I had a Gifford S100 system running Concurrent CP/M and with telephone-voice interface boards that ran a voice-based dating site I designed and programmed in Pascal. The CPU was a 286 running at 8mhz. It was in NYC, with phone number 212-ROMANCE.
It's interesting to imagine an alternate world of CP/M-86/S-100 vs. MS-DOS/ISA for x86 PCs.
The brilliant bit to me is how modular S-100 seems to be with everything on a swappable card. This seems to enable a more granular upgrade and system build approach compared to what we have with modern PCs where most core components are stuck on a single motherboard.
I think I might be interested in a modular, semi-disaggregated PC based on a PCI backplane.
I was actually doing computer repair when the IBM PC came out. It had a backplane, but most of the system was on the motherboard. Repair people were horrified that many if not most of the logic chips were soldered in, not in sockets like the Apple II. At the time, I had repaired Apple II's, and dumb terminals that had soldered IC's. Half of my time was spent painstakingly de-soldering chips from boards so a replacement chip could (hopefully) solve the problem.
On the other hand, most Apple II repairs consisted of what we jokingly called "the laying on of hands," which meant reaching in and pushing all of the chips back into the sockets with the palms of your hands. That was always accompanied by rubbing the gold fingers on the disk controller card with a pencil eraser because the gold plating was inferior. Many Apple II repairs were 5 minutes. Third on the list was power supplies, which we never figured out how to repair at a component level, and were justifiably afraid of.
At the time, IBM published some data based on their own years of computer service. They serviced a lot of their own boards, and made their own memory chips. Their data showed that chips were actually more reliable than sockets, so they limited sockets to those which had potential for fallout in production (the 8088), and the memory chips which were an optional add-on. And the ROMs, which could potentially be upgraded.
> This seems to enable a more granular upgrade and system build approach compared to what we have with modern PCs where most core components are stuck on a single motherboard.
You end up paying for that increased flexibility in a number of ways: greater physical size, increased part count, increased manufacturing costs. I think even if the industry went with CP/M-86/S-100 instead of MS-DOS/ISA, sooner or later people would have put the CPU card and the backplane on the same PCB, and we would have ended up in more or less the same place – the cost advantages of doing so would have made that development inevitable.
> I think I might be interested in a modular, semi-disaggregated PC based on a PCI backplane.
You can buy those, they exist – there are passive PCI backplanes [0] into which you can plug a CPU card. They are used for industrial control applications. You are going to pay a premium for that though. As well as increased number of parts/etc, the lower volumes they are manufactured in adds to the cost.
Some early PC compatibles were based upon the same backplane/card approach as the S-100 computers. I have a Zenith/Heathkit Z-160 transportable in my collection. It has an 8 slot 8-bit ISA backplane and everything including the cpu is on ISA cards. This backplane is shared with the Zenith/Heathkit Z-150 desktop. They also released AT class desktops, by this time the portable PC form factor had become too small to support this approach, which used a 16-bit ISA backplane before eventually abandoning the approach for their later computers. Most likely for all the reasons you point put.
IMO S100 is under-represented in the history of microcomputers. That history is usually represented as Apple, Tandy, Commodore and Atari vs IBM PC & clones. But in my opinion, the IBM PC & clone / MS-DOS ecosystem is a fairly direct successor of the S100+CP/M ecosystem.
The "motherboard" of most S100 systems AFAIK were simply a a power supply and bunch of slots. Nothing onboard, your RAM fit in a slot, your CPU card fit in a slot, your video card or serial port for your terminal fit in a slot, your disk controller fit in a slot, etc.
The original IBM 5150 had RAM and CPU onboard, as well as a tape I/O port.
I did run into an old Zenith PC clone at a Goodwill a while back. I think dated from 1987.
Really weird system - the motherboard was an S100-style backplane that only had the Motorola MCxxx-whatever clock chip (a real one and even socketed). On a big ISA-style card was a 286 with a lot of other chips, another big ISA-style card that looked like RAM, and a third that had all the discrete components of your PC chipset - the PIT, the two UART chips, the two DMA controllers. I forget if the PICs were there or on the CPU board.
The BIOS was also weird ... had a built in monitor complete with assembler and disassembler commands.
