POLY-88 Emulator (8080 based system from 1976) C Source Code and supporting resources

Discussions
1

Over this past summer, I got in contact with Bob Bybee. He is an original owner of a POLY-8813 system and highly engaged and active in the heyday of that system between 1976 and 1980. About two years ago I obtained a POLY-88 from another original owner, who passed on to me a massive collection of tapes, documents, and an “almost” working system. That system has now been restored and I’ll post more information about it at some point in 2026.

But for now, I promised Bob Bybee to help maintain a github repository of his POLY-88 related emulator work and image collection. That collection is now available here:

voidstar78/POLY88_RESOURCE: Repository of PolyMorphic Systems POLY-88 related resources

And I’ve prepared an initial “release” of his emulator (as a pre-compiled product and a few associated samples). I had hoped to have this prepared before Christmas 2025, but alas am a few days late. And I’m also still on travel - so I plan to refine and double check all this on my return, and prepare more examples and scripts of how to use the emulator (perhaps even a video).

But I am interested in contacting others who recall anything about the POLY-88 system and their recollections. I have another contact in Illinois that also has quite a collection of material, and between I think we have most of the highlights of the system covered. But I’m also interested in stories from contemporary users of this system, what appealed to them, what workarounds they did for the poor case design, etc.

3 Likes
2

Nicely done! I do hope we get to hear some stories.

As a side note, that’s a great image of your collection of cassettes!

image
image1920×1465 637 KB

1 Like
3

Thanks! And that’s only about 1/3rd of them :smiley: The others are being archived by other folks. One hard thing about the POLY-88 is they had two different storage formats for the tape, and so it’s not always obvious which format is being used (sometimes the format is annotated in the handwritten notes, but not always)

Also, I assume the POLY-88 is fairly unknown in Europe? That is, I don’t think PolyMorphic Systems grew large enough to get involved in international orders.

4

I would think so - although I haven’t done any specific searching. I’d be inclined to go a bit further, and say that S-100 barely came into the picture for me (in the UK) other than by way of BYTE which is of course an American magazine.

(I notice Transam, a UK company, had an S-100 offering - the Tuscan.)

5

Contact Mike Douglas.

6

I’m not an expert about bus hardware, but my understanding was both the Apple bus and S-100 (unrelated to each other) both had some limitations. For example, on Apple, one couldn’t place any card into any slot? (at least, in the only Apple II system I got, I was told the cards it had, they had to remain in their slots) And the “flaw” of S-100 was each card had to include its own power regulator components (increasing the cost and generated heat?). I never dug into how they differ from the prior PDP bus architecture and the later PC ISA bus.

7

While PC slots (and its 16 bit version, “ISA”) have all pins in common, the Apple II slots (and the later NuBus) decode addresses on the motherboard and have a separate select signal for each slot. That means that you can just install two identical cards and they will show up in different addresses, so there is no need to fiddle with DIP switches or straps to configure a card so it doesn’t conflict with others. Besides their own memory address space for hardware registers, each board got 256 bytes to have the drivers in an onboard ROM (in the case of the NuBus the drivers were written in Forth so they would work no matter what processor your machine had).

Software using a card would have to look for it in all possible slots, but most lazy programmers didn’t bother. That is what made it necessary to put some cards in a given slot and not in any other.

ISA cards eventually replaced the DIP switches and straps with configurable hardware which could be set by the plug-n-play operating system instead of the user. PCI cleaned that up having a dedicated address space with which to access the configuration hardware.

Two downsides of the Apple scheme are that the maximum number of slots is fixed and that all cards get the exact same memory space. The PC alternative allows each card to use as many or as few memory (or i/o) addresses as it needs (as well as interrupt and dma lines).

The S-100 power regulation scheme soon fell out of favor, but I wouldn’t call it a flaw. Early boards had very different power requirements (+12V, +5V, -5V, etc) and it was easier to let each one handle its needs. And trying to generate 5V at the supply itself might mean you only had 4.5V by the time you got to a corner of a far away board. With everyone settling on +5V for everything and specially when multi-layer boards came along it was just easier to do the regulation just once. Note that we have come full circle and are back to the S-100 power architecture: your board receives +5V from a USB-C plug and then has a 3.3V regulator for some i/o chip, a 1.8V regulator for the SDRAM, a 1.2V regulator for the FPGA core logic and so on.

3 Likes
8

To be fair, for many early cards, there was no established mechanism for identifying the card; programmers usually just used some few bytes in the firmware of the card. This was not robust to changes in the firmware, or to all third-party cards with compatible interfaces.

Later in the life of the Apple II (post-dating a lot of very popular cards like the Disk ][, both serial cards predating the Super Serial, _etc.), Apple defined the “Pascal 1.1 Firmware Protocol ID Bytes” in Apple II Miscellaneous Technical Note #8. In the revised version linked here, Apple notes that this format was too small for the ultimate popularity of the Apple II line, and that it no longer tries to maintain a complete registry.

Post hoc, there are tools such as Card Cat that successfully identify a very wide range of cards; certainly it can be done!

