29 Apr
2004
29 Apr
'04
5:57 a.m.
[I'm putting the users@ and developers(a)lists.agnula.org on Cc: for
informational purposes, and the team(a)fsfeurope.org to hear the opinion
of the FSFE on the matter. If LADders prefer not to continue the
discussion on linux-audio-dev, I'm sure nobody will object to removing
the latter list from the Cc:s :) ]
[For the list on Cc: we are talking about the redistributability and
GNU GPL compliance of the alsa-firmware package, as well as of
firmware in general, I'd say]
>>>> "Thomas" == Thomas
Charbonnel <thomas(a)undata.org> writes:
First of all thanks for Dyne:bolic :) All README
files from the
alsa-firmware package grant copyright to the respective
companies with the statement 'Redistributable under the GPL', so
I guess the answer is yes. As far as I'm concerned we received
several verbal and mail confirmations from RME that we could
redistribute the files, and Matthias Carstens (who I just met
last week) promised me an official written statement.
I absolutely don't want to start a legal debate here, given that it
would probably be off topic and the issue has already been (and is
being) widely discussed on the debian-legal mailing list, but please
notice that AFAICT distributing binaries under the GNU GPL license
means that the distributor must
(a) Accompany [the program] with the complete corresponding
machine-readable source code [...]
(b) Accompany [the program] with a written offer, valid for at least
three years, to give any third party, for a charge no more than
your cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code [...]
(there is also option (c), see the GNU GPL v.2 for further details)
The point here is understanding what the `source' of a piece of
firmware is.
The GNU GPL defines the `source' as the "the preferred form of the
work for making modifications to it". Now the debate on debian-legal
has been whether the hex-expressed firmware discovered in various
kernel files was actually hand-modified by the "distributor" with a
hex editor, or a higher-level language was used. If the latter is
true, then the GNU GPL has been breached (because I've never seen the
source code of the alsa-firmware package, please correct me if I'm
wrong).
So, saying that the firmware is "distributable under the GNU GPL" is
not sufficient `per se' to prove that the firmware itself is Free
Software.
My personal position is one of being a bit more pragmatic. A large
part of the hardware we use actually has firmware embedded into it,
the only difference being that we don't see it and we don't need to
upload it (for example, AFAICR the Pentium IV automatically translates
standard Intel machine code into an internal, risc-like, set of
instructions - nobody is asking Intel for the source code of *that*
firmware).
The issue is thorny and I agree that a Live CD without alsa-firmware
is not particularly efficient. On the other hand, I do see legal (as
well as ethical, if one wants to go down that route) problems in
distributing non-free firmware. I'd like to understand the various
options a bit more before launching ourselves into the
"users-need-it-so-lets-package-it" frenzy (I'd rather tell users that
they must bug the companies they buy hardware from to release the
`source code' of the firmware needed to operate those cards under
GNU/Linux, if we discover that the firmware is actually non-free).
Usual caveats apply: IANAL, YMMV, etc, etc. If anybody has a deeper
insight into the matter I'd love hearing it, since the problem has
been a PITA for us for a long time (see the lists.agnula.org archives
and devel.agnula.org `demudi' project's bug lists if you are
interested).
Bye,
--
Andrea Glorioso andrea.glorioso(a)agnula.org
AGNULA Technical Manager http://www.agnula.org/
M: +39 333 820 5723 F: +39 (0)51 930 31 133
"Libre Audio, Libre Video, Libre Software: AGNULA"