OSL 2.0 and linking of libraries

jcowan at reutershealth.com jcowan at reutershealth.com
Thu Apr 1 19:58:42 UTC 2004

Forrest J. Cavalier III scripsit:

> As far as I understand it, when
>    moduleA + moduleB = statically linked executable
> executable IS a derivative work of both moduleA and moduleB.

That's what's at issue.  There aren't any cases in point, so we are forced
back on analogical reasoning.  It seems pretty clear that a tarball is not
a derivative work: it is a compilation, like a collection of articles or
an anthology of short stories.  The copyright in a compilation as such
extends only to the choice of components and their arrangement: in the case
of a tarball, the latter point is mechanically determined and irrelevant.

But it's not clear why there should be a distinction in law between a tarball
and a static executable.  In both cases, it's tolerably easy to identify
individual parts of the result with the corresponding original inputs, although
these do not look the same (as a result of compression in the one case, and
compilation on the other).

In short, if I supply you with the source of a GPLed program plus binary
versions of certain non-GPL modules, and allow you to do the linking, it
doesn't seem that this offends the Copyright Act as conditioned by the GPL's
grants.  Yet this is exactly the case that the LGPL was designed to handle.

And I'm willing to bet that the LPGL's requirement for providing the
proprietary parts of a program in linkable form is far more often honored
in the breach than in the observance.

You let them out again, Old Man Willow!                 John Cowan
What you be a-thinking of?  You should not be waking!   jcowan at reutershealth.com
Eat earth!  Dig deep!  Drink water!  Go to sleep!       www.reutershealth.com
Bombadil is talking.                                    www.ccil.org/~cowan
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3

More information about the License-discuss mailing list