Path: utzoo!mnetor!uunet!husc6!bloom-beacon!mit-eddie!ll-xn!ames!pasteur!agate!eris!mwm
From: mwm@eris (Mike (My watch has windows) Meyer)
Newsgroups: comp.sys.ibm.pc
Subject: Re: A Nutty Idea
Message-ID: <8711@agate.BERKELEY.EDU>
Date: 14 Apr 88 01:16:16 GMT
References: <231@unsvax.UUCP> <9804@tut.cis.ohio-state.edu> <470@splut.UUCP>
Sender: usenet@agate.BERKELEY.EDU
Reply-To: mwm@eris.UUCP (Mike (My watch has windows) Meyer)
Organization: Missionaria Phonibalonica
Lines: 76
Keywords: volunteer GNU-like PD freeware

In article <470@splut.UUCP> jay@splut.UUCP (Jay Maynard) writes:
<In article <9804@tut.cis.ohio-state.edu> bob@allosaur.cis.ohio-state.edu.UUCP writes:
<>In article <231@unsvax.UUCP> tlhingan@unsvax.uucp (Eugene Tramaglino) writes:
<>>If everyone on the net with the appropriate skills volunteered 5
<>>(five) hours per week of their time, we could write our own OS,
<>>editor, utils, etc, ... I hereby volunteer my 5.  Now, we need a good
<>>co-ordinator...
<>
<>Please send mail to info-gnu@prep.ai.mit.edu.  There's plenty of work
<>for you to do.  GNU won't be PD, but it will be free, as are the
<>utilities that are already available.
<
<Well, almost. The only problem is that RMS holds PCs in contempt, and
<refuses to even consider making GNU work on them.

True - but only if you read PC to mean "a computer that costs
significantly less than a cheap new car" rather than the all-to-common
misreading "things that look like the IBM entries into the personal
computer market." In the latter case, that statement is so much horse
pucky. You even see half of it yourself:

<he will require them to support demand-paged virtual memory.

Not true, but close. What he will require is a large, flat virtual
space. No demand paging necessary - all you need is an MMU. True,
certain obsolete micros (from Intel, Zilog & Moterola) can't support
that. RMS doesn't think it's worth bending his brain and his software
to fit those cpus.

>(Goodbye, GNU/Mac.)
 
Well, only on cheap Macs. A Mac ][ (or an Amiga 2000 + 68020) with an
MMU option and 16 meg of real memory should run the proposed GNU OS
quite well, thank you.

Any of the current generation of '386 boxes would be quite reasonable
to put the proposed GNU OS on. Assuming that saddling them with a
support architecture designed for obsolete Intel processors doesn't
cripple the '386 to badly.

RMS has stated that GNU will almost certainly not run on an '86 or
'286. It won't run on an 8080 or a z80, either - and for the same
reasons. The cpus are inadaquate to support the scale of program that
RMS is working on.

On the other hand, he has stated that GNU could be made to run on a
'386 box. You could probably convince him or some part of the FSF to
do the port by giving them the kind of hardware that other people have
made available - nothing special: a 16MHz or faster '386 box with 4 to 8
meg of memory and 60 to 80 meg of disk would be about the right thing
to consider.

<A better place for your efforts would be in extending Minix; it's not PD
<either, but it's cheap, and you even get a text in OS construction with
<it.

If you want a nice tool for learning OS internals & the like on cheap
hardware, this is true. On the other hand, trying to cram all the
features of 4.3BSD into it and making it run on an 8086 is not
something for the weak of heart to consider (I don't even think it's
something for the sane to consider). But that's what extending Minix
to be what RMS has in mind for GNU would be like.

I support both Minix & GNU. They both serve a purpose - they are just
different. Minix is small and cheap and clones an antiquated OS. But
you can get it now, and run it on hardware that costs less than a used
car I'd trust my life to. GNU is planned as being a modern OS, and
running on hardware that is currently very expensive, but getting
cheaper all the time.

	<mike
--
I went down to the hiring fair,				Mike Meyer
For to sell my labor.					mwm@berkeley.edu
I noticed a maid in the very next row,			ucbvax!mwm
I hoped she'd be my neighbor.				mwm@ucbjade.BITNET