Unix vs Linux - the movie.
John S. J. Anderson
jacobs+usenet at genehack.org
Sun Jul 30 10:58:15 EST 2000
-----BEGIN PGP SIGNED MESSAGE-----
>>>>> "Stephane" == Stephane Bortzmeyer <bortzmeyer at pasteur.fr> writes:
Stephane> In practice, I think that the peer review is less efficient
Stephane> when you cannot make changes and distribute them: you cannot
Stephane> test your ideas so widely. And people are less likely to
Stephane> work on a program if they cannot distribute and use the
Well, that depends on what you think the point of peer review is. To
my mind, the _primary_ purpose of peer review is to make sure that
accurate results are being presented. Access to source code not only
facilitates that process, in most cases (all cases?) it is *required*
in order for success.
A secondary purpose (or possibly a secondary source) of peer review is
the use of a set of results as a platform for further experimentation.
If this subsequent experimentation gives anomalous results, then you
can double back and examine the primary results more closely, to see
if they might contain undetected errors. Here, a license that
restricts your ability to make and distribute changes _is_ a
Nevertheless, I still think source code access (even if only by the
anonymous referees prior to publication) should be a required part of
the peer review process.
Stephane> I would say that freedom is not binary. There is space
Stephane> between free and "completely unfree".
As an aside, there is an interesting interview of Craig Burton in the
August 2000 issue of Linux Journal (page 16). In it, he outlines his
view on Open Source and code access -- basically, the same issue that
we've been talking around here.
He places software in a matrix defined by two axes: the X axis
represents ownership, and runs from Proprietary to Public Domain. The
Y axis represents code access, and runs from Closed to Open. The
useful thing about this model is that it cleanly separates the two
issues that have become tangled together in this discussion.
Again, I think the issue here (or at least the issue I'm concerned
about) lies on the Closed<->Open axis. The Proprietary<->Public Domain
axis is somewhat interesting, but orthogonal to the issue of peer
[ John S Jacobs Anderson ]------><URL:mailto:jacobs at genehack.org>
[ Genehack: Not your daddy's weblog ]------><URL:http://genehack.org>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.2 (GNU/Linux)
Comment: Mailcrypt 3.5.5 and Gnu Privacy Guard
-----END PGP SIGNATURE-----
More information about the Bio-soft