article on GPLv3, Linux kernel, and Devices Rigged to Malfunction

Ben Finney ben at
Sun Oct 22 22:44:12 UTC 2006

On 22-Oct-2006, Alex Hudson wrote:
> On Fri, 2006-10-20 at 20:08 +0100, Ciaran O'Riordan wrote:
> > Option #2 is market pressure, which is a much more difficult.  Our
> > job would be much easier if our kernels were GPLv3'd [...]
> I'm not sure it would make much difference.
> The kernel is currently GPL'd, yet it doesn't encourage people to write
> free software drivers. Whether or not NVidia/ATI's system for writing
> non-free drivers is technically legal I don't know, but it seems no-one
> with affected copyrights cares enough to do much about it.

Greg Kroah-Hartman, at least, is talking about the presumed illegality
of non-free drivers in Linux at every opportunity. (Why he's not
chosen to take legal action against infringers, I'm not sure.)

> OpenBSD seem to have been measurably more successful in opening up
> wireless drivers than Linux, and their license doesn't even require
> copyleft.

This is true. You are probably right that the effectiveness of free
software copyright holders in getting free driver information from
vendors is more to do with personal attributes than kernel copyright

> I also worry that a lot of GPL violation appears to be in the embedded
> arena. It's unlikely that a new GPL version would increase compliance in
> and of itself, and I would suspect that un-Tivo-ising something after
> the fact may or may not be possible.

A consistent history of copyright violation lawsuit wins against these
infringers would educate the individual vendors involved, and would
send a clear message. I can't imagine a successful prosecution result
that didn't include cessation of distribiting infringing works as part
of the judgement.

> There is a limit to which you can use copyright on a work to govern
> precisely how it gets used (hence the pro-DRM laws).

How it gets *used*, no. But copyright is, and has been from the
beginning, all about governing how one can *derive* and *distribute*
works, which is what we're discussing here.

Don't make the mistake that the Linux developers are making: of saying
the term "use" for the action of distributing a derived work. That's
both counter-intuitive to how most people think about software use,
and conflates acts that are strongly governed by copyright with acts
that are not. Best to keep them distinct.

 \        "We have to go forth and crush every world view that doesn't |
  `\             believe in tolerance and free speech."  -- David Brin |
_o__)                                                                  |
Ben Finney <ben at>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
URL: <>

More information about the Discussion mailing list