Dan Knight - 2002.11.27

In the beginning of personal computing, there were no operating systems, for there was almost nothing to operate on. Then came floppy drives and controllers, creating the need for Apples and Ataris and TRS-80s and the like to communicate with random access media.

Each brand had its own OS: Apple DOS, TRS-DOS, and so forth. And then came CP/M, a standard OS that computers designed around the 8080, Z80, and 8085 could use. And it was good.

Then came the IBM PC with three OS alternatives: CP/M-86, the UCSD p-System, and something from Microsoft called PC-DOS.

Today Good Idea Bad Idea looks at what happens when users choose operating systems.

Good Idea

IBM didn't think much of the PC industry in those days and fully expected their 5150 to fail, so they did all they could to keep development costs down. They used an existing Intel CPU (a 16-bit one crippled for an 8-bit bus) and other off the shelf parts, which also paved the way for clones.

And, rather than create their own OS, IBM gave users a choice. They could choose CP/M-86, the logical outgrowth of the industry standard 8-bit CP/M. Or they could choose the OS of academia, the UCSD p-System. Or they could choose the upstart OS, which was cleverly named PC-DOS and sold for a lot less than the alternatives.

The rest is history. Microsoft's DOS became the de facto standard, and today Microsoft operating systems run about 90% of the world's personal computers.

Success: MS-DOS. 'Nuff said.

Success: Microsoft Windows, especially since version 3.1.

Bad Idea

Competing with the industry standard operating system.

Failure: CP/M was the first casualty of the success of the IBM PC and Microsoft's DOS.

Failure: All of the other 8-bit computers and their operating systems. Only the Apple IIGS managed a slightly successful transition to the 16-bit world, bit it was too little too late.

Failure: The UCSD p-System.

Failure: Digital Research's CP/M-86 and the multiuser MP/M-86.

Failure: Digital Research's DR-DOS, mostly due to fallacious error messages generated by the Microsoft Windows 3.1 installer claiming Windows was not compatible with DR-DOS.

Failure: OS/2, initially a joint project of IBM and Microsoft to create a next generation OS to replace DOS. Microsoft took what it learned, applied it to Windows, and left IBM with a great OS that failed in the marketplace.

Conclusion

Throughout this period, Unix has persisted, the Mac OS has become the leading desktop alternative to Windows, and Linux is the leading alternative OS for industry standard hardware. Between them, these operating systems account for maybe 10% of the entire personal computer market.

As Digital Research and IBM discovered, you cannot compete directly with Microsoft; they will roll over and crush you. The Mac OS remains viable because it runs on proprietary hardware. Linux is viable primarily because it is free, does not require proprietary hardware, and gives tech types a stable, viable, open source, virtually free alternative to Windows.

<this article available in a printer friendly format>

---

<home>

Good Idea Bad Idea was inspired by a few too many episodes of Animaniacs - if there is such a thing. GIBI is an irreverent look back that the history of computing in the hopes that we'll learn from the mistakes of the past rather than repeat them.

Recent Good Idea Bad Idea Columns

• Good Idea Bad Idea looks at operating systems, 11.27. Good idea: giving consumers a choice in operating systems. Bad idea: competing with Microsoft.
• Good Idea Bad Idea Looks at disk format confusion, 11.21. Good idea: increased capacity for removable media. Bad idea: too many formats to pick from.
• Good Idea Bad Idea gets on the bus, 11.15. Good idea: an industry standard bus. Bad idea: a bus nobody else uses.

Other Recent Content on LEPC

• Value and cost: With a PC, you get what you pay for, Katherine Keller, Thinking From the Box, 02.26. With the right choices, your PC can be easy to upgrade and avoid becoming a doorstop.
• Networking 101, Dan Knight, Online Tech Journal, 02.10. An introduction to ethernet, hubs, switches, routers, and wireless networking.
• Pixels and points, screens and paper, Dan Knight, Online Tech Journal, 02.06. What you see on the screen corresponds to what you get on the printed page. A brief history of points, pixels, and the changing face of computer displays.
• Deliver us from evil: Thoughts on computer self defense, Dan Knight, The Knight Line, 01.16. "Should you have the right to take active steps to stop a computerized attack on your computer system?"
• Why we need anti-spam legislation, Dan Knight, The Knight Line, 01.02. "Unlike rain, sending spam is a human behavior which can and should be covered by law."
• Tying up loose ends on bitty boxen, Buttercup, and the state of tech, Michelle Klein-Häss, Geek Speak, 12.02. Mandrake Linux rocks, Buttercup gets a new motherboard, and the sorry state of the tech sector today.
• more in our editorial archive

Good Idea Bad Idea articles ©2002 by Cobweb Publishing.  Low End PC (LEPC) launched September 2001. The entire LEPC site copyright ©2001-2003 by Cobweb Publishing, Inc., unless otherwise noted. Copyright of individual articles resides with the author. All rights reserved.
  Advice presented in good faith, but what works for one may not work for all. Computers are like that. Please report errors to the webmaster.
  Letters sent to LEPC may be published at our discretion. Email addresses will not be published unless specifically requested. If you prefer that your message not be published, clearly mark it "not for publication." Letters may be edited for length, context, and to match house style.
  LINK POLICY: This site allows and encourages links to any public page, so long as the linked page does not appear within a frame that prevents bookmarking the linked page.
  PRIVACY POLICY: In brief, we don't collect any personal information unless you explicitly provide it.
  Low End PC is an independent publication and has not been authorized, sponsored, or otherwise approved by Microsoft, Intel, or anyone else. Many company and product names may be trademarks or registered trademarks of the individual companies and are hereby acknowledged.
  For more details, see our Terms of Use.