LWN.net Logo

Pedigree chart wishful thinking

Pedigree chart wishful thinking

Posted Jan 29, 2003 18:49 UTC (Wed) by nas (subscriber, #17)
In reply to: SCOsource and Linux by rfunk
Parent article: SCOsource and Linux

Not only is the link between Minux and Linux dubious, I think the link between Minux and the original Unix code is also questionable. AFAIK, Minux was a reimplementation of a Unix-like system and was not derived from an existing system. Andrew Tanenbaum used it for teaching students. Any old-timers out there to confirm?

Hmm, I wonder if SCO got permission to use Éric Lévénez's chart? It looks like they took it and highlighted the branches that lead to Unix. Nice research work boys. If they didn't get permission, I think they have their own copyright problems. :-)


(Log in to post comments)

Pedigree chart wishful thinking

Posted Jan 29, 2003 21:00 UTC (Wed) by emkey (guest, #144) [Link]

Thats certainly my recollection. The only reason its distribution was restricted was because of the agreement Tanenbaum had signed with... Prentice Hall I believe who distributed the book. Even then you could give copies to other under some circumstances. If he'd release Minix to the public domain back then we likely wouldn't ever have heard of Linus.

Pedigree chart wishful thinking

Posted Feb 5, 2003 11:14 UTC (Wed) by Soruk (subscriber, #2722) [Link]

AST now has - since he got permission from Prentice-Hall to do so, Minix has been relicensed under a BSD-like licence.

I haven't used it in a number of years, but it's still the only freely-available UNIX-like OS thast actually works for sub-386 systems.

Pedigree chart wishful thinking

Posted Jan 29, 2003 23:41 UTC (Wed) by da4089 (subscriber, #1195) [Link]

i imagine the issue is patents, not copyright, in which case it is the design of features (and possibly their expresion via system calls) that might at issue? for instance, i think DMR has a patent on the setuid mechanism.

as i recall, AndyT's stated goal was to make Minix AT&T 7th Ed compatible at the programming level.

as an aside, i would have thought that Minix was least likely to have IP problems, being implemented as a set of cooperating servers, rather than a monolithic kernel.

Orginal Unix patents should have expired

Posted Jan 29, 2003 23:59 UTC (Wed) by nas (subscriber, #17) [Link]

It looks like the setuid patent has expired. I still think SCO is silly to imply that IP flowed from Unix to Minux to Linux.

Pedigree chart wishful thinking

Posted Jan 30, 2003 7:57 UTC (Thu) by eru (subscriber, #2753) [Link]

> AFAIK, Minux was a reimplementation of a Unix-like system and was not
> derived from an existing system.

Minix is not just a re-implementation, but internally a very different
design: the kernel consists of a set of processes that pass messages
between them. By contrast, Linux uses the same monolithic kernel approach
as the original Unix. The only thing that Minix shares with Unix is the
API. The original Minix version implemented only the old features
that 7. Edition Unix had already in the 1970's.

> Andrew Tanenbaum used it for teaching students. Any old-timers out there
> to confirm?

Yes, and he published a text book on operating systems, using Minix as
his concrete example. The book actually lists the entire Minix kernel.
It ran on a PC XT or compatible: even a hard disk was not necessary
in the original version. I installed it on mine soon after it came out.
A great learning experience.

Copyright © 2008, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds