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

[Alex Hudson]

On Tue, 2006-10-24 at 08:29 -0400, simo wrote:
> On Tue, 2006-10-24 at 13:14 +0100, Alex Hudson wrote:
> > The opens up a new, legally sound, route out of violation: people like
> > Eben Moglen can go to violators and say, "If you come into compliance,
> > you know that within two months your legal liability in this area is
> > gone - the slate is wiped clean". The FSF, nor anyone else, can go back
> > on an agreement they make and attempt to sue a violator outside that
> > time.
> 
> This is wrong.
> You are clean after 2 months if no copyright holder complains with you.

AFAICS, you're just restating the inverse of what I said - you can't sue
anyone outside the 2 month window.

But whatever. I didn't mean to infer that you couldn't be sued within
that two month window, however you read it, just that you cannot be sued
outside it.

> This is not a problem of designing secret systems, all these systems are
> pretty well known, what is unknown is the key to sign the hash of the
> data. And the point is moot anyway, releasing keys in this case is
> exactly like release the source code. You can expect the same level of
> infringement imo.

Possibly, but you're missing the point. It's only like releasing source
code in so far as the GPLv3 defines source code to include authorisation
keys.

Very few free software authors and distributors who will be roughly
compliant from day one will design in authorisation codes. It seems to
me that they will be used almost exclusively by those who are also
likely to violate the GPL by not distributing source code.

Cheers,

Alex.