Re: open source, open science conference

[Brian Gough <bjg@network-theory.co.uk>]

Mark Galassi writes:
 > I agree: first of all, they might be violating the certification mark
 > since most of the talks are on OpenGL and stuff like that (SGI is
 > sponsoring it).  Second, I prefer the term "free software".  Were you
 > thinking of something else?

I agree with you, although I hadn't thought about that certification
mark thing (didn't all that trademark stuff go down the tubes ??).

My thoughts are that in scientific software we should be aware of the
danger of the term "open source" being (mis)understood as "source code
is available".

  We have the example of a widely-used library (NR) that has
source-code available, but where the license seriously restricts the
users' freedoms.  Because of these restrictions people have not been
able to collectively improve the software (resulting in much wasted
effort, etc).  The authors of that software even wrote that copying a
program "is not a friendly thing to do".  So it's quite possible for
people to make source code available while promoting ideas opposed to
free software.

 > Yeah, but James might be making some progress in that direction.
 > There is a possible path which (if navigated very accurately) can lead
 > to a public domain stamping of the software.
 > 
 > In any case, if Brookhaven works on GPLing their software, we can see
 > if they get good results and then try their approach.

Public domain seems the most realistic option, there is plenty of
precedent for that, the legal dept understand it and it is generally
accepted that this is how government funded work should be made
available.  The bad scenario of promoting the GPL would be that the
lawyers might look at it closely and decide that they can write a much
"better" license.... this seems to be a recent trend (NPL, QPL, APSL, ...).