Re: Patent clauses in licenses
Andrew Suffield <asuffield@debian.org> wrote:
> On Sat, Sep 18, 2004 at 12:12:53AM +0100, Matthew Garrett wrote:
>> The implication of the post I replied to was that any license that
>> allows the removal of some set of the rights it grants should be
>> non-free. The GPL is an obvious counter-example, since it allows you to
>> lose all rights associated with it.
>
> Termination for non-compliance, in a publically redistributed work, is
> just a reflection of copyright law; it doesn't really change what you
> can and can't do. (You can always get another licensed copy). Every
> free license does this, really.
RMS has in the past claimed that failure to abide by the terms of the
GPL results in a permanent loss of those rights (in respect to a
specific piece of software, at least). If you're going to disagree with
the copyright holder of what is probably still the largest single body
of GPLed software in Debian at present, I'm going to want evidence of a
decent legal standpoint for this opinion.
> The use of a termination clause to introduce other restrictions (other
> than "you must comply with the license"), rather than simply writing
> those restrictions in directly, indicates that they probably aren't
> things you can write in directly, such as restrictions on use
> (copyright abuse aside for the moment; that doesn't help us, it just
> employs more lawyers). Such things are non-free restrictions (the set
> of things you're not allowed to restrict in a copyright license is
> fairly small).
As far as I can tell, your argument is that "You may not initiate patent
suits against the licensor" is equivilent to "Initiating patent suits
against the licensor will result in the loss of your rights under this
license". I would tend to agree. You then appear to claim that the first
is obviously non-free, and as a result the second is non-free. I see no
obvious reason that the first point of this assertion is true.
If you want to claim that the only restrictions on freedom we currently
accept are those that are entirely controlled under copyright law, you
may be correct (the Apache License 2.0 is an obvious counter-example,
but you could always claim that that's counter to normal policy and
thus some sort of error). I'd suggest that you're wrong here.
The restrictions that the DFSG allow are the restrictions that were, at
the time, associated with software that most people accepted as free.
Any claim that they were accepted as free for other reasons (such as
being entirely within the realms of copright law) entirely ignores the
context in which the DFSG were written.
We don't accept restrictions as free because they use one branch of the
law - we accept restrictions as free because they are either unimportant
or because they protect free software more than they hinder it.
--
Matthew Garrett | mjg59-chiark.mail.debian.project@srcf.ucam.org
Reply to: