CP/M vs DOS

A story over at hardwaregeeks.com reports that the developer of S-DOS is suing an author who wrote that DOS rips-off CP/M.

(In the early days of personal computers -- pre-IBM PC -- CP/M was the reigning operating system -- short for "control program for microcomputers." PCs were called "microcomputers." In the 1970s and early 1980s, it was very cool to have a slash in computer product names -- just like was cool for product names to have a lowercase "i" in more recent times.)

S-DOS (short for "Seattle Disk-based Operating System) was very important to the history of computing. After Bill Gates told IBM he could provide them with an operating system for the then-secret IBM PC project, Mr Gates went out and bought S-DOS from Seattle Computer Products, and renamed it MS-DOS (Microsoft). IBM licenced it, and renamed it PC-DOS (Personal Computer). Later, Windows (through to v98) was a graphical shell grafted onto MS-DOS with some task-sharing code added.

There were many simiarlities between CP/M and DOS. My first computer, a Victor 9000, came with both. There was no documentation provided for DOS v1, just an empty binder. So, because CP/M was "the" operating system for PCs at the time, and there were books written on it, I began with CP/M in April, 1983.

Within the year, DOS v2 added documentation and subdirectories, which caused me to switch over.(Subdirectories were areally big deal, allowing us to finally sort files into groupings.) As well, the success of the IBM PC caused most other computer companies to switch over to MS-DOS. Victor was much more agressive than Microsoft in adding new features to DOS, and we Victor owners benefited from dozens of Unix-like features that other computer owners lacked.

(The equivalent to subdirectories in CP/M was really annoying. I forget what it was called -- "sectors", perhaps? They divided the diskette in to independent areas, almost like partioning hard drives today; you could switch between sectors by pressing Ctrl+C -- but that was the same keystroke for exiting most software, so it was a pain.)

It was easy to switch from CP/M to DOS because:

* DOS was being enhanced with more features; CP/M was not.
* DOS command names were near-identical to CP/M command names.
* Microsoft included a utility to convert CP/M files to DOS -- sound familiar?

Was one the clone of the other? I suspect it was more of case of DOS building on its predecessor, like Lotus 1-2-3 building on VisiCalc.

DOS also borrowed concepts from Unix, such as the command names for making (mkdir), changing (chdir), and erasing (rmdir) subdirectories, and the nomenclature for root (\), parent (..), and current subdirectories (.).

PS: The earliest versions of AutoCAD were available for CP/M-86, the 16-bit version. AutocAD also ran on Z-100 (I think it was) operating system, as well as DOS, and for a time, Unix.

Comments

[Rod Levin]

FYI,
Z-100 was a computer from Zenith Electronics. It ran both CPM and Z-DOS. Z-DOS was a private lable version of MS-DOS that came out at at the same time or a bit before the IBM PC-DOS version. The Z-100 was cool at the time because it could, with the right monitor, do HiRes graphics at 640x480 and 16 colors. Pretty High tech for the time...

[Evan Yares]

Before AutoCAD was called AutoCAD, it ran on CP/M on an S-100 bus computer, using a processor board developed and sold by John Walker and Dan Drake's company, Marinchip Systems.

The reason for using this board was that it used the Texas Instruments 9900 processor, which had a hardware adder -- a necessary function for getting any kind of reasonable performance out of microcomputer-based CAD.

Even today, some of the limitations you find in AutoCAD (such as viewres) have roots in decisions made back in the late 1970s to overcome limitations in the hardware and operating systems of the time.