RE: Re: Open Source Licensing

From: I.R.Maturana <irm[_at_]in3activa.net>
Date: Mon, 01 Sep 2003 13:20:12 -0400

IRM>
> > Your answer implies that a contract may override the
> > licensee's rights on a derivative work.
Larry Rosen>
> That's exactly what I meant. I can license you to create a Spanish
> translation of my book but not a German one. I can license you to
> obtain my specific approval of quality before you distribute your
> derivative works (although my approval cannot be withheld for

I think you do not understand the issue raised by OSL and similar licenses. The issue is not about the primary author's rights, but about the secondary author's rights.
There are rights attached to the derivative work itself, which does not belong to the primary author. Both are authors, and are "equal in rights" under (C)Laws and Treaties.

The rule is: once your Open license allows a translator to create a derivative
work (i.e.: a Spanish translation), the Spanish translator holds on this secondary work (and only on it) the same rights than you hold on your original version (and only on it).
When we say that EN-author and ES-translator have "equal rights" on their respective versions, this does not mean that the rights are "shared", nor "merged".
You cannot substitue Authors of derivative works as you do for copies of your own work. Nobody, except the Law, may "license" individual rights.

As a consequence, a primary-Author who wants to distribute, or to control the distribution of a derivative work, will have to get a license from the secondary-Author.
We are not talking about copies (nor about the right-to-use duplicate works).
Copies always belongs to the same author. However, secondary works are independent, fully protected works, which are not the property of the primary author. Therefore, to distribute secondary works, you need a license from the secondary author, even if the license will be "reciprocal".

Your wording about "reciprocal licensing" does not work this way. By reading your answer, and the OSL itself, it appears clearly that OSL assumes that the primary-author holds full rights on secondary-works. In fact, OSL consider derivative works as if they were a special case of duplicative works. While OSL can control distribution of duplicative works, OSL is not directly enforceable for distribution of derivative works.

To help you to understand the difference, let us compare the OSL system with a network architecture. Rather than a simple architecture where a single primary-work is always in the center (a.k.a., a star system :),
let us consider a tiered-architecture, where secondary works become the base for other works (--for works unrelated with the primary work).

This example is typical in Software development, but let us continue with the
translation equivalent:
- Given a Spanish translation derived from English; - We want to produce a French Translation derived from Spanish.

Faced to this model, the French translator has to request the right-to-derive
from both authors: English author _AND_ Spanish translator. (Under OSL model, there is even no need to ask English author !)

In other words, once an author grants the right-to-derive, you get two (2) authors. Outside the OSL system, French translator will have to request licenses from the English and Spanish authors. Inside the OSL system,
French translator only need approval from the Spanish author (_not_ from English author!).
In short, the OSL cannot endanger the Spanish Author's right to grant derivative-rights to the French translator.

I also would notice that the fact that OSL is "copyrighted" does not override
any of the Spanish Author rights on the Spanish version, nor it will override
any right on the French Author's future right. (Well, this would be an excellent argument to do not follow the OSL !:)

Larry Rosen>
> Take a look at Bowers v. Baystate, one of many cases that holds that
> licenses can restrict the rights of licensees.

IMHO, this case applies to reverse engineering, and has nothing to do with derivation. A shrink-wrap license implies a preexistent work. This sample is meaningless on future works.

(Hope this long message will help you understand the issue that the OSL and related licenses are raising upon their own foundations... :)

Best regards-

I.R.Maturana -- http://www.in3activa.org (LPT-ESv1r3 - http://www.in3activa.org/doc/es/LPT-ES.html) Received on Mon Sep 01 2003 - 21:20:12 GMT

This archive was generated by hypermail 2.2.0 : Mon Mar 26 2007 - 00:35:50 GMT