MINIX operating system

December 25, 2007 – 5:51 pm by John

When I was looking for blags to add to our blagroll, I came across, Tim Swanson's web page. Even if I hadn't heard of Tim Swanson, I would have added him without hesitation because of his libertarianism and obvious fixation on "The Simpsons." He could hate baseball and beer and still be a fantastic guy just because of those two qualifications.

Tim Swanson is an economist, I guess, who teaches something in Korea and/or Texas A & M and writes semi-regularly for the Ludwig von Mises Institute. (I hope to write about his latest column soon.)

He is a fantastically prolific blagger, having maintained his own blag for at least four and half years. One of his post categories is Open Source, so I clicked on that to see what he'd written on those topics. The first post had an ad underneath it saying, "MINIX 3 is a new reliable free operating system. Smaller than Linux. Try it. It's free!"

MINIX is Andy Tanenbaum's operating system that he invented in 1987 for his computer science class at a university in The Netherlands. He invented it to be a stable, secure, fast, lightweight microkernel operating system. Even though it wasn't designed for commercial sale, it was widely known in CS circles. He had a rather public debate about Linux with Linus Torvalds after the latter published Linux because Linus wrote a monolithic kernel, something Tanenbaum disapproved of and found very...backwards.

This is a fascinating story of the history of UNIX and its offshoots and clones, how some people tried to smear Linus and his OS, and how many people are still mistaken about Andy and Linus's relationship (they have no problem with each other).

O'Reilly compiled a "transcript" of the entire Tanenbaum–Torvalds debate, which was "fought" via an internet newsgroup in 1992, a month after Linux was published. This discussion, which I've read before, contains many more people's input than just Andy and Linus. I don't really get it, except the vague idea of what a microkernel does differently from a monolithic kernel.

I found Tanenbaum's thoughts on microkernels, monolithic kernels, and the future of OSes much more accessible. I was and still am very intrigued and attracted by the whole goal of the "reliable OS," which is: if one process stalls it won't affect any others, things run lighter (less space) and faster, and there will be no "reset" buttons on computers produced in the future because they aren't likely to freeze or crash. On the other hand, it is a little unsettling to think about Microsoft regaining a lot of its credibility and even becoming favored again amongst geeks if its Singularity OS is successful...

Bookmark and Share

Trackback URL for this entry is:

Sorry, comments for this entry are closed at this time.