Every single chip was socketed. It was ... bizarre. Had an MFM drive running DOS 2-point-something and an ISA VGA card in one of the slots. Still worked.
Prior to the Z-286 line Zenith made a line of PC compatibles with an 8 slot 8-bit ISA backplane design, the Z-150 desktop and Z-160 transportable. The backplane board lacks the RTC and keyboard connection of the Z-286 backplane board and is therefore completely free of active components.
I agree the history is biased towards the more mainstream home systems, as opposed to the sized-down business systems, which is what the S-100 computers amounted to.
In fact, the distinction between "home" system (a mostly non-extensible computer in a plastic case with a ROM BASIC as an OS and a focus on graphical games and edutainment) and "business" system (an extensible, if not a kit, system in a metal case with CP/M or MP/M as an OS and a focus on business software) is lost, even in the histories.
From a hardware POV, PC was more similar to S100 than it was to Apple II. The S100 bus was a dumb extension of the 8080 pins, the PC a dumb extension of the 8088 pins and the Apple II a dumb extension of the 6502 pins. The 6502 and 808x micros had very different timing cycles.
In comparison the fact that PC & Apple II used a motherboard architecture while S100 used a backplane architecture is very superficial. It was possible to buy S100 systems that used a motherboard architecture, it's essentially just hardwiring a connector. Here's one: http://www.s100computers.com/Hardware%20Folder/HeathZenith/M...
The other comparison point was in ecosystem: the S100 had a very rich ecosystem of vendors putting out systems and expansion cards, just like the PC Clone ecosystem.
Finally, the S100 and PC were open systems, whereas Apple was quite closed.
Not super S100 focused but, in the early 80s, I used an Alpha Microsystems S100 multi-user system with a WD16 processor. Later, the company I worked for, a video production company, upgraded to the 68000 based Alpha Micro. The whole company ran on that thing, accounting, writing and I even wrote a small program allowing it to control a Chyron Scribe video character generator in the edit suite.
I remember liking the system very much but admit that I didn't have much with which to compare it. I thought their OS, AMOS and AMOS/L, was great.
Someone I know who knew Bill Gates fairly well at that time told me that Alpha Microsystems was a company that truly scared Gates.
In 1986, I met Dick Wilcox, one of the Alpha Microsystem founders as he was buying a Pitts S2B at the Auburn, WA airport.
That was almost 20 years ago now, I don't know how much of their business actually survives. They have a website live still, but it seems quite non-functional: http://www.alphamicro.com/
I spent many hours coding purchasing and inventory systems for our company in the late 80's on our AM-100/L.
I tried to get it working again just for fun in the early 2000's but it had some hardware issue and Alpha Micro was less than enthusiastically interested in helping me with technical information. Ah well.
It's still sitting in a storage room to this day. Probably hopelessly defunct, but I just can't bring myself to get rid of it.
My first machine as a mad 14 year old was an S-100 system I build from bare board kits and programmed (initially) from a front panel I designed. My dad eventually took pity on me and got me a serial terminal.
I had a Z80 CPU that could handle a blazing 2 MHz IIRC but I remember I could never run at that speed, though I could do more than 1 MHz if I disconnected my front panel.
Heh, I had a IMSAI 8080, pretty sure it has a S100 interface for add in cards. I started with a paper based terminal then upgraded to the Hazeltine 1410 I think it was. Amusingly the paper based system (teletype) was 110 baud, but had flair/drama. I remember playing star trek and listening for each position that was a "." for empty space or ** for an enemy.
Playing the same game at 300 baud on the hazeltine 1410 was pretty boring in comparison.
My first job out of college was at Scion Corporation, who made the MicroAngelo S-100 graphics boards. I did Z80 programming for the firmware. They were nice boards but very expensive so I never got one for myself. The company nearly died when their follow on board for the IBM-PC was a failure, but that's a longer story.
What I liked about those days of computing is that you could invest time in a system and not expect it to be obsolete by the time you felt comfortable with it. And also that you could actually know the whole system. In that sense modern software development is very quick to give you some result but you're building on so many layers of potential trouble that the result never really feels quite as solid and deterministic. Old computers failed because the hardware failed, modern computers fail in weird organic ways when some subsystem far down in the stack temporarily gives out and you'll never know what actually caused the problem.
