Prove DOS "rips off" CP/M code, win $200,000!
Apparently researcher Bob Zeidman, who has previously investigated alleged similarities between MS-DOS (AKA 86-DOS and IBM PC-DOS), is now offering a $200,000 reward to anyone that can prove code was directly copied:
http://thenextweb.com/insider/2016/08/0 ... urce-code/
However, the whole idea that MS-DOS/86-DOS was "copied" directly from CP/M is insane in about a dozen different ways....
---
Imagine for a minute that you are sitting in front of your companies' brand new 16-bit 8086 based S-100 card. But there is no software at all for it yet. You can't really sell your hardware without software.
You obtain a port of ROM Disk BASIC from this clown down the street, but you still have no OS. And CP/M currently only supports the 8-bit 8080/Z80. So what do you do? You start hacking out your own quick and dirty OS!
You are familiar with writing CP/M programs. You are familiar with what CP/M expects, how it works externally, and you have CP/M technical documentation at your disposal. And you want to make it easy for you and others to port existing CP/M applications to your new OS.
Perhaps you have even seen some CP/M code. But even if you did, that 8-bit 8080 code is not at all useful on your 16-bit 8086. It is quite different, otherwise CP/M would already run on it!
So you start hacking away, implementing somewhat similar function calls, trying to make it give similar results to CP/M, but making a few improvements such as using a File Allocation Table file system similar to that of your ROM Disk BASIC. You implement a similar Basic Input/Output System (printer, serial, keyboard, terminal interface code) specific to your hardware - something that each OEM like your company is required to do with CP/M since there is no hardware standard. As far as you are concerned, everything you wrote you wrote yourself.
Oh, and that clown you licensed BASIC from just showed up and thinks your new OS is interesting.
So you polish up your new "86-DOS 1.0" and before you know it, it turns out it is Big Blue IBM themselves that wants to use your new OS. But it is that clown down the street from Microsoft who will license it to IBM and eventually a bunch of other companies that want it.
And now Gary Killdall is pulling out his hair yelling "you ripped me off!!!".
---
Now fast forward to the far off year 2016, and everyone has all kinds of crazy ideas about what might have happened.
Apparently someone thinks they saw some hidden code that would print a Digital Research copyright banner. If anyone really thinks that, they were tricked. Like typing in some mysterious hex code in Debug and running it. Where could one even hide that in a few thousand K of binary code? One could not just copy Z-80 code in to an 8088/8086, it would have been translated and rewritten. Which, even still, means every byte would have been mostly understood.
Did Tim Patterson at Seattle Computer Products or anyone else that worked on 86-DOS follow sufficient clean room procedure? Possibly not, as such procedure did not really exist yet. Were there similarities? Duh, yes. Did they directly copy any code? No.
Well, if you think you can prove otherwise perhaps you can win $200,000.
http://thenextweb.com/insider/2016/08/0 ... urce-code/
However, the whole idea that MS-DOS/86-DOS was "copied" directly from CP/M is insane in about a dozen different ways....
---
Imagine for a minute that you are sitting in front of your companies' brand new 16-bit 8086 based S-100 card. But there is no software at all for it yet. You can't really sell your hardware without software.
You obtain a port of ROM Disk BASIC from this clown down the street, but you still have no OS. And CP/M currently only supports the 8-bit 8080/Z80. So what do you do? You start hacking out your own quick and dirty OS!
You are familiar with writing CP/M programs. You are familiar with what CP/M expects, how it works externally, and you have CP/M technical documentation at your disposal. And you want to make it easy for you and others to port existing CP/M applications to your new OS.
Perhaps you have even seen some CP/M code. But even if you did, that 8-bit 8080 code is not at all useful on your 16-bit 8086. It is quite different, otherwise CP/M would already run on it!
So you start hacking away, implementing somewhat similar function calls, trying to make it give similar results to CP/M, but making a few improvements such as using a File Allocation Table file system similar to that of your ROM Disk BASIC. You implement a similar Basic Input/Output System (printer, serial, keyboard, terminal interface code) specific to your hardware - something that each OEM like your company is required to do with CP/M since there is no hardware standard. As far as you are concerned, everything you wrote you wrote yourself.
Oh, and that clown you licensed BASIC from just showed up and thinks your new OS is interesting.
So you polish up your new "86-DOS 1.0" and before you know it, it turns out it is Big Blue IBM themselves that wants to use your new OS. But it is that clown down the street from Microsoft who will license it to IBM and eventually a bunch of other companies that want it.
And now Gary Killdall is pulling out his hair yelling "you ripped me off!!!".
---
Now fast forward to the far off year 2016, and everyone has all kinds of crazy ideas about what might have happened.
Apparently someone thinks they saw some hidden code that would print a Digital Research copyright banner. If anyone really thinks that, they were tricked. Like typing in some mysterious hex code in Debug and running it. Where could one even hide that in a few thousand K of binary code? One could not just copy Z-80 code in to an 8088/8086, it would have been translated and rewritten. Which, even still, means every byte would have been mostly understood.
Did Tim Patterson at Seattle Computer Products or anyone else that worked on 86-DOS follow sufficient clean room procedure? Possibly not, as such procedure did not really exist yet. Were there similarities? Duh, yes. Did they directly copy any code? No.
Well, if you think you can prove otherwise perhaps you can win $200,000.
Comments
Tell that to most coders especially the object oriented and watch the mixed reactions.
CP/M was originally written PL/M. Later releases replaced in places with ASM. The earliest BDOS source saved online was 9k of PL/M which by 2.2 became 76k of assembler in two files. About half the complete CP/M 2.2 source code (including utilites) was written in PL/M. PL/M compilers were only from DRI and it took DRI a long time to get an 8086 release of it. PL/M source code would have been useless to Tim Paterson.
The original charge was that QDOS was just a conversion of CP/M. QDOS includes a utility to convert 8080 to 8086*, Converted code was very distinctive with inefficient structures no human programmer ever would use. Mechanical conversion would have been obvious.
$200,000 should cause any proof of wrongdoing to turn up or finally end all the rumors. Several legal clones of CP/M were done. I doubt TIm Paterson was such a poor programmer that only he could not manage to create a legal clone.
* Wordstar 3.3 was the most famous victim of mechanical conversion. Got to DOS early but then spent 4 years only able to do a minor foreign language update because the code became very difficult to work with.