Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!djfiander
From: djfiander@watmath.UUCP
Newsgroups: comp.os.minix
Subject: Re: First Impressions of Minix
Message-ID: <12904@watmath.UUCP>
Date: Sun, 19-Apr-87 11:34:50 EST
Article-I.D.: watmath.12904
Posted: Sun Apr 19 11:34:50 1987
Date-Received: Sun, 19-Apr-87 18:36:26 EST
References: <288@grebyn.COM> <893@tekfdi.TEK.COM>
Reply-To: djfiander@watmath.UUCP (David Fiander)
Organization: U. of Waterloo, Ontario
Lines: 27

In article <893@tekfdi.TEK.COM> mhorne@honda.fdi.tek.com (Mike Horne) writes:
>Bob, when you add a new device and/or device driver to an existing UNIX
>(or clone thereof) kernal, you have to re-make the kernal.  Usually this
>does NOT entail recompiling the whole OS, but rather just compiling the
>new driver source and the existing object files for the parts of the OS
>that aren't affected by the new driver.  It is a fairly simple process
>(re-making the OS, not writing the driver!).  What you want is a good OS
>which is as simple minded as MesSDOS but powerful enough to sport things
>like multi-tasking, multi-user, mind-reading, etc.  Sorry, Bob, that time
>hasn't quite come yet.

Why does an operating system have to be simple-minded to support dynamically
loaded device drivers?  I have a friend who owns a tandy CoCo running OS-9,
A multi-user, multi-tasking OS running in 64K of physical memory on a 
Motorola 6809 which supports all the standard Unix System calls.  If he wants
to read an MSDos diskette he loads a disk driver which understands that format
(without having to reboot, let alone relink), sticks in the disk, and reads it.
When he is finished, he unloads the driver.  This is the way drivers should work
as far as I'm concerned.

-- 
"Are you police officers?"
	"No ma'am, we're musicians."
- The Blues Brothers

UUCP  : {allegra,ihnp4,decvax,utzoo,clyde}!watmath!djfiander
CSNET : djfiander%watmath@waterloo.CSNET