Note: lists.debian.org is refusing to allow me to subscribe: please
forward this post.
MJ Ray scripsit:
> Sublicensing means that you are still bound by the original licence,
> but you can offer any licence in the specified range to those you
> distribute to.
Quite so, and I should have clarified that point. If Alice licenses
a work to Bob under a sublicenseable license, and Bob sublicenses the
same work or a derivative work to Charlie, then Charlie cannot receive
from Bob under a license that contradicts Alice's license. In the
case of the AFL, as of the MIT license, it is so permissive that it
is almost impossible to contradict it.
> The wording in the AFL looks like the range of
> permitted sublicences consists of only the AFL, but maybe I
> misunderstood it.
No, it permits sublicensing the work and derivative works under any
license that doesn't contradict the AFL, since it would require particular
words (as found in the OSL, the AFL's sibling license) to restrict it.
The license need not even be a free license. So works under the AFL
can be treated much like those under the BSD or MIT licenses.
I have urged Larry to add explicit words "under any license whatsoever" to
the next version of the AFL, to make the consequences of sublicenseability
clear.
By contrast, the MPL is also sublicenseable, but the Original Work and
Modifications can only be sublicensed under the MPL itself.
> I'm surprised if the author of the AFL thinks it can
> be replaced by any licence, as that would seem to be a trivial way to
> defeat its overbroad patent termination.
The AFL, being a contract, cannot and does not attempt to defend Licensor
against patent lawsuits by third parties who are not licensees.
I'd be interested to learn why you think the patent-termination language
of the AFL 2.1 is overbroad. It says:
This License shall terminate automatically and You
may no longer exercise any of the rights granted
to You by this License as of the date You commence
an action, including a cross-claim or counterclaim,
against Licensor or any licensee alleging that the
Original Work infringes a patent.
So it takes away Bob's right to use Alice's software if Bob claims that
that very software infringes Bob's or Charlie's patent.
> I think you are right that it's an unusual practice for free software,
> though, but I'm not a lawyer.
There's an awful lot of MIT-licensed and MPL-licensed software out there.
Any of it can be sublicensed (with varying restrictions on the sublicense);
whether it is or not, nobody knows.
IANAL, TINLA.
Glenn Maynard scripsit:
> For example, the choice of venue clause in the AFL doesn't exist in
> the GPL. If I receive a work under the AFL from John, integrate some
> code from gcc, and send the result to Bob, can Bob sue John without
> being bound by the choice of venue? If not, it's GPL-incompatible.
> How does "sublicenseability" help here?
Of course, your distribution to Bob is under the GPL.
The fact that Bob receives code that was originally licensed by John
under the AFL doesn't make Bob a licensee of John. The only ways
for Bob to become a licensee are to explicitly manifest his assent
(in which case he is bound) or to do one of the things mentioned
in AFL section 1 provided there is no other way to do those things.
In fact, however, Bob can make and distribute copies and derivative
works under the GPL license which he has from you.
=====
To respond to a few points made by others: the fact that the AFL is
a contract does not make it violate the Autocrat Test, since the licensee
makes no promises under the AFL (it is a *unilateral* contract). The
right to use, as well as all the copyright and patent rights, are
explicitly granted. The only thing a licensee can't do is remove the
copyright notices, which is also prohibited by many other free licenses.
The OSI board definitely does analyze licenses themselves; the fact that
the AFL's author has an institutional connection does not mean their
judgment of his licenses is biased, any more than the fact that RMS has
an institutional connection with the FSF means that the GPL and LGPL are
not free licenses.
All Figlet authors are within U.S. jurisdiction, or were at the time when
they were authors.
The claim that *all* authors must agree to a change of license must
depend on some legal theory that I'm not aware of and that is not
stated. We are dealing here with a single module that has been repeatedly
patched, not with independently licensed modules (apart from the zip stuff,
which is now out of the case). That seems to meet the U.S. statutory
definition of joint authorship, by which any author can act unilaterally
but is responsible to the other authors for the profits (not an issue here).
--
John Cowan [EMAIL PROTECTED] www.reutershealth.com www.ccil.org/~cowan
I am he that buries his friends alive and drowns them and draws them
alive again from the water. I came from the end of a bag, but no bag
went over me. I am the friend of bears and the guest of eagles. I am
Ringwinner and Luckwearer; and I am Barrel-rider. --Bilbo to Smaug
