Andrew S. Tanenbaum: The Impact of MINIX 

He explains why microkernel is better. Lack of this architecture is a severe weakness in Linux.

@@ijoyner I don't believe he is biased, I think microkernel architecture has a point against monolithic ones.

Linus was his student.

His _Computer Networks_ book is the one I value most. I have the 2nd, 3rd and 4th editions, and they represent a series of snapshots of what was important at various stages of the art over the years. He should probably give up on microkernels by now.

@@lawrencedoliveiro9104 What do you say now that Google is openly developing Fuchsia, an operating system based on a microkernel architecture, named Zircon?

OS of no OS :)

It's never to late.

David Sebastian I figure he has to be comical as hell.

Ehhh. Someone else would have come up with similar shi@

Why? He wanted to educate people, not make money.

Olaf Klischat Why should he bash himself? MINIX was strictly designed as a teaching tool at a time when the latest UNIX systems could not be used for such purposes. Not placing MINIX under GPL or one of its compatible licenses was a virtue since it allowed the codebase to be kept under control so it could be useful to students. The problem for Tanenbaum was there wasnt an affordable UNIX system at the time. MINIX - being a low cost UNIX-like system attracted a lot of attention for its affordability which meant that people came to it wanting a full UNIX system, and got the ugly end of the stick when they realized that was never the purpose of MINIX, and Tanenbaum rejected a lot of suggestions for the sake of keeping the code simple and able to run on machines his students could afford. He actually refers to this in the Linux is obsolete debate when he says Linux will keep people who who BSD UNIX away from MINIX.

@@ScottThorpe Why? Because now he wants to make money with Minix. Clearly he made it open source too late and now in his own words "is trying to appeal to a niche market". He made a huge mistake not open sourcing Minix when he had a chance and he will die regretting it.

@@onufrybonekip3717 meanwhile MINIX runs inside of quite literally every Intel processor made in the last few years :P I don't think he regrets anything.

Minix wasn't nearly as locked down as other systems, as the source code was available for the price of a book. Of course it was way before the open source movement that he was operating in, and he's seen that BSD licences have a lot of good to them, which is why he changed his mind. I do think Linux is bloated, especially for things like smart TV's where you almost certainly don't need floppy disk and 40 different file systems support in the kernel itself. For android phones, sure, but they have gigs of RAM and a half dozen cores.

No, just a member of a tribe that claim to be chosen by their god.

Xerox mostly ignored graphical interfaces with icons, windows, and mice, even when they had it right in front of them. See Xerox Alto.

The management ignored and gave away what was being done at Xerox PARC. They showed Apple (who were doing similar work, but not so ambitious), and Jobs said "What are you guys doing here, you could rule the world with this technology". Most of the Xerox PARC people left Xerox to go down the road to Apple to actually build products.

He learnt so much from writing his own OS. Since he put every line of code in it himself, he can teach it the best.

BSD was still under the AT&T license at the time Tanenbaum wrote MINIX (1987). "Until then, all versions of BSD incorporated proprietary AT&T Unix code and were, therefore, subject to an AT&T software license. Source code licenses had become very expensive and several outside parties had expressed interest in a separate release of the networking code, which had been developed entirely outside AT&T and would not be subject to the licensing requirement. This led to Networking Release 1 (Net/1), which was made available to non-licensees of AT&T code and was freely redistributable under the terms of the BSD license. It was released in June 1989." en.wikipedia.org/wiki/History_of_the_Berkeley_Software_Distribution#4.3BSD

Because BSD are monolithic kernels.

There was no freely available BSD until 1992 (386BSD, the basis of all current BSDs today).

Tanenbaum wanted to make MINIX free software, but Prentice-Hall refused to allow it. It isn't fair to say that this was done for his own profit; it was done because the publisher refused to have it any other way.

​@@kovoc8877 He could have marketed his book without the software and published the software independently. It is normal that book companies protect their IP.

@@michaelkaercher The source code of MINIX is integral to how 'Operating Systems: Design and Implementation' is structured; I have doubts as to if Prentice-Hall would have been willing to allow Tanenbaum to retain his copyright on it. Regardless, the idea of Tanenbaum and VU Amsterdam being able to supply the massive demand that MINIX elicited seems unlikely. I suppose that he could have published MINIX on USENET, but that would not have succeeded in getting MINIX in the hands of everyone who wanted it. I think that Tanenbaum made the right choice--and that this choice was not motivated by profit. His goal was to get MINIX in the hands of as many students as possible, and I think he accomplished that goal in spite of the circumstances.

Nope, BSD was licensed to only some US universities. It wasn't freely available.

The first open source BSD (386BSD) didn't exist until 1992 - a whole year after Linus already started writing Linux.

His brain is overclocked.

@@17plus9 he is the man

I can understand him perfectly at 2x speed.

Dejan Lekic The problem was that GNU had been in development for about 8 years and Hurd still did not show any signs of materialising (RMS spent three years thinking that Mach could be used and Hurd did not boot until 1994 -by which Linux was starting to become a significant commercial force) and there was the AT&T lawsuit with FreeBSD. Linux happened to be at the right place at the right time and rightfully won where other OS's had failed.

I agree. Something else would have come up

No, Android TV and the micro kernel minix are totally different things

Hey man, you might be at the wrong adres to post your comment. I believe you might have the MINIX 3 operating system and the mini computer mixed up. They're both veeeeeery different things :)

Linux isnt