Academic free software
Posted 6 Dec 2001 at 16:42 UTC by isenguard
Most academic software is released under a "free for academic use"
license, which makes it impossible to use such software in a free software
project. Some notable counterexamples exist. I am looking for more
examples of academic software that has been successfully released
under truly free licenses.
The majority of academic researchers release the software they create
as part of their research, but they generally do so under a license
that restricts use of the software to academic purposes. Among other
things, that makes it impossible to incorporate research software
directly into free software projects.
I and a colleague are in the process of writing a paper advocating
that our particular research community adopts free software licensing
as a general practice. The benefits to the free software community
of having more free software are clear, but the arguments for making
academic
software free are less well defined, and different from those applicable
to the commercial world.
Can anyone point to successful free software projects primarily
authored by academic researchers, for which the release under a free
software license has benefited the research being undertaken?
The R project is the one I know
best, but there must be others.
I would say that ET++ and InterViews
fall in this category. These OO frameworks were released under
BSD-style licenses, and used by many parties. They re-invested this
technical capital into DesignPatterns.
BSD Unix, posted 6 Dec 2001 at 18:08 UTC by goingware »
(Master)
Well, there is BSD Unix, which was originally a research project of the
University of California at Berkeley (BSD = Berkeley Software
Distribution).
While it wasn't originally freely redistributable, because it contained
AT&T source code, eventually the released large parts of it without the
AT&T source, and then they wrote replacements for the AT&T source so the
whole think could be released.
hm, posted 6 Dec 2001 at 18:11 UTC by i0lanthe »
(Journeyer)
I know the Sphinx project here at CMU is now Open Source, but I don't
know whether they have perceived exciting benefits as a result (speech
recognition is way out of my area)... you would have to ask Kevin Lenzo
about it to find out, probably.
There is some bioinformatics software which is free. One is
the NCBI toolkit, which is developed by the NIH (a government
organization) so is in the public domain. Another is HMMER,
from Wash. Univ. St. Louis, released under the GPL. Some more
are EMBOSS, Artemis, SEALS. While not strictly academic
software, you might also look into bioperl, biopython, and
other projects supported by open-bio.org. Another place to
look is bioinformatics.org. This should be enough for you
to find other non-restrictive software.
You may also be interested in a proposal requesting that the
various public funding sources require grant recipients
release their software under an open source / free license.
This is at
www.openinformatics.org.
Despite my advocacy, I happen to disagree with the petition.
My reasons are posted in the last few weeks of the O'Reilly
bioinformatics mailing list, which is at
http://labs.ore
illy.com/pipermail/bioinformatics/.
X11 and CMUCL, posted 6 Dec 2001 at 20:36 UTC by jwalther »
(Journeyer)
X11 itself was released under the MIT license, which is Open Source.
CMUCL is released under GPL I believe. Both tremendously important in
their respective areas. In fact, I'd say they both dominate their
categories.
Kerberos, posted 7 Dec 2001 at 02:32 UTC by krftkndl »
(Journeyer)
MIT Kerberos is free software (the KTH implementation too). I believe
this has affected the project in a positive way.
Octave, posted 7 Dec 2001 at 14:07 UTC by shd »
(Journeyer)
Octave: http://www.octave.org
KRoC, posted 7 Dec 2001 at 17:14 UTC by azz »
(Journeyer)
KRoC, the Kent Research occam Compiler, was released
under the GPL (with libraries under the LGPL), and runs very nicely on
Linux.
MacAnova, posted 8 Dec 2001 at 18:54 UTC by Mulad »
(Apprentice)
MacAnova, another S-like
statistical package, is released under the GPL. I actually had trouble
finding the license when I downloaded it, but I contacted one of the
authors, and that's what he said.
phpwebsite, posted 9 Dec 2001 at 09:07 UTC by deekayen »
(Master)
A project that started up right around the same time as PHP Nuke:
http://phpwebsite.appstate.edu/
Greenstone is
digital
library
software from the
New Zealand Digital
Library, a research group the
University of Waikato. They also
have a
machine learning group who have released a package called
Weka.
Both are on-going
research projects by numbers of researchers; both are also used by real
people in the real world.
Ptolemy, posted 10 Dec 2001 at 18:52 UTC by jbuck »
(Master)
I helped start the Ptolemy
system-level CAD project, which has been extremely influential in the
research community. If it hadn't been free software it wouldn't have had
such an impact, in my opinion.
Definitions, posted 10 Dec 2001 at 23:38 UTC by Donwulff »
(Journeyer)
What is exactly meant by "Academic Software" in this context? If this
refers specifically to software used to directly further research -
statistics packages and such - the list may be fairly short. If it's
any software produced by research projects, or even any software
produced in the academia, then the list is quite long one. Open Source
has typically been its strongest among academic settings. BSD style
license has very deep roots in the academia. I'm personally involved in
development of X-Smiles prototype
XML-browser for "exotic devices", which may not be very popular, but
certainly succesful and ongoing effort. Altough the BSD license has not
garnered us many outside developers, it has allowed us to plug in ready-
made packages without which the project couldn't exist. It also means
we have to deal with outside decisions for packages we use, such as the
contention that embedded/handheld environments are not worth supporting.
Also, the question assumes these programs fall into only one of two
possible categories: Restricted to research, and Open Source. Quite a
few, possibly most, fall in the grey area in-between. They place
additional restrictions, not qualifying them to be Open Source by
official definitions, but still allowing widespread use. For an
example, the
license for TINKER, used among other things in the Folding@home
distributed project that's quickly gaining popularity. And, finally
ofcourse, there's the question of what constitutes as "succesful" ;)
From what my instructor, Lavy Libman, told me it is the Technion policy
that any projects done by the students as part of their Bachelor degree,
(and were not initiated by an industry firm) will be released under an
Open-Source license. Which license is left to the decision of the students.
For once, the IP-Noise
Simulator project which Roy Glasberg and I conducted, is such a case.
Both my Emu speech database
system and the more recent Annotation Graph Toolkit are
open source projects. There are a number of other tools in this area
(annotating linguistic data) which are open sourced, there's a useful
index at LDC, Upenn.
These tools are mainly aimed at helping Linguists work with large
collections of multimedia data. We're finding now that open source
efforts are helping standardise the tools and the data formats used
for annotation.
One of our reasons for beginning Emu was the high cost of the
standard
speech annotation software which only ran on unix workstations and
required a per-seat licence. This made it useless for teaching
anything but a handful of students. Our software now gets used in
teaching labs around the world as well as for research purposes.