Subject: Re: OSDD?
From: Brian Behlendorf <brian@hyperreal.org>
Date: Mon, 24 Aug 1998 18:11:17 -0700

At 02:32 AM 8/24/98 -0400, Brian Bartholomew wrote:
>Here are the terms I want to buy software under, which amount to
>"capped profit".  Suppose five guys sit in a room for a year and write
>a spreadsheet for Linux.  Suppose they spend $1M doing this.  You can
>sell rights to run this software which include the source, but you
>must not collect more than $2M.  If you have 50K purchasers, each
>purchaser must pay no more than $40.  At this point the software is
>GPL'ed.  Then you find another program to write.  I have a few tweaks
>to add but this is the basic idea: repay your investment, make a big
>profit, price is simply related to cost and not tied with support,
>don't bleed me forever.  You are incented to get the repeat customer
>by making your current offering good.

This is uncanny.  Wednesday night I had a conversation with a founder of a
rather large software company (one brought up here recently).  His company
was ready to unleash a new "framework" specification and technology onto
the world, one which is actually pretty nice, well thought out, and could
be a real enabling force.  

He was asking me (and others) what sort of license he should put on it.  We
started talking about all sorts of things: the bootstrapping problems
inherent in launching new frameworks, trying to distinguish between
"commercial" and "non-commercial" uses of the framework, the new
provider-consumer dynamics inherent in the open-source software model, etc.  

Some among us said that so long as the framework cost money for any type of
use (commercial or non), that it was not possible to call it "open source",
and that it would suffer the same types of confidence problems and market
problems that other types of framework technologies, like Java, have suffered.

I suggested a revenue cap, just like you did, above.  Tell the world how
much you want to make on it to recoup your investment, plus your typical
margin, and be public about how well you reach that number, and then after
you hit that number, make the IP libre.  

The response was that this idea, and most of the other business models
espoused by the open source community, were "pre-Industrial" - that the
Industrial Revolution (and I'm paraphrasing here, so be gentle) allowed
someone with a bright "idea" to be able to reap rewards persistantly over
time due to having developed that "idea".  And that this ability was
crucial to running a 20th-century business, particularly a technology
business.  Patents were invented to recognize and protect this aspect of
business.

The dim view of this is that business people want it both ways - a level
playing field, but unique, protected advantages.  That this concept is what
in fact has allowed certain large companies to develop into bohemoths,
weilding the bully club of IP rights that far transcend moral rights.  And
that being renumerated over and over for the same bit of work is just as
wrong as being punished twice for the same offense.

The bright view of this is that it can allow small companies with
exceptionally bright employees the chance to compete against the big boys
should they be lucky enough to come up with such a valuable idea.  Without
this protection, good ideas could be appropriated by the big guys who have
more resources to shephard the idea into revenue-producing reality.  This
concept reinforces the value of intelligence in the business community.  No
longer is your earnings potential limited to how many hours/week you worked.

So what's the right answer here?  Is wanting to earn money purely from an
idea somehow irreconcilable with open-source?  

Here's another angle to it.  Is source code a form of speech?  Of
expression?  If so, who among us would argue that a film actor should make
back "just enough" money from a movie's release, thereafter being free?  Or
a musician should only make up to a certain amount of money from their
music, and after that point (save for performing live, etc) it's free?   

There's a very strong cultural moral that favors being recompensated for
the same work performed over and over. That's a moral that may have to be
overturned if you tell a software company that in the future they should
only expect to get paid once for their investments.

There is also something of a halting problem here.  Until it can be shown
that such a pledge of "capped profits" would increase sales, such a pledge
would be a pledge to leave money on the table, money you've rightfully
"earned" under the way business works.  It's also unlikely you'll hit your
$$ targets, because many people will wait until it's free, even if in doing
so they prevent it from ever becoming free.  Who wants to be the last
person to pay?

	Brian


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
"Common sense is the collection of prejudices  |     brian@apache.org
acquired by the age of eighteen." - Einstein   |  brian@hyperreal.org