Re: [tlug] [OT] Say _no_ to the Microsoft Office format as an ISO standard

["Stephen J. Turnbull" <stephen@example.com>]

Jean-Christophe Helary writes:

 > How do you convince developers that giving others the right to close  
 > the code they took so much time to write is a benefit ?

Getting them all the way to "benefit" is hard.  It helps for them to
have advanced training in economics, which is obviously a non-
starter. ;-)  However, I can outline the argument.

- First of all, the code *they wrote* *cannot* be closed in the
  abstract unless they close it themselves (and possibly not at all).
  It is always available in its original form.

- Second, some instances of the code can be closed (eg BSD -> BSDi),
  but you have to wonder why customers would pay for it.  Sure, the
  first few times it happens there may be some fraud involved, but
  *caveat emptor* is a reasonable answer to that.  Emptors do caveat,
  you know, and pretty quickly they learn about lock-in and the like.
  There are whole textbooks written on the subject.  So there must be
  some reasonable tradeoff here for the buyers.

  Why not leave it up to the buyers to decide in what form they want
  their code, given that you have given them the choice?[1]

So it's hard to see that letting others produce closed-source products
from code released under open source is a loss at all.[2]  Losing the
viral property might hurt, but that's not the question you asked.

As for where the benefit comes from, the argument is implicit in the
above.  Giving consumers choice is a benefit; since distributing
publically as open source but allowing closed-source redistribution
gives the consumers more choice than enforcing open-source
redistribution, there is a benefit.

If you want to discuss the paybacks required by the viral clause,
that's arguable either way.  Above, we have already logically proved
that there are net benefits to permissive licensing if *all* wannabe
licensees look at the viral clause and say, "no thanks; I'll develop
my own closed code."  It seems highly probable (but cannot be
logically proven!)  that there are net benefits to viral licensing if
*all* wannabe licensees accept the viral clause and open their own
code.  The real world is, of course, somewhere in the middle; it's an
empirical question where.

Interestingly, when I proposed to Stallman that this research be done,
he begged me not to do it, and if I did it, I should withhold my
results from publication, because it could only undermine the purely
theoretical arguments of the free software advocates.

 > And I did not declare that use of "liberal-open-source-license code"  
 > was stealing, it is closing the code and/or not contributing back  
 > that is stealing.

Just who do you claim is being stolen from?  What property is being
stolen?  And where do the victims' rights to the property come from?

And how do you claim that the *licensee* is stealing?  Surely, if any
theft is occurring, it is the *licensor*, not the *licensee*, who is
committing it.  AFAICS the licensee can rely legally and ethically on
the licensor's permission, and not be held to be guilty of theft.

That is, if I, or Jeffrey Friedl, or Larry Wall, or Guido van Rossum,
or Eric Allman, or Keith Packard, or HP/Sun/DEC/Fujitsu/..., or Donald
Knuth, et al, ad nauseum, wanted to prevent closing and/or
noncontribution we'd use copyleft licenses.  If *anybody* is
"stealing", it is *us*, by authorizing the allegedly antisocial
behavior.  And not our downstreams who are only doing what *we*
authorize them to do.  I take *strong* exception to that accusation.

 > For total disclosure though, I have just created such a surplus value  
 > creating thing: I just incorporated, here, in Takamatsu, and I can  
 > tell you that fair redistribution of the surplus value I'm intending  
 > to create _is_ my aim.

Fair redistribution of zero is very easy.  Fair redistribution of a
negative surplus is even more painful than redistributing zero is
easy.  If you want to enjoy the redistribution of surplus, you'll do
better to aim at creating it, first. ;-)


Footnotes: 
[1]  Given the axiom that each person knows better than anyone else
what she needs, you cannot turn that question around.  You need to
explain the benefits of removing choice, otherwise there is no way to
justify any course but to offer the choice.

[2]  Seeing that those arguments are correct, and knowing what the
important underlying assumptions are, is what requires advanced
training.  I assure you that the above can be expressed in such a way
that any competent economist would agree the argument is correct.
Most would agree with the "important" assumptions, but that is not
universal.