copyrightable APIs? (was RE: namespace protection compa

Rod Dixon rodd at
Fri Apr 20 21:01:44 UTC 2001

On Fri, 20 Apr 2001, Chloe Hoffman wrote:

> >From: "Forrest J Cavalier III"
> As I suggested, infringement I think is the nub of the question - not
> copyrightability. In my view, there is certainly a significant difference
> from an infringement perspective, assuming no express or implied license or
> other kind of estoppel, between writing a program accessing an interface and
> implementing a set of APIs (such as the Java API) - in my mind the former
> being much further away from infringement than the latter.
If by "implementing a set of APIs," you mean copying them, I would agree,
but isn't that always the case? The more you copy, the stronger the case
of infringement. In fact, that is a conclusion that we might all agree
with. Hence,
copyrightability is still an important part of this equation, but,
perhaps, in a different manner. The distinction is not the question of whether the API, itself, is
copyrightable, but whether the infringement analysis should
appropriately filter out
uncopyrightable subject matter. In other words, looking at just the
question of infringement, we still need to decide whether an
implementation (or use, or copy) of a set of APIs should require the
filtering (filter prong of the abstraction/filtration test) out of source
code that is not copyrightable. I think the answer is
yes. But, I would also go further than this.


More information about the License-discuss mailing list