Derivative/collective works and OSL

John Cowan jcowan at
Thu Feb 3 16:49:31 UTC 2005

rich apodaca scripsit:

> Can anyone give me examples that illustrate the boundaries between
> "derivative works" and "collective works" as these terms relate to
> software?

Ah, that's a difficult question, a very difficult question.  Experts
disagree, and no court has ruled.

> For example, it seems clear that if I take library source code,
> add material to one or more source files, and distribute the result,
> then I've created a "derivative work".

Almost certainly.

> But what if I write an application that just links to the library? For
> example, what if my work simply uses Java "import" and "extends" or
> "implements" statements? Or what if I use both "import" and "extends"
> and I additionally use "new FooClass()" where "FooClass" is a class
> defined in the library?
> Are the two situations above different? In in each case, have I created
> a "derivative work" or "collective work"?

The FSF says it's a derivative work, Larry Rosen (and I) say it's a
collective work.

> I am interested in this topic because the Open Software License (OSL)
> seems to rely on copyright law to define the term "derivative work." As
> far as I can tell, a "collective" software work would fall outside
> the scope of the OSL's reciprocity requirements (sections 1c and 5).
> If so, do the LGPL and OSL reciprocity requirements actually differ
> in any meaningful way?

Neither the OSL nor the GPL nor the LGPL binds collective works to
reciprocity, everyone agrees on that.  But:

	God bless the King, I mean the Faith's Defender;
	God bless (no harm in blessing) the Pretender.
	But which Pretender is, and which is King,
	God bless us all! --that's quite another thing.
		--John Byrom (1773)

There is a similar dispute, though not so problematic, at the other end:
does mere compiling of source code create a derivative work, or is the
object code the original work in a different medium, as a paperback
book is the same work as a hardback original?  Nobody knows the answer
to that either.

The reason it matters is that pretty much everyone agrees that a tarball
is a collective work, and if when compiled it is still a collective work,
then it is not derivative of any of the works contained in the tarball.

                Si hoc legere scis, nimium eruditionis habes.

More information about the License-discuss mailing list