Metaprogramming and Free Availability of Sources

[Laurent Martelli]

>>>>> "RMS" == Richard Stallman <rms@gnu.org> writes:

  Faré>     But there is no absolute natural right to access other
  Faré> people's information.

  RMS> I agree that privacy is an important right, and I would object
  RMS> to any law requiring people to release all software that they
  RMS> write.

  RMS> But when someone chooses to make a program available in some
  RMS> way, it is no longer a matter of privacy.  Then he or she
  RMS> should be required to show users what is in the software they
  RMS> are being invited to use.

But what about the freedom to distribute a 'binary' version of a
software ? As long as people are free to choose to use it or not to
use it, I can't see any problem. If we follow your reasoning, we'll
have to require that cars are sold we the complete building
instructions and detailed physical characteristics, because it's
probably the only way to know how the car behaves when used on a
road. But you could tell this by just inspecting it and analysing it. 

The same goes for software. One coulud argue that you can tell what
the software does from the compiled version. We could even go further
and say that one source code may given different compiled code
depending on the compiler used. But the same goes for the compiled
code, depending on the os and shared libraries. So it seems that
there's no such thing as a best form of a program do determine what it
does. 

And should people be forced to show their UML diagrams (or any other
crap) ? 

I think that when there's no clear separation between concepts, we
should not try to impose one, whatever the goal you pursue. 

My definition of freedom includes the freedom for others no to like my
definition of freedom. 

-- 
Laurent Martelli
martelli@iie.cnam.fr