PDA

View Full Version : Thoughts on Operating Systems



mbarton
October 16th, 2004, 06:21 AM
Hi, folks. I've been doing a lot of reading lately and came across some rather nasty comments about early personal computers from some of the old mainframe hackers. Apparently (this is news to me), there was some harsh resistance to PCs from the oldschool hackers because they shipped with proprietary operating systems. Their argument was that they didn't want the company to have that much control over what they did with their computers.

Did any of the vintage computers ship with a free operating system? I'm more than a little curious if people here have debated the issue and have any thoughts on it. I'm no expert on this topic; thus my quest for info. :-)

Erik
October 16th, 2004, 08:38 AM
The original IBM PC did, of course, in a way. You could get it without DOS, but almost nobody ever did. CP/M-86 just wasn't all that popular. Most PC Clones shipped with MS-DOS.

The Osborne 1 did for sure since CP/M was part of the bundle.

The Kaypro line followed suit, as did some other similar machines (Zorba, etc.)

I don't recall any prior to that (around 1981ish) unless you consider things like ROM Monitors and the like. Those could, of course, be replaced although on systems like the Apple and Atari, they rarely were.

Erik

Terry Yager
October 16th, 2004, 11:20 AM
I hate to revive an old arguement, but I guess it depends on how you define "Operating System". Most of the early systems shipped with some kind of minimal system in ROM. Other systems shipped with the very popular CP/M Operating System, which was not "free" but usually included in the cost of the new system, along with other bundled software. Each manufacturer had to provide thier own custom-written BIOS for the machine, because every machine was different.

I can't imagine any OldSkool mainframe hackers being upset about proprietary software. With the advent of CP/M, there was a lot more standardization of software for micros than there was for mainframes. Any CP/M computer could run any CP/M program, unlike the mainframes & minis which all required thier own software. IBM programs wouldn't run on DEC which would not run on Data General, etc.

--T

Exluddite
October 16th, 2004, 02:41 PM
What would have stopped them from installing another operating system? Some sort of liscencing snafu?

Terry Yager
October 16th, 2004, 03:37 PM
Nothing was stopping them except the non-existance of any other OS. Some companies did use thier own software, most of which were similar to CP/M, like Cromemco's CDOS or Zenith's Z-DOS (& HDOS). Some ambitious programmers would roll-thier-own version of OS for thier systems. I remember reading in some of the old buyer's guides from that period that one way to reduce the cost of your system would be to buy it without the bundled software. This didn't happen often tho, because the bundle was such a bargain. The few dollars reduction in price was nowhere near the value of the software, if you were to purchase it separately.

In the mainframe world it was a different story. There was (& still is) an alternative OS available which was ported to work on most systems. This accounts for the popularity of UNIX. It was considerably cheaper and some will argue, better than the computer company's own software. When IBM charged $10K for an OS and you could get UNIX cheap or free, it was kinda a no-brainer.

--T

mbarton
October 16th, 2004, 06:50 PM
Thanks for all the great responses. This is quite interesting material!

I guess the question was a bit zany, because GNU/Linux was nowhere close to being finished when many of the classic PCs were in stores. The C-64 had a version of Microsoft Basic on it, if I'm not sadly mistaken. As I understand it, Unix wasn't really a free operating system, but it was loosely regulated enough (and, as was mentioned, cheap enough) to seep across most mainframes of the late 70s and 80s. I've encountered in on many a university VAX system. From what I hear, Unix started tightening up the security and licensing scheme until the old hackers felt a stirring need to create a free alternative. Maybe I should try to find a "history of unix" or some such to beef up own.

I've never been too sure about CP/M. It was on a Commodore 128 I owned for awhile, but I never used it. Could you describe CP/M very roughly? What it was good at? Bad at?

Terry Yager
October 16th, 2004, 08:55 PM
CP/M's greatest claim to fame is it's portability. It can be brought up on any hardware based on the Intel 8080/85 or Zylog Z80 family of processors. It is able to do so because of it's modular design. It consists of three separate modules: Basic Input-Output System (BIOS), Basic Disk Operating System (BDOS), and Console Command Processor (CCP). Of the three modules, only the BIOS is specific to a certain kind of machine. The CCP & BDOS are the same on all machines. All a manufacturer (or individual) has to do is write the BIOS (or modify an existing BIOS) to match the hardware installed in a particular machine.
It is capably of running any kind of applications, including graphics, if the hardware supports it, but most applications are text-based. Any program that the programmer wants to run on a variety of different company's machines must be text-based because only a few machines supported graphics. Many word processors, spreadsheets, communications programs, programming languages, etc. were written for plain-vanilla CP/M. Anything a computer can do can be done in CP/M (given the limitations of the antique hardware).

--T

mbarton
October 17th, 2004, 06:04 AM
Thanks for the great reply, Terry. I take it, then, that CP/M is "open source" and free to distribute and modify? It's not owned by a company or individuals?

