Quoting Jim Cheetham <[EMAIL PROTECTED]>: > Doesn't anyone know their history round here? :-) > > Dr Tanenbaum teaches Operating Systems amongst his other interesting > computing subjects :-) In order to teach Unix, he needed the sources; > these were not available for his students to experiment with. > > So, he invented Minix - a Unix-like operating system designed to be > valuable in a teaching situation. A copy is included with his > accompanying book (probably should try to get that one day), and I > think > that is (certainly was) the only vector. > > Minix is a microkernel architecture ... so it's similar to The Hurd.
Robert Switzer in his book "Operating Systems: A Practical Approach" discusses some things he wishes Dr Tanenbaum had gotten right with Minix. One of those was an inconsistent application of the idea of independent servers - the File System in Minix is an independent server, and one other thing, but the rest of the code is monolithic. > > Linus took an OS class elsewhere, and wanted to play with Minix - but > couldn't get (or perhaps afford) a copy of the book. So he started his > own kernel from scratch, which grew in popularity. He could afford a copy of Minix, and the book was apparently part of his OS course. It was one of the things he learnt quite a bit from - some of Linux 0.0.1 looks like Minix 1.0, some of it looks altered. He says his knowledge came from applying all the independent patches needed to compile Minix-386. The LINUX IS OBSOLETE thread had some strong words of his about the difficulties of doing something useful with Minix on the 386. So he used it as the vehicle to write Linux, writing Linux in gcc-based assembler. > > However, Linux is "just" a reimplementation of the monolithic kernel > style present at the time Linus was learning. Andy thinks it's a shame > that all the energy that's going into Linux isn't going into anything > truly innovative, in kernel terms. I think Andy's got something there, but on the other hand, Linux _is_ advancing the OS kernel. Just think how stagnant OS studies would be if we were restricted to studying a blue-screening semi-micro-kernel under terms that prevented us from doing anything useful. > > The Wikipedia article is a little light - I suspect the authors were > only interested in him where his interests overlap with Linux. There's That's a shame. > not much about his other acedemic work. > http://en.wikipedia.org/wiki/Andrew_Tanenbaum They should get Woodhall to rewrite it then. He's Tanenbaum's collaborator in the second and third editions. > > -jim > > On Thu, Feb 16, 2006 at 09:09:47PM +1300, Zane Gilmore wrote: > > Wasn't Andrew Tanenbaum one of Linus's operating system theory > professors? > > > > I seem to remember him making comments about not giving Linus > > "full marks" for Linux because he used a monolithic kernel structure. > > > > It was around the time that that guy was tring to claim that Linus > could > > not possibly have written Linux. > > > > > > > > > > Jim Cheetham wrote: > > >On Tue, Feb 14, 2006 at 02:06:30PM +1300, Carl Cerecke wrote: > > > > > >>Andrew Tanenbaum is a Big Name in the research field of Operating > Systems. > > > > > > > > >But not necessarily in the field of Linux :-) > > > > > >-jim > > > > > > > > > > "Sharpened hands are happy hands. "Brim the tinfall with mirthful bands" - A Deepness in the Sky, Vernor Vinge "I me. Shape middled me. I would come out into hot!" I from the spicy that day was overcasked mockingly - it's a symbol of the other horizon. - emacs : meta x dissociated-press
