[Israel.pm] Selling the Perl Resource Kit

Shlomi Fish shlomif at iglu.org.il
Wed Feb 11 07:30:20 PST 2004


On Wednesday 11 February 2004 17:05, perl-il at amos.mailshell.com wrote:
> Shlomi Fish wrote:
> > minix, a small UNIX-like OS for x86's as an example. minix in turn
> > served as
> >
> >the basis for the first version of Linux which was written by Linus
> > Torvalds.
>
> This isn't quite true, as far as I remember - Linus actually startted
> from scratch. 

No, he didn't. He converted the minix code to run on 386 and gradually rewrote 
all of it. (in a similar fashion to the fact that the BSD guys started from 
AT&T UNIX and gradually rewrote all of it). That way, the Linux kernel 
eventually shared no code with minix, but while the minix code was there, it 
enabled the Linux 0.0.1 parts to actually function.

Still, re-writing an entire codebase like minix, is still quite a feat. I am 
in great awe of Linus Torvalds even for that. (albeit the early versions of 
the Linux kernel were much less functional than the more modern ones).

> The big noise about linux when it came out, as I remember, was that:
> 1. Linux takes advantage of the 386 "real mode" (or whatever it's
> called) - which ment real
> division between "system mode" and "user mode". Minix was designed to
> run on pre-386
> which basically ment it was a toy system. I think it also had a
> "valuntary" scheduler (i.e. a
> program could take control of the system if it miss-behaved).

Right. The 8086/8 does not give way too well to multi-tasking. That's one of 
the most significant additions in the 386.

Note that even the 386 and onwards architecture has a lot of unhappy leftovers 
from 8086. It's a very horrid architecture from my impression. I studied it 
in the excellent course "Micro-computers" in the Technion. In the same 
semester, I studied the course "Software Systems" which teaches about VAX-11 
assembler. VAX-11 assembler is so sweet and symmetrical that you could cry 
out of joy when coding in it. The Pentiums assembler is much more limited and 
the architecture is a complete mess.

> 2. Linux is GPL - Minix was not true "open source" in the current sense,
> that's why I'm pretty
> sure that I remember correctly that Linus couldn't take code from Minix
> (Minix is currently
> in "Public Domain", but its LICENSE was ammended last at 1997, prior to
> that it had
> a stricter copyright, see http://www.cs.vu.nl/pub/minix/LICENSE and
> other stuff at
> http://www.cs.vu.nl/~ast/minix.html and minix.org).
>

Right. Linus Torvalds GPLed the kernel quite quickly after releasing it to the 
public. Minix indeed had a problematic not-entirely-free license. If it had 
been open-source it is possible we would have been using a system called 
"minix" today...

That teaches a lesson that open source prevails, and a core component that 
isn't open-source does not go very far. In any case, Tannenbaum never made 
claim for ownership of the Linux kernel source, but rather criticized 
Torvalds for the decisions he chose to incorporate in the kernel. (monolithic 
kernel, etc.)

Regards,

	Shlomi Fish

---------------------------------------------------------------------
Shlomi Fish      shlomif at iglu.org.il
Homepage:        http://t2.technion.ac.il/~shlomif/

I don't believe in fairies. Oops! A fairy died.
I don't believe in fairies. Oops! Another fairy died.




More information about the Perl mailing list