Return to BSD News archive
Path: sserve!manuel!munnari.oz.au!uunet!cs.utexas.edu!sdd.hp.com!swrinde!elroy.jpl.nasa.gov!ames!agate!agate!phr
From: phr@soda.berkeley.edu (Paul Rubin)
Newsgroups: comp.unix.bsd
Subject: Re: Restrictions on 'free' UNIX / 386BSD (Re: selling 386BSD)
Message-ID: <PHR.92Aug17110648@soda.berkeley.edu>
Date: 17 Aug 92 16:06:48 GMT
References: <PHR.92Aug15214245@soda.berkeley.edu> <YSDIBS4@taronga.com>
<PHR.92Aug16224207@soda.berkeley.edu> <WLEIPPE@taronga.com>
Organization: CSUA/UCB
Lines: 23
NNTP-Posting-Host: soda.berkeley.edu
In-reply-to: peter@taronga.com's message of 17 Aug 92 10:54:59 GMT
> This sort of distribution is not practical for an operating system.
>Why is it practical for compilers and not OS's?
Because the compiler distribution is subsidised by the OS distribution. If
the OS itself has no protection against unsupported copying, what is there
to subsidise it?
Hardware distribution.
And for clonable machines (Sparc, MIPS, x86) the O/S has to be a
profit center in its own right.
The x86 proves this is totally wrong. Nearly all x86 vendors just
ship ms-dog on their machines with almost no profit---they are in the
hardware business and are perfectly happy to send you a box with -no-
OS if you ask for it. We already see from Cygnus that companies can
profit from distributing and supporting free software. Obviously
enough people are willing to pay for support to keep such companies
in the black. Conversely, if not enough people want the support,
the support must not be so important, and then there's not much excuse
for a proprietary OS.