• Please review our updated Terms and Rules here

PDP-8 Timesharing - is anyone still doing this on real hardware?

TJ_Mossman

Experienced Member
Joined
Dec 1, 2013
Messages
370
Location
Bristol, England
Does anyone still have a timesharing system running on their PDP-8?

I know Josh Dersch modified TSS/8 to run from an RK05 on the LCM's 8/e, but since the LCM closed down I don't know of any other timesharing 8's that are still running.

With any luck I'll get ETOS running soon, as I found the required TSC8-75 board in a scrap pile. If it works all I should need to do is get my RK05 running and install a bunch of serial cards...

I'd also be interested if anyone has any information on OMNI-8, as the only reference I've seen to it is here.
 
I ran EDU-something on an SBC6120/IOB6120 for a while. Not sure if that counts, however, as there's no actual timeshare support in the 6120, and so it is more of a multi-user BASIC than a general purpose timesharing.

Getting all the ASCII terminals set up was moderately hardware/space intensive.

I think Josh has had the TSS/8 running at home recently, rather than at LCM.

Vince
 
There was someone once out there sharing an 11/45. Who boy, was it slow. Just...slow. Wow slow. "How did anyone get anything done" slow.
 
It'd be an interesting exercise to roll up a Telnet/SSH-to-serial bridge wherewith someone could set up an Internet-facing timesharing system on vintage minis that cannot themselves support a TCP/IP stack...wonder if the folks at some of the vintage-computer museums have already done something like that.
 
There was someone once out there sharing an 11/45. Who boy, was it slow. Just...slow. Wow slow. "How did anyone get anything done" slow.
 
It'd be an interesting exercise to roll up a Telnet/SSH-to-serial bridge wherewith someone could set up an Internet-facing timesharing system on vintage minis that cannot themselves support a TCP/IP stack...wonder if the folks at some of the vintage-computer museums have already done something like that.

I once had thought to do that with my 11/84 system. I did have it on my own local network, but never went further. I also have a "Black Box" for it that provides a way to connect many terminals to it via TCP/IP. Never actually got that working and am not sure it is a working box. I'd rather have set up my ex pdp-8e with TSS/8 but at that timeI dd not know of the RK05 patch to allow swapping to RK05.
 
I ran EDU-something on an SBC6120/IOB6120 for a while. Not sure if that counts, however, as there's no actual timeshare support in the 6120, and so it is more of a multi-user BASIC than a general purpose timesharing.

I remember stumbling across that on your site a couple of years ago, it's certainly on my to-do list if I can get an IOB6120.
I wonder what it would take to add the timesharing support to the SBC6120? I've idly thought about it but never looked into it enough to know if it's a matter of slapping a handful of TTL on a daughter-board, or if it would rely on internal signals that aren't accessible on the bus.

It'd be an interesting exercise to roll up a Telnet/SSH-to-serial bridge wherewith someone could set up an Internet-facing timesharing system on vintage minis that cannot themselves support a TCP/IP stack...wonder if the folks at some of the vintage-computer museums have already done something like that.

LCM did that with several of their systems. My long-term goal is to run a BBS on my 8/f and use an Avocent Cyclades console server to handle the Telnet/SSH side. The console server supports "port pooling", so incoming Telnet connections get automatically routed to a free serial port.
 
I believe timesharing is not possible on the 6120. There are certain registers and signals that are not available externally. Then again I could be totally wrong, so someone prove me wrong.
 
I believe timesharing is not possible on the 6120. There are certain registers and signals that are not available externally. Then again I could be totally wrong, so someone prove me wrong.

I think you are right about that. ISTR that the 6120 MMU swallows 62xx IOT instructions, making it difficult or impossible to detect and respond to them.

I have drawn MMU designs that are meant to supply time-sharing for the 6100 CPU. Never tested them, though.

Vince
 
Wonder why it was designed like that? Guess they couldn't sell the bigger systems for time-sharing if the 6210 could handle it?
 
I'd suspect more that the target market for the 61x0 line was embedded systems and dedicated office equipment (i.e. the DECMate, which was sold primarily as a standalone word processor despite being a fully-functional general-purpose computer,) and there was just no call for it in time-sharing applications by 1975,* when the -8 architecture was already over a decade old, let alone by the time the 6120 rolled out. But it's a good question, and I'd be curious to hear more about the behind-the-scenes story with that line of parts.

* (Then again, DEC rolled out the 8/A with timesharing option just a year prior...)
 
Last edited:
My understanding is that the Living Computer Museum is cold and dark and that all the employees have moved on to other jobs. The last news article I could find was from October of 2020 and was about the death of David Cameron, one of their engineers.

I just hate the idea that it probably won't be back.

I will probably fire up TSS/8 one of these years. It wouldn't be too difficult to hook up the 4 port M8319 (KL8A) to 4 ports on a PC and allow telnet access.

Sorry that does not help you today.
 
Wish I had a pdp-8 to fire it up on. :) My large 8i system ended up there at LCM. I hope it survives.
 
I have ETOS running on my 8/E. Played a little with Multos/8. Haven't tried the RK05 version of TSS/8 on real hardware.

It was on the machine I had online. Due to java in browser not being supported any more and other upgrades its not currently online. Getting closer to being back.
http://www.pdp8online.com
 
I have used TSS/8 and ETOS on a PiDP8. I think that it needs to be said, "it's not what the dog said, it's that the dog spoke at all". I'll bet that 99+% of the time, people at the terminals were getting listings or results at a blazing 110 baud or two-finger typing at 10 WPM on a Teletype. Computers (especially RAM and disk) were so expensive I'm sure that schools were very happy to get anything that actually worked and was usable for as many students as possible. I was a high school timesharing student in the early-mid 70's and I will tell you that a slow computer would beat the absolute crap out of no computer.

TSS/8 really was something of a "toy", the BASIC was really limited in capability and space, but it looked to me to be pretty good at teaching both the students and administrators about computing and environments. ETOS was just amazing in its time. I believe that if I had spent some time tweaking the parameters of ETOS I could have had it running remarkably well. True, it's unfair to compare a PiDP8 or the underlying emulator to real hardware, but I'll bet that the PDP/8A's extra memory capability for swapping would have been a game-changer. too bad it showed up at the end of that era.
 
I don't have anything running yet, but I had hoped to get a TSS/8 system on a PDP/8. Need to have at least one ASR33 for authenticity (IMHO), and ta couple serial ports that could have any other type of connection. In mid 70's
Just need to find the time and space to do it, I think I have most of the necessary hardware.

I cut my teeth with an ASR-33 and dialup line to Ripon Colleges TSS/8 system. learned Fortran, Focal, PDP/8 assembly, and BASIC, which was good enough to do quite a few things including a computer dating programming for a high school dance in 1975. I wrote my own ODT8 that I thought was better than the ODT that was available.
good luck to OP.
dale
 
I used TSS/8 a few times when I was in High School in 1973 over a dialup with an acoustic coupler on an ASR-33. I remember being told that this 8/e had 24k and a drum as the swap device. At 110 baud you really couldn't tell other people were using it until you got to 15 users. Then it became a dog. That machine was my first experience with assembly. And that is probably part of the reason I like assembly to this day.
 
Back
Top