Subject: Re: small worlds and better than ransom
From: "Forrest J. Cavalier III" <mibsoft@mibsoftware.com>
Date: Fri, 05 Oct 2007 16:03:19 -0400

Thomas Lord wrote:

> The "ransom" models of funding R&D don't quite work.  They
> say "I will release a program that does X but only if the public
> ponies up $Y."   Variations might let the exact value of $Y float
> but all ransom variants have in common that "X" -- what the
> program being sold does -- is known in advance.   (I consider
> "bounty" models to be a special case of "ransom" models, for
> this analysis.)
> 
> The only ways to know in advance that your program will do X
> is if either it is a fairly uninteresting program that you plan to
> write or if you've already done the bulk of the interesting work.
> Either way, you aren't selling the activity of doing R&D: you're
> selling past work.  (In the case of "bounties," if X isn't boring
> and trivial and isn't already done, then you aren't selling development
> at all -- you're selling your liability that feature X will be
> ready by some specific time.)

Saying "I have written software with feature X, ransom is $Y" reveals a great 
deal about the uncertainty of completing a competing product (nearly 0%), and 
the effort needed to produce it.

In Open Source, a lot of your competitors are not in it for the money.
If a small group knows it can spend a couple of weekends creating
an GPL'ed X competitor, they might do it for fun or the glory.  They
lack $ but have time.  They don't want to waste either time or money
on failed efforts, but you just told them what it costs to finish.

How do you avoid that problem?

If your $Y is bigger than a couple of weekend's of effort, then you
don't have as much competition.  It is also true that you have a
very limited market of buyers who want to spend $10,000 or more to
free your work.

One strategy is publicizing your "X" only to specific potential buyers.
This is establishing the supplier-customer relationship, which you
describe.

I'm not trying to be critical, I like your analysis, but what does this
have to do with Free Software?   Customers like it when they can
buy an "X", and not let competitors have "X" for free.