A discussion of hybrid open source and proprietary licensing models.

[Marshall Van Alstyne <marshall AT MIT.EDU>]

Ryan thanks for your participation.  This RL experience is highly 
valuable.  So on to a few questions :) ...

At 03:02 PM 9/15/2004, you wrote:
> My concept of the split was like a word processor and the book you write
> on it.  If the code for the word processor is open, everyone can make it
> work better, and positive change will create a momentum for overall
> improvement.  But my book is my personal work, from which I expect to
> extract value in the form of compensation.  I don't want the output of
> the word processor to be open, otherwise I'll kill the value
> proposition.

This "feels" right to me.  I also note that the same issue comes up even 
with programs like GCC where the extensions to the compiler itself fall 
under GPL but programs compiled with GCC do not.

But let me play devil's advocate

Can anyone provide the word processor/book or compiler/executable intuition 
with a principled theoretical or economic argument?  The reason I ask is 
that the dividing lines are not always clear.  When should this principle bind?

At one extreme, a book or compiled executable need not include *any* code 
or content from the tool used to produce it.  In that case, any number of 
similar tools could have produced it and there seems to be no special debt 
to terms in using the tool.

Suppose, however, that the output itself was only possible with one 
particular tool (say the very 1st compiler).  Suppose further, that the 
output starts to include bits from the tool (say parts of the 
executable).  When does the new work no longer differ enough from the tool 
or from some other form of content?

> There is actually less and less of this kind of behavior because the
> publishers as a group are becoming more aware of the concept of shared
> utility.  There's an economic incentive for them to collaborate - they
> all want to use each other's work.  This is a social impetus, not a
> license-driven trend, and if someone proves a business model that shows
> disregarding the value of the "cathedral" makes money, it could be lost.
> But so far, the trend has been in a positive direction.

FWIW, a piece of economic modeling we're doing seeks to show why they 
really *should* collaborate.  There's a bit of a prisoner's dilemma in that 
each wants to share the other's work but not necessarily to share their 
own.  But, the whole system benefits if each does share.

> I had the advantage of total crisis.  The D&D brand was nearly dead when
> I got my hands on it, and the company had spent tens of millions of
> dollars to acquire it.  In such extreme circumstances, management is
> often willing to cast caution to the winds on a "hail Mary" type play.
> It didn't hurt that the CEO, Peter Adkison, could see the long term
> value to himself personally of being able to use D&D in commercial
> products regardless of his relationship to the company that owned it.
> In fact, Wizards of the Coast's first product (written by Peter) was a
> book that endeavored to create a unified system for many RPGs and which
> was battered by real and threatened litigation as a result.

Again, very cool RL experience.  I just hope this isn't what it takes to 
get mgmt attn.

MVA