Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!mcvax!botter!star!ast
From: ast@cs.vu.nl (Andy Tanenbaum)
Newsgroups: comp.os.minix
Subject: Re: HD problem finally fixed! and HD boot question
Message-ID: <485@ast.cs.vu.nl>
Date: Wed, 26-Aug-87 05:00:56 EDT
Article-I.D.: ast.485
Posted: Wed Aug 26 05:00:56 1987
Date-Received: Fri, 28-Aug-87 00:45:36 EDT
References: <4240@ncoast.UUCP>
Reply-To: ast@cs.vu.nl ()
Organization: VU Informatica, Amsterdam
Lines: 54

In article <4240@ncoast.UUCP> allbery@ncoast.UUCP (Brandon S. Allbery) writes:
Wwhy are the following not in the 1.2 commands package?
>
>	cal	cdiff	ed	expr	test
>
Cal.c, expr.c, and test.c are in the 1.2 distribution.  I found and
fixed the bug in cal.c.  Ed is not there because it is large and I didn't
have time to test it out thoroughly yet.  It will be in 1.3 (sigh).  I
didn't include cdiff because (1) there were various fixes floating around
from various people, and I sort of lost track of the "definitive" version.
If someone will send that to me, I can include it in 1.3.  That incidentally,
won't be for many moons.  I have other fish to fry in the meantime.

> What, officially, is the status of "patch"? 
I didn't include it because it is so large.  I have been trying hard to
minimize the number of distribution diskettes, because the price P-H has
to charge is related to their manufacturing cost.  Also, patch belongs to
Larry Wall, so I have to discuss that with him at some point.

> I still have plans to redo the tty stuff.  It occurs to me that the changes
>to FS as proposed in an earlier fix may be a good idea; the current setup is
>stylistically nice but worthless as a useful construct.  
Please see if you can make a tty driver that only changes the tty.c file and
closely related files (klib88.s, mpx88.s).  The previously posted versions
have (1) had bugs and (2) required a major overhaul to the whole system,
which I haven't had the time to do.  (I have this fear that if I change
10 files at once, obscure bugs will crop up and I will have no idea where to
look.)
As to the splitting of FS into a separate module, it is very useful for
portability.  The Atari port, for example, changed virtually nothing in
FS.  If FS were integrated with the kernel, the port would have been
MUCH harder.  As to the performance of message passing, I have been using
a Zenith Z-248 AT clone.  I can now compile the "Hello world" program
in 10 seconds.  I am not displeased.  I'll post more statistics later.


>Recently, Andy Tanenbaum posted a patch to Minix to load the root filesystem
>from the hard disk.  This is nice, but I'd like to know if there will ever
>be an official way to have the root filesystem be ON the hard disk.
Probably not an "official" version.  I never really had problems on a PC
with 640K, and with 512K AT and extended memory, it really flies.

[Re: bootstrap loaded].  I am not especially interested, but if you are,
by all means do it and post it.  One of the advantages of booting MINIX from
floppy is that you can have an active hard disk partition for DOS, so that
to boot DOS, you don't use the floppy, and to boot MINIX, you do. Personally
I find that easy enough.

I have recently included soft scrolling in tty.c for the benefit of
EGA users with nonstandard cards (the regular 1.2 version works fine on
IBM EGA cards and true compatibles).  I'll post the new tty.c/klib88.s
when I have time.  (Having time is a recurrent theme with me).

Andy Tanenbaum (ast@cs.vu.nl)