CP/M requesters.

The CP/NET clients could also be run in a diskless configuration with the system stored in ROM, then named CP/NOS (with NOS for Network Operating System).

Similar, MP/NOS contained MP/M without local disk facilities. Like CP/NOS, MP/NOS performed the disk functions through the network.

So then the candidates for the "OS with a name similar to CP/M" would be "CP/NET" and "CP/NOS". A manual for CP/NET can be found here

However, that doesn't fit the description of the CLI: Both CP/NET and MP/M had commands very similar to CP/M. (However, it's possible that the "commands were short abbreviations, sometimes just a letter or two, with arguments that were often numeric." language was something that ran on top of the OS). It also would involve some kind of networking, not just dumb terminals connected to a central machine, which would be MP/M.

So again, not a complete fit, but possibly it helps in remembering more details.