Re: Knuth statement on renaming cm files and Licence violation.
I don't really have anything to say that hasn't been said by others in
the past, but since some folks are (understandably, really) tired of
the issue, I thought I'd try to explain the standard debian-legal take
on it. Someone correct me if I'm inaccurate.
Russ Allbery <rra@stanford.edu> writes:
> What brought me into this argument, though, is basically a question of
> principal, namely that licenses should be evaluated as if they were
> enforcible even if written in a way that may make them legally
> unenforcible for some reason and even if the person doesn't intend to
> enforce them. And actually, I doubt that people really disagree with me
> on that.
No, I don't think anyone disagrees with that. I think the problems
comes up because your scenario is a bit ambiguous; it describes some
very different situations:
A) Someone releases software under a license that debian-legal thinks
is free (quite self-evidently, perhaps -- the pine license is a
real-world example of this, IIRC). But the licensor comes to us and
tells us that our interpretation is all wrong, that the license does
not permit X, Y, or Z, which would clearly make it non-free.
Debian-legal will (and has) decide that this makes the license
non-free, even though the same license in use by someone else would
likely be considered free.
B) Someone releases software under a license and claims it is free,
but debian-legal looks at it and doesn't think that the license
permits the freedoms that the licensor says it does. This puts us in
a awkward spot, we don't know whether the licensor doesn't understand
the freedoms or the license. Generally someone tries to talk to the
licensor and figure out what's going on before deciding for sure
whether the license is free (and hopefully gets the ambiguous
requirements cleared up in the process).
C) Someone releases software under a free license, but with addenda
that make requests that are not legally binding. These addenda would
perhaps make the license non-free if they were legally binding, but
since they are not the license is considered free. The author may do
this because he thinks that the law is too blunt an instrument to
accomplish his purposes, and so will use more social mechanisms
(public condemnation, boycott, etc.) to further his ends. My
understanding is that most folks on debian-legal are quite happy with
this approach, and in fact will often recommend it in cases where the
author has a particular concern with respect to the software.
Respect for the author's opinion means a different thing in all three
of these situations, and it isn't always clear which scenario is
active in any given situation. The question in the current situation
is which of these scenarios applies. If it's A (the license is
ambiguous and/or Knuth claims it places restrictions that aren't DFSG
free, even though debian-legal doesn't see that in the license text)
then the license may well be non-free, and further discussion
(preferably with the copyright holder, which would presumably be
Knuth) is required. If it's C (which to my eye seems as likely, if
not more so, as a reading of the text) then there's no problem.
To be honest, given that there's such disagreement on the issue, I
don't really see how it can be resolved without contacting Knuth,
explaining the issues to him, and getting his verdict.
Note, though, that I think everyone here agrees that the extra
requirements in scenario C are *morally* as important to follow as the
legal ones. But nonetheless there is an important distinction between
a request (however strongly worded and important) and a legal
requirement.
--
Jeremy Hankins <nowan@nowan.org>
PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03
Reply to: