[License-discuss] relationship between opensource code and the copyrighted works it produces?

[Karl Fogel]

forget color <forgetcolor at gmail.com> writes:
>I'm a visual artist and composer. Code is my medium. I'm interested in
>releasing some of my code as open source, but don't quite understand
>the licensing and copyright relationships between an open source
>codebase and the artworks that the code may produce.
>
>There are two types of situations common in my work, and they may
>require different treatments. First are works where I don't really
>intend anyone else to work on the project with me, but want to make
>the code available for learning, understanding, and potential
>adaptation into a new (but different) project. Second are works where
>I want collaboration on the coding, but generally only to help improve
>the work I made with it.
>
>Therefore, in both cases my intention is to maintain copyright on the
>artworks I create using the code. 
>
>For example, if the code produces a browser addon called Widget99 then
>I want Widget99 to be a distinct copyrighted work of mine that nobody
>else reproduces and pitches as their own. If I make a code-drawing
>system and called it CDS then I want CDS and the works I produce with
>it to be copyrighted as well. 
>
>However, in each case the code I used to write it could be used in
>other's projects, whether by concept or direct adaptation.  But only
>as long as their new works don't infringe my copyrighted works.
>
>Is what I want something that can be (or should be) supported by an
>open source license, or is what I want inherently non open source? How
>is a copyrighted work produced by code affected by that code's open
>source availability?
>
>Any suggestions for licenses to look at are appreciated. The only
>thing I found that might be close for my Widget99 example is something
>like the Open Art license (http://three.org/openart/license/), but it
>doesn't appear to be supported by FSF or OSI.  I can imagine GPL or
>similar could be used for my CDS example, but only if the copyrighted
>works are still protected despite the open source nature of the code.

My understanding (I am not a lawyer) is that copyright only applies to
creative works -- specifically, to works resulting from human
creativity, or to the portion of a work that results from human
creativity.  

This is why, for example, the information in a phone book cannot be
copyrighted, but a song reciting those numbers could be.  Or why when an
editor does a new edition of a public domain work (say, a Beethoven
piano sonata), the edition is copyrightable even though the underlying
work of Beethoven is still public domain.

So I don't think you need to do anything special here.  The resultant
works are copyrightable (under any terms you choose) to the degree that
they result from your creativity.  If they are completely automatically
generated by open source code, I'm not sure they'd be copyrightable, but
I'm also not sure that fact would have anything to do with the license
of the code that generated the works.

-Karl