Terry Yager
October 17th, 2004, 07:14 AM
CP/M is owned by someone, it's not Public Domain, but it is considered freeware now. When it was owned by Calderra a few years back, they released it, thanks to the efforts of the late Tim Olmstead. Binaries and source code are available here:

http://www.cpm.z80.de/

--T

carlsson
October 18th, 2004, 09:46 AM
Re: Commodore Basic - yes, it was bought from Microsoft but in some one-time license which let C= distribute it with how many machines they wished and also modify it to their needs (which they did, of course).

I wonder if you mean "source code available", "freely hackable" or just "without extra cost" when you ask about free operating systems. While CP/M and others today may be freely distributable (how on Earth that is possible if Caldera once owned the rights?!), it probably was not everyone's property in the 1970-80s. If it had been, IBM would never had to contact Digital Research (boss golfing/vacation/whatever according to the tale) about getting a license - they could just have slapped CP/M onto IBM PC without asking! :) Bill Gates would surely had loved that.

I think the commercialisation of Unix happened in the early 1980s when the two branches and dialects AT&T (commercial use?) and Berkeley (university use?) formed.

BTW, on a VAX or PDP network, wouldn't it be more common to see TOPS and later VMS than Unix, or do I mix up things now?

Terry Yager
October 18th, 2004, 10:40 AM
See the link above for a copy of CP/M's current license agreement, from Lineo, the current owners.

--T

tachyon
October 18th, 2004, 02:38 PM
Having been involved with mainframes, midrange computers (VAX, AS/400, HP3000/9000, etc) as well as personal computers, I can tell you that the majority of the 'opposition' was based on the the 'Not Invented Here' principle. Since the mainframe managers didn't bring the machines into the company, they didn't control them. In the companies I've worked for in the past few years, PC's usually fall under the telecom groups. The people who put telephones on people's desks.

An IBM PC has roughly the computing ability of a small IBM 360 series mainframe. When you realize that a 360 was a batch oriented multiuser computer, as opposed to an interactive single user computer in the PC, it puts the issue in clear perspective. If one person has a computer as powerful as the entire Data Processing department, it's a threat to the DP folks' jobs.

The issue you mentioned was the response of the Unix side DP people, who were often scientists and engineers who had been forced to take up DP because the 'professional' DP people were welded to their IBM's, while the scientific and engineering people usually used less expensive DEC systems. Of course, a DEC system was many orders of magnitude more expensive than a hobbyists' CP/M computer and because the software was rather rudimentary, didn't have a lot of competition.

It was when programs such as dBase II, VisiCalc and WordStar became popular that there was truly a threat to the Data Processing Heirarchy.

I've worked on many programs on PC's and even Kaypro CP/M computers that were written for use on a manufacturing 'shop floor' to do one specific thing. Since the computer and custom software was often less expensive than a single tool head for one of the manufacturing machines, they flew 'under the radar' of the DP departments.

As CNC machines and many pieces of manufacturing equipment became more complex and included one or more PC's to reduce costs over dedicated computers, many DP departments complained that they didn't have the 'source code' for the applications. Since they couldn't change the programs, they didn't control them.

Larger shops with business oriented systems had long dealt with system software that they had no control over, and they pounced on the proprietary applications since they had source for the GL/AP/AR systems they ran.

That's what your sources talk about, not a lack of standards or actual problems; but a simple lack of control.

In one company I worked for, the first 'official' PC network was provided by a vendor setting up an automated distribution center (warehouse). The IT department was flabbergasted because there was a large chunk of technology that they didn't have experience with.

Change is difficult for anyone to adjust to, and most people resist change. When something directly threatens your livelihood, you fight it any way you can. The most common way is to find lies with just a kernal of truth to them to spread to attempt to keep the change from happening.

-tachyon

alltare
October 21st, 2004, 09:34 PM
In the REAL early days, Altair BASIC, later to become Microsoft BASIC, wasn't free, but I recall that paper tape and cassette versions could be had cheap when bundled with MITS memory boards (and maybe with some other products). Free or not, everyone seemed to have a copy though, and this led to Bill Gates' 1976 "open letter" on piracy. I don't remember how Disk BASIC was marketed.

The S-100 industry eventually shifted away from MITS' hard-sectored floppy discs to the soft-sectoring scheme that's almost universal today. This, coupled with CP/M, caused MITS' OS to become the odd man out. CP/M for the MITS hardware was available, but I don't think it was very popular.

Even MITS' sister company, Icom (under Pertec's roof), made soft-sectored floppy systems including the early "Frugal Floppy". Their BASIC, called DEBBI (Disk Extended Basic By Icom), was free with hardware purchase, if I recall correctly. CP/M could easily run on their systems, too, but neither MITS nor Icom bundled it.

steve
=====================



Hi, folks. I've been doing a lot of reading lately and came across some rather nasty comments about early personal computers from some of the old mainframe hackers. Apparently (this is news to me), there was some harsh resistance to PCs from the oldschool hackers because they shipped with proprietary operating systems. Their argument was that they didn't want the company to have that much control over what they did with their computers.

Did any of the vintage computers ship with a free operating system? I'm more than a little curious if people here have debated the issue and have any thoughts on it. I'm no expert on this topic; thus my quest for info. :-)