Date: Sun, 20 Mar 2005 10:25:59 +0100 From: Paul de Weerd <weerd@weirdnet.nl> To: Charles Swiger <cswiger@mac.com> Cc: Theo de Raadt <deraadt@cvs.openbsd.org> Subject: Re: Adaptec AAC raid support Message-ID: <20050320092559.GA1565@ghoul.weirdnet.nl> In-Reply-To: <fa59f75d5bf8592efdd54e49bc51d149@mac.com> References: <200503192050.j2JKoesb003713@cvs.openbsd.org> <fa59f75d5bf8592efdd54e49bc51d149@mac.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--tKW2IUtsqtDRztdT
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Sat, Mar 19, 2005 at 04:29:59PM -0500, Charles Swiger wrote:
| > 3) by not insisting at all that vendors open things at least a
| > bit, Scott is not like Bill Paul or others who have opened
| > up a lot of hardware, but is a lot more like Sam Leffler who
| > has perpetuated this (and today, FreeBSD has one 802.11g/a
| > driver -- and it uses binary bits).
|=20
| Yes, well, I prefer the former approach myself, but I am not going to=20
| complain that Sam has written a wireless driver using binary firmware=20
| rather than one that is completely open. I appreciate the work he's=20
| done, even if I would like to see a completely open series of wireless=20
| drivers.
There's a problem with this approach.
Vendors will see the efforts of developers using binary-only stuff.
And it sends a message : These guys can use our stuff, they will buy
our products, and we do not have to give out documentation.
This makes it harder on developers and on users than need be. But
since the vendor has seen that what they've done so far (releasing
binary-only stuff, eg a linux-only program that can run on your BSD
system via linux-emu (but only if you run on i386)) works for us, why
should they then supply more information/documentation ? What is in it
for them ? It works, doesn't it ? Why give the vendor such a hard
time, they did their job, see this-and-that project can work it out,
why can't you ?
It's just like Windows. You buy hardware, run Windows, but have no
idea what's going on. The vendor supplies the driver and all is well.
However, we have an open source operating system. We can see the
internals of the system, we can go in ourselves and fix bugs. We can
take the code and port it to other architectures, port it to other
operating systems, do whatever we want. This is what we want to do. We
do not want to be tied into software we can not examine ourselves.
Otherwise, why run BSD ? Why run Linux ? There's Windows for you, it
comes with drivers so you do not have to write the code yourself, so
you can not be bothered to read the code, to take it and port it or
fix its bugs or adapt it to suit your needs.
I decided long ago to use open source software. The longer I use it,=20
the more I value the freedom open source software gives me. I
therefore appreciate open sourced drivers. And I appreciate the time
it takes the developers of my operating system to ask vendors for
documentation, then take that documentation and use it to write those
drivers. What value is there in trying to support a vendor that is
unwilling to share the documentation the developers need to write
drivers ? They don't support us, why should we support them ?
It's sad that I spend money on hardware I later find is not supported.
And of course, I would like to use this hardware rather sooner than
later. But I would prefer that support to be open source. If that's
not possible, then I'll just go support another vendor who truly
supports open source. I'll toss out the unsupported hardware, tell my
friends, family and co-workers not to buy stuff from that company and
endorse the vendor that is willing to open up their stuff.
This grass roots approach has turned out to be pretty succesful in the
OpenBSD world. We now have a whole lot of drivers for wireless cards
where we were unable to use most cards a year ago. Vendors have
learned that they can make more sales if they open up their
documentation so they are happy. The wireless card I bought last year
now works, so I am happy. Others can take the code and make it work on
their systems (hardware architecture/operating system) so they can be
happy. I feel that this is in large parts due to Theo's approach (and
the other developers) to this issue, so I thank them for it.
At least now I know what RAID controller not to buy.
Cheers,
Paul 'WEiRD' de Weerd
PS: that "completely open series of wireless drivers" you talked about
is now available at your local OpenBSD mirror. Feel free to take it
and port it to your prefered system.
--=20
>++++++++[<++++++++++>-]<+++++++.>+++[<------>-]<.>+++[<+
+++++++++++>-]<.>++[<------------>-]<+.--------------.[-]
http://www.weirdnet.nl/ =20
--tKW2IUtsqtDRztdT
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (OpenBSD)
iD8DBQFCPUGmmw12l2HFcK0RAii4AJ4utZE5fFOMz/mR6fYWVWtUVgtqpgCggLje
elPmXoZqv70z9wXRUyNkCeY=
=yvIU
-----END PGP SIGNATURE-----
--tKW2IUtsqtDRztdT--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050320092559.GA1565>