Hacker News Reader

Comment by yuhong | Hacker News Reader

http://www.os2museum.com/wp/before-os2-was-os2/

I assume that most users of S-100 systems were technical users who know for example how to modify the CP/M BIOS code, right?

View on HN

A number of small companies sold packaged S-100 systems with customized CP/M provided. I bought an S-100 bus box from "California Computer Systems" which included a serial port board and disk interface board, and they provided a BIOS set up for those devices. They also provided the source, and over a couple of years I modified the BIOS several times, adding devices and improving things generally. But lots of people bought turnkey systems like the Sol [1] and IMSAI [2] and never touched the BIOS.

[1] http://www.oldcomputers.net/sol-20.html

[2] http://imsai.net/

I do wonder if there is room for a solution like the S-100 bus these days. That is perhaps the one thing that irks me about todays computers, that the motherboard is dictated by the CPU socket soldered on.

The capabilities built into today's chip sets and motherboards is just insane. This weekend was the first time I plugged a board into a PC in over 10 years, to add USB 3.0 to a system too old to have had it already. The last PC I put together was based on a mini-ITX motherboard, simply because I knew from experience that I wouldn't need anything more expandable than that.

I don't think it'd be easy considering the speeds you expect from the various buses. You could put everything on a PCIe card and let it drive a backplane through a specialized connector, but that wouldn't be like S-100.

A couple weeks back I was thinking about building a backplane that would allow a bunch of BBC Microbits to cooperate. That, too, would not be like an S-100 bus.

My first computer was put together from pieces, but at its heart was a CCS case, backplane, and power supply. It was a happy medium between totally geeked out (IMSAI) and dumbed down (Sol). Of course I wrote my own BIOS from scratch.

Yes, at least in the early hobbyist phase. When you bought a machine (or parts of a machine, the line is a little blurry with S-100 bus), you usually also got the ASM source code for the BIOS. If you wanted to add a printer or other peripheral usually you had to make changes to the BIOS and recompile. It was actually pretty easy, the source code had a lot of comments and was easy to modify.

Later on, vendors such as CompuPro sold complete "turnkey" S-100 bus systems that were intended for less technical users.

Depends on context. In the early days many s100 users assembled the boards themselves and doubtless wouldn't have blinked any eyelid at hacking 8080/z80 bios code if they needed to (I remember the first kid at school to get a computer around that time, programmable with physical switches, binary LED display). By the time the commercial software market for CP/M became established with killer apps like word star, the users would have been rightly horrified by bios hacking.

Of course, MS was also not as willing to provide the MS-DOS OAKs/BAKs to random users than DR was with CP/M. When they finally stopping copying the model before OS/2 1.0 was released, they still did things like putting the floppy and hard drive drivers in the same file.

Reference for people like me who didn't recognize the acronym OAK: "OEM Adaptaion Kit", <http://www.os2museum.com/wp/ms-dos-oaks/>

And BAK stands for Binary Adaptation Kit.

I wrote a bios to try to get CP/M up and running on my IMSAI. It does require a lot of deep knowledge about the machine which admittedly was more common in those days. However I suspect most people bought a pre-packaged system.

If you were doing most of your coding in BASIC its unlikely you'd be able to modify the BIOS.

Reply on news.ycombinator.com