In article <9110072...@gandalf.informatik.rwth-aachen.de> mich...@gandalf.informatik.rwth-aachen.de (Michael Haardt) writes:
> If it becomes a stand-alone system, I will quit MINIXing.
Me too ;-)
torva...@klaava.Helsinki.FI (Linus Benedict Torvalds) writes:
>> I can (well, almost) hear you asking yourselves "why?".
I have been thinking of doing it (and partially actually
did), but had no guts to do the whole work ...
> There is no need to ask "why?". Sure, MINIX 386 might by
> nice, if only there would be official support.
Official support means you have
1) Product
2) Team to support it
Case 1 is easy, but for the case 2 I can not say the same ...
They have to eat too ;-) i.e. work hard and get some
salary too. Then they need shops etc. to sell the
product/support.
I don't need any "official support" .. the magical
words mean ugly things to me. I have SEEN Sun's
official support ;-)
> At the moment, you have to ftp a lot to get a 386
> system working with all needed utilities and nobody knows when
> there will be the next offical version.
Network anarchy is the best official support.
There certainly will be many unofficial good patches
to Linux at nic.funet.fi. I try to keep them in
good order.
> Most MINIX systems are in a twilight zone, somewhere between
> 1.5 with heavy patches and 1.6.16 with patches.
Yes, but that's because they are "officially
suppiorted" patches .. NOT the whole sources.
I try to keep also whole patched sources available.
I do remember how I personally spent about 20 hours
to patch 1.2 to 1.3d - but most of it was fooling
around ftp bug (had to split it, and again split the
splitted parts etc. after some real hacking ...).
> The result is a mix of POSIX and system 7 and heaven know what
> else. Not that I am not happy with that MINIX, but there is no
> announcement which promises to set a new standard.
There are/were many things which cause/caused this.
Andy didn't like to have "bignix", wanted to keep the
systems first plain PC/XT (8088) level and as floppy
system.
I have been using Minix as a teaching too. But but but,
it's really hard/unusable for that ! It takes too much
time to fool around the kernel for a student, like
if the work is to do simple driver to kernel .. the
poor student has to make changes to million places around
header/source files. And when another poor student
makes a simple driver .. eh eh .. and you want to
have one kernel with both drivers in it ...
Some of the best are available at cs.hut.fi:pub/minix
Also making Minix more commercial is not good thing;
people like I don't want to make any commercial
product better for other people to buy it. Nothing
against dealing or getting money out of software,
but others to gain my coding need to buy the product.
> Without PH copyright, one can just give sources to someone
> else for explaining a feature or discussing a new
> implementation, without these diffs and diffs to diffs and
> diffs to ...
You need always diffs to some level ;-)
I wouldn't read the whole sources to find out how
a feature is implemented.
> Summary: It depends on your ability to ftp, on the archive you
> use and on your personal capabilities, which system you will
> get.
Linux is now certainly for hackers only.
> What about setting up a mailing-list?
Now we have mailing list for Linux, it's:
Linux-activi...@niksula.hut.fi
And please, DO REMEMBER, REQUESTS to:
Linux-activists-requ...@niksula.hut.fi
Otherwise I will @@%$#@$#!$ you ;-)
Net etiquette is to send requests to mailing-list-request
address, otherwise thousands of people on mailing list
will read your "please add me" etc. postings.
> If only enough people work on the library, it should be
> complete soon.
I think libraries are ready in couple of months.
But help is welcome !
> Michael
arl