'Refresh the page' is the modern day equivalent of 'control-alt-delete' from PC era, but before then stuff usually 'just worked' once it worked and it kept on working until the hardware gave out.
I think this is part of what's behind the surge of interest in retrocomputing. As you say, those old 8-bit systems could be understood in their entirety by a single person. I know it's part of what has drawn me into the world of Arduino and other microcontrollers. It's fun to play with a simple system and get instant feedback from a blinky light.
Having said that, 512x480 with non-square pixels is a blight on the Universe.
Was that in / near the University of Maryland in the late 1970's? I was working at a computer store in Rockville as a teenager. Chuck Rieger came in and showed me and the other employees the MicroAngelo. It seemed to be very expensive and very complex, if memory serves me right.
Scion was in Reston, Virginia, but Chuck was a professor at the University of Maryland as well. He eventually quit UMD to work at Scion full time. He's the reason I got the job, he was one of my UMD profs. The MicroAngelo was easy to program but expensive because it was basically a full computer. It had an onboard Z80 and 32K (later 64K) of RAM. I worked there from 1981 to 1983-ish, when they fell on hard times and had to lay a bunch of people off.
I'm immensely interested in the history of computing and computers. If it was ever the case you'd consider being interviewed about your experience (either on this topic, or others-- your HN comment history makes me think you've got a lot of interesting experience) please reach out. I don't know that I'd necessarily want to do the interviewing (it's nothing I've practiced), but I'd definitely try to help facilitate it. These stories are too valuable and interesting not to be told.
A Sol-20 with monitor sold at an estate auction near me for so cheap not too long ago. I had no idea what it was and had my eye on the Apple IIc instead. I did end up winning a box of S100 bus cards and Sol-20 tapes though, including a "Micro Soft" extended basic cassette. The lot I won also came with an S100 bus "music synthesizer board" which is pretty awesome. I ended up selling most of it off to fund my stupid Vintage Apple collection though.
S100 was also a widely used industrial bus. My first job out of college (1980s) was writing software for a machine tool where the controller was a custom s100 80186 motherboard and a bunch of ROM and static RAM s100 boards. We used a ton of static RAM (probably 4-6 64k boards, I can't remember exactly) that were battery backed up. You could power cycle the controller and just start back where you left off.
I'm not sure if it's what you mean, but the RC2014 ecosystem ( https://rc2014.co.uk/ ) is a modern-retro design fairly heavily inspired by S100 systems. It's a backplane + boards Z80 system (...with a few other CPU boards made for it) intended for modern hobbyists. Little bit expensive because it's boutique (Minimal system is about $100, a big nice CP/M box with good I/O would be $300ish). I haven't had the disposable time and money for one, but they look like good fun, and honestly pretty easy and cheap compared to real period equipment these days.
I've long been interested in that particular project; but as you say it is just that little bit too expensive.
So instead of spending the money I've been spending my time, making something similar (but much much simpler). I'm having fun, even with just a clock, Z80, and EEPROM. My next step will be handle I/O with a keyboard and LCD, hopefully by the end of the month.
In industrial applications there are still plenty of systems built on good old ISA and PCI buses with the host on one card and a whole pile of IO of all kinds of plumage on others.
Yes, that will never die, if only because it is a staple in the defense and aerospace industries and I wouldn't be surprised at all if there is still development done in that arena, iirc even the A380 had a VME rack somewhere in its guts.
I'm working at a particle accelerator, and we still have a few VME racks, but we plan to get rid of them next year. Industrial and scientific control systems are full of really old stuff.
My late friend, Lloyd Smith used offer a service where he split the DC power bus and added a second power supply to S-100 systems, because the stock power supplies were undersized for a full card cage.
I regret giving away my IMSAI 8080 S-100 box a little over 15 years ago. I should have looked on eBay. If I had, I would have known how collectable they are. I may have been able to get a few hundred bucks for it anyway. It was still in working condition, but all of my (8") boot disks were destroyed by humidity. It booted a few times, but then didn't. While troubleshooting, the disk drive ate all the remaining boot disks.
My mentor worked with drone projects in the early 80s. They flew S100 bus computers. The drone craft were at least half the size of an ultralight manned airplane.
