Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!seismo!mcvax!botter!ark!polder
From: polder@ark.UUCP
Newsgroups: comp.os.minix
Subject: Re: Re: portablity bug in mkfs.
Message-ID: <973@ark.cs.vu.nl>
Date: Fri, 10-Apr-87 20:17:56 EST
Article-I.D.: ark.973
Posted: Fri Apr 10 20:17:56 1987
Date-Received: Sat, 11-Apr-87 21:22:03 EST
Reply-To: polder@cs.vu.nl (Paul Polderman)
Distribution: world
Organization: Free University, Amsterdam
Lines: 19

In article <515@ark.cs.vu.nl>, levy@ttrdc.UUCP (Daniel R. Levy) writes:

> In article <956@ark.cs.vu.nl>, polder@ark.cs.vu.nl (Paul Polderman) writes:
> > When mkfs is compiled on a system where sizeof(int) is not
> > 2 bytes, (e.g. a VAX) mkfs will build a wrong filesystem
> > when using indirect blocks.

> Can someone please explain to me why one would try to compile MINIX source on
> a VAX.  Just curious.

It is often convenient (and faster!) to develop user (not kernel) software
for MINIX on another system.
If the include files are on-line on your local system (as they are here,
MINIX was developed here), you can use all the defines used in MINIX.
The system-calls of MINIX are also available on your local UNIX-system.
So there is nothing to prevent you from writing software for MINIX
on your local system.

- Paul Polderman (polder@cs.vu.nl)