🚀 go-pugleaf

In comp.lang.functional Steven G. Johnson <stevenj@alum.mit.edu> wrote:
> So why not use the GPL, like gcc uses?  (I assume by "commercial" you mean
> "proprietary."  Non-proprietary commercial use, similar to what Cygnus
> does with gcc, would benefit you because you would get to see and use any
> enhancements they make.)

Three points should be noted on this:

  Cygnus produces a compiler for a main stream language. It's not a
  problem for them if the competition makes use of e.g. new optimization
  algorithms - simply because the competition cannot be cheaper than
  Cygnus already is...

  If a firm wanted to improve Ocaml (or help improving it), there
  wouldn't be any problem for INRIA. But Ocaml is not main stream -
  if competitors from this (= the main stream) side made direct use of
  their algorithms (by copying the code), INRIA could get problems -
  you shouldn't forget that it is a research institution that has to
  report about the usefulness of their developments. If their intention
  is to provide an alternative product (not just a substitute), they
  are probably well-advised not to indirectly support the competition
  (competition = producers of main stream products).

  In the main stream market money is made with development environments,
  books, consulting, etc... The competition would have a strong interest
  in getting a better compiler if this could help them sell more of
  the rest (Cygnus also has to earn money...).  So GPL will definitely
  not distract them from grabbing INRIA's code (to INRIA's possible
  disadvantage).

Best regards,
Markus Mottl

> PS. I am a happy user of OCaml (see http://theory.lcs.mit.edu/~fftw).  Thanks!

PS. That's fine!
    By the way - I have just taken a look at "genfft": really nice! I like
    the application of monads in your simplification algorithms...

-- 
Markus Mottl, mottl@miss.wu-wien.ac.at, http://miss.wu-wien.ac.at/~mottl