Re: why NetBSD include binary/blob driver or firmware by default?

[David Young <dyoung%pobox.com@localhost>]

On Fri, Jan 23, 2009 at 10:53:14PM +0100, Joerg Sonnenberger wrote:
> On Fri, Jan 23, 2009 at 11:26:31PM +0200, Cem Kayali wrote:
> > While searching about this issue, i noticed that NetBSD, too, includes  
> > drivers and firmwares without source code, most or all has suffix _uu in  
> > src tee.
> 
> The following is not an official statement of TNF, but my own position.
> 
> The only driver that is not open source should be ath(4) and that is in
> the process of being replaced. This does not count drivers that ship
> binary firmware images. Contrary to what the Free Software Foundation
> wants to make you believe, it is completely acceptable to have a
> binary-only firmware as long as the license allows redistribution and
> use for any purpose.

Whether closed-source firmware is acceptable or not depends on
one's purpose.

If one's purpose is to innovate in wireless networking, one may
benefit by modifying WLAN firmware's source code, or by rewriting
it from scratch.  Running one's own code on a microcontroller that
is tightly coupled to the MAC and PHY creates innovation opportunities.

If the purpose is to support microcontroller-based SCSI adapters
on space-constrained installation media, one may benefit by removing
non-essential features from the source and compiling a slimmer
version of the firmware.

If one wants to fix bugs in the firmware, source code may be
essential.

If you desire only to distribute an operating system that allows
redistribution and use for any purpose, then source code is probably
not essential.

> Source code for the firmware images would often not be very
> useful, as they require very special tools for processing.

Sure.  But I think that oftentimes the processing tool is no more
special than gcc.

Dave

-- 
David Young             OJC Technologies
dyoung%ojctech.com@localhost      Urbana, IL * (217) 278-3933