Subject: RE: Procedure for using an approved license
From: "Lawrence E. Rosen" <>
Date: Sun, 20 Oct 2002 13:11:37 -0700

> From: Rod Dixon [] 
> there is a lot being said here. To clarify one point at a 
> time, the use of "derivative work" should be in the copyright 
> law sense, not an unusual meaning gleamed from a 
> license...whether it is the MPL or any other license. In this 
> respect, the issue is relatively simple; namely, did the 
> copyright holder grant the right to create the work and did 
> he or she grant the right to distribute the work.

Thanks, Rod, for your clarification.  We attorneys have a responsibility
to our community to be precise in our definitions, at least with
reference to terms of art such as "derivative work."

Mitchell is correct in suggesting that some licensors may want the
reciprocity obligation (to publish source code) to apply to more than --
or to less than -- derivative works.  The GPL authors, for example, seem
to want to include works that link together in some ways; they are
entitled to do so as long as they define their terms clearly and so long
as their definitions are consistent with the copyright law that governs
their license.  The MPL, by contrast, wants to limit the reciprocity
obligation on a file-by-file basis, also a legitimate objective for a
license as long as the term "file" is clearly defined.  

In drafting the OSL, I tried to steer clear of terminology that was
technology-specific and to use terms of art from copyright law wherever
possible.  I did not want to clutter the concept of derivative works
with terms such as "larger work" or "work based on the work" or "file."
When I wanted a specific software concept, however, such as "Source
Code" or "External Deployment," to inform the application of the
copyright law to this license, I tried to define it clearly.  

Perhaps Mitchell, in the next version of the MPL, will be able to define
more clearly what she intends to encompass by the "derivative work"
reciprocity provision in that license.  That will help projects to
decide which license to adopt for their software.  Perhaps, too, we
should work together to define that term precisely for our needs and use
an agreed definition in both the OSL and the MPL?  

As for the current version of the OSL, I thought it best to let the
courts clear up the concept of derivative works in the edge cases, since
they will anyway if someone litigates this in an important case.  I have
written an article explaining my own views of how that will turn out; it
will appear in my Linux Journal column in a few months.  


license-discuss archive is at