Re: Generic Kernel API

From: Reid Linnemann <lreid_at_cs.okstate.edu>
Date: Thu, 10 Nov 2005 13:46:12 -0600
Marcin Jessa wrote:

> 
> Sure, but the point is to use native FreeBSD drivers, even if they were
> in closed source binary form and not drivers written for an entirely
> different O.S.
> 
> Marcin

I'd like to pose my own semi-educated opinion about this topic:

If hardware vendors are given the ability to provide closed-source, 
unfree-licensed binary drivers for hardware, they will probably gladly 
do so. They get the bonus of sales to FreeBSD users without having to 
give up control or knowledge of their products. I can see the potential 
that with those benefits in mind, eventually no vendors would opt to 
support the open implementations of drivers for their products, instead 
placating users only with binary drivers which are subject to 
non-berkeley licensing and that - being that no source is provided to 
FreeBSD - will be unsupportable by the FreeBSD project itself. Users 
could be inandated with license agreements and use policies, and their 
FreeBSD systems be held at the mercy of miserable 3rd party support.

Maybe this isn't even a remote possibility, I'm not involved in release 
engineering at all and I definitely am not a soure of authority on this 
issue, but the mere thought of this situation makes me shudder. I just 
do not want to see FreeBSD no longer be free (as in speech) through a 
dependency on restricted binary drivers. I'm not saying kernel 
interfaces shouldn't be stabilized to promote compatibility, but I am 
saying that caution should be used lest FreeBSD eventually be mangled 
and twisted into 'EULABSD'.

Anyway, that's my half-educated opinion. If it's wrong, fantastic! I can 
sleep better at night.

-- 
Reid Linnemann
Senior Systems Analyst
Oklahoma Department of CareerTech
405-743-5422
rlinn_at_okcareertech.org

-Ars longa, vita brevis-
Received on Thu Nov 10 2005 - 18:46:26 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:47 UTC