CCL:G: License what you want. Open what you want. No comparison?
- From: "Pablo Echenique"
- Subject: CCL:G: License what you want. Open what you want. No
- Date: Sun, 3 Feb 2008 13:44:37 +0100
Sent to CCL by: "Pablo Echenique" [echenique.p!=!gmail.com]
I must say that the conversation is really interesting and many
relevant points have been introduced but, with all due respect, I
think people is taking it too personally.
On the one hand, we have two options: Open Source vs. Proprietary Source.
It could be interesting to discuss which one is best, whether or not a
sustainable firm can be based on the first, whether or not the
Proprietary Source has a place in an efficient global Science
But, whatever anwers each one of us finds to these question, the
decision to adopt one option or the other is a free one.
The Open Source movement, with which I agree in most essential points,
is being increasingly infected with radical supporters which, like in
Middle Age, ride every day to Jerusalem to kill the unholy Proprietary
In my opinion, all development models MUST be tried before deciding
which is best. And maybe the final, not so surprising answer is that
the most efficiency is obtained in a world in which all options have
at least one strong representant.
Moreover, if Open Source is the only answer to scientific software, we
will know it when we prove it, not when we declare it.
In brief, what is best for Science do not depend on ethical
principles, it depends on facts.
In this sense, I believe Mr. DeLano's insights are very useful for the
discussion, but so are Mr. Frisch's, because Gaussian Inc. is a
success history too.
Regarding my own free choice, I give away all knowledge, data and code
I produce, since, being paid by the government, i.e., by the people, I
believe it is the most ethical choice (and probably the most efficient
However, I do not find anything unethical in the way a private firm
such as Gaussian Inc. works. In the free market, nobody forces you to
buy a product. You don't like Gaussian's performance, decisions, name,
colour of the CD case, etc.? Just don't buy it. You have other
I prefer to live in a world in which Gaussian exists, so I have one
more option, which I can freely choose to take or not. I use Gaussian,
GAMESS, Aces and Dalton, and I consider Gaussian a good overall
application. My University has voluntarily signed a contract with
Gaussian Inc. in which the conditions of use are clearly stated. If I
freely chose to break some of that conditions (like comparing results
to other codes), it is perfectly licit that Gaussian takes the
appropriate actions. It is the same that I would do.
This being said, like Dominic Ryan, I do not understand that
particular condition and I think that it is at the root of the
disconfort many users feel about Gaussian.
I am not as pretentious as to give advices to Mr. Frisch about which
decisions are best for Gaussian, since it is completely obvious that
they have thought much more about the topic than me, and they have a
lot of relevant data that I ignore.
I just express my belief that the "no-comparison" clause is the most
Not opening the source is something that a lot of developers do and
they are not as fiercely criticized as Gaussian.
My two cents,
Instituto de Biocomputación y
Física de los Sistemas Complejos (BIFI)
Departamento de Física Teórica
Universidad de Zaragoza
Pedro Cerbuna 12, 50009 Zaragoza
Tel.: +34 976761260
Fax: +34 976761264