Certainly some programs were just lazy; I think early in the life of the Apple II (and a lot of very popular software is quite old!) it was more forgivable than later, when you could assume that your user had at least 48k of RAM and a floppy disk drive. :wink:

1 Like
9

I agree 100%. My goal was to point out that this was a software issue, not a hardware one. There were other systems back then which did have dedicated slots into which you could only plug specific cards, but the Apple II mas (mostly) not one of them. I wrote “mostly” because slot 0 was special for memory expansion, being replaced in the later Apple //e machines with the “aux” slot and slot 7 was the only one with the sync and color ref signals (so the only one where PAL/SECAM conversion cards could go).

10

Not entirely although as an 8080 machine with its own obscure OS it never afaik made it over. I have done some emulation work on it including taking apart various bits of code to build models for the SCSI interface and some of the other later bits.

EmulatorKit has the work I’ve done so far, all open source, it can run CP/M and PolyDOS with some configurations although I am still working slowly on some of the hard disk interfaces.

It’s an interesting OS in that it’s very very crude in the original form but the people who wrote it were very much familiar with “proper” operating systems. It grew some interesting things as a result over time like basic volume management and virtual drive assignments but always on top of the pretty clunky original OS. The two user mode is also really really hacky.

The source that exists is also worth a read, it’s got some entertaining comments and some quirky and funny named commands.

1 Like
11

nubus drivers for the PPC bootstrap may have been forth (or openprom which is kind of forth but) but the actual system drivers were not, and the M68K machines didn’t use that approach at all.

Also in the ROMs for each nubus card though are the icons for them, which is fun.

12

I’m glad someone remembers Polymorphic. I used to drive past their building by Ward Memorial Blvd. (CA-217) on my way to UC Santa Barbara. They lasted years past the point where I thought their computer made sense and then finally went out of business. The sign was still there for years more.

1 Like
13

In that “critical year” of 1976/1977, the 12 “early trinity competitors” that I’ve come up with are:

Alpha Digital Systems, Ohio Scientific, Horizon, Versatile 2, Sphere Corporation, IMSAI 8080, Veras Systems, Vector, Processor Terminal, Systems Research, Polymorphic Systems, and SWTPC (Southwest Technical PC, I think)

Each of these has an interesting story and were right there to compete.

Like one number we have is that Sphere made about 1000 units, but then “mysteriously” disappeared (the owner rather suddenly closed shop; it’s been awhile since I followed the story- but I recall my impression was he didn’t run away with the money, but I think had gotten overwhelmed with loans and had to call it quits).

IMSAI kind of trounced MITS (Altair) as a clone- well, more than a clone, a bit more of a turnkey system. MITS surrendered and sold to Pertec Computing, who also couldn’t really save it.

And there was Processor Technology with the SOL-20.

There are rumors that part of the early Apple success was they had the foresight to buy up a large supply of RAM chips (this being around 1979) to such an extent that it hindered the various other computer-startups at the time. I’ve never really been able to verify that - but it was a combination of other things also: the Apple logo itself and relation to Isaac Newton I think had a lot of consumer appeal (i.e. Apple always had a keen sense or marketing appeal). Then VisiCalc was first for the Apple (followed by Commodore a year later) - it’s interesting that “almost everything” (software wise) was developed on a mini before it appeared on micros (and micros had to make reduced or compromised versions). But VisiCalc was an exception, as far as I can tell that was truly an original innovation from microcomputers first. Around the same time was the Microsoft Z80 card for the Apple - so now it could do CP/M things also.

I think Vector, Ohio Scientific, SWPTC, and PolyMorphic hung around for a few years - until the IBM PC came out. The “trinity” (IMO) gets remembered since they had the resources to survive beyond the IBM PC - essentially up until 1993. At that point, Tandy switch basically to PC-clones, Commodore ran out of resources, and Apple wasn’t too far behind.

2 Likes
14

4K Dram was all ways in short supply,But at $20 a chip only main frame users could afford them until
after 1976.1977 ?? $10 1978 ?? $5 The Z80 and the 6502 or 6800 was better with dram than other micros so that implied more designs using 4k dram.

15

and Cromemco! Cromemco Z-2 - Wikipedia

16

The 1970s ended with the microcomputer market split between three kinds of companies:

  1. organic growth, where money from sales goes into making more products (all the companies you listed which ended up disappearing)
  2. big companies (Radio Shack, Atari, Texas Instruments, Commodore)
  3. venture capital backed startup (Apple)
17

Technology maturing every 5 or 6 years, meant you also needed to develop the next computer version
with this model sales. Apple got lucky here. IBM knew to do that once the PC started selling. Others just cloned what was popular, or just vanished.

18

As an quick note, someone needs to edit this topic title. It was an 8080 based system not 8088, of course.

I have a few pictures of my Polymorphic Systems 8813 here: Polymorphic Systems – Poly 8813 – VintageComputer.ca I did check and I have not yet archived my floppies pictured in my blog post. I need to get to those soon. Great S100 system though.

1 Like
19

(Good point, topic titled edited - I assume that’s ok, @voidstar)

Nice photos of your system - as ever, it would be good to have those floppies imaged and ideally shared!

1 Like
20

RAM chip supply was a big issue, and also the tools needed. The lack of supply of the intended RAM is one of the things that sank Nascom.
Some of the official tools like the testers for the 4116 were also very expensive. Memotech famously built their own using microcontrollers rather than buy the mindnumblingly costly tools

1 Like