[Fedora-legal-list] Can autotalent be packaged for Fedora?

David Cornette david at davidcornette.com
Fri May 14 19:27:29 UTC 2010

I am thinking of packaging autotalent
<http://web.mit.edu/tbaran/www/autotalent.html> for Fedora, but I first wanted
to make sure that license and copyright permitted it.

The source distribution contains the following files:

The README file says "By Thomas A. Baran", but does not say anything about
copyright or license.  The file COPYING states the following, which is followed
by the text of the GPLv2:

The bulk of the code in Autotalent is released under GPL2.  However,
the FFT routine was taken from PureData
(http://crca.ucsd.edu/~msp/software.html), which was released under a
license that is similar to the BSD license.  So with the exception of
the mayer_fft.* files, everything should fall under GPL2.



The files autotalent.c and Makefile both state that they are under GPLv2+.
The file ladspa.h is under LGPLv2.1+

The file mayer_fft.h has no attribution, copyright or license information at
all.  The file mayer.c has this to say:

/* This is the FFT routine taken from PureData, a great piece of
software by Miller S. Puckette.
http://crca.ucsd.edu/~msp/software.html */

** FFT and FHT routines
**  Copyright 1988, 1993; Ron Mayer
[Documentation of functions and their signatures omitted]
** NOTE: This routine uses at least 2 patented algorithms, and may be
**       under the restrictions of a bunch of different organizations.
**       Although I wrote it completely myself, it is kind of a derivative
**       of a routine I once authored and released under the GPL, so it
**       may fall under the free software foundation's restrictions;
**       it was worked on as a Stanford Univ project, so they claim
**       some rights to it; it was further optimized at work here, so
**       I think this company claims parts of it.  The patents are
**       held by R. Bracewell (the FHT algorithm) and O. Buneman (the
**       trig generator), both at Stanford Univ.
**       If it were up to me, I'd say go do whatever you want with it;
**       but it would be polite to give credit to the following people
**       if you use this anywhere:
**           Euler     - probable inventor of the fourier transform.
**           Gauss     - probable inventor of the FFT.
**           Hartley   - probable inventor of the hartley transform.
**           Buneman   - for a really cool trig generator
**           Mayer(me) - for authoring this particular version and
**                       including all the optimizations in one package.
**       Thanks,
**       Ron Mayer; mayer at acuson.com

/* This is a slightly modified version of Mayer's contribution; write
* msp at ucsd.edu for the original code.  Kudos to Mayer for a fine piece
* of work.  -msp


The license for Pure Data may be found at
<http://www-crca.ucsd.edu/~msp/Software/LICENSE.txt>, and it is indeed a 3
clause BSD type license.

However, it is not included in the source distribution of autotalent, and the
ownership of mayer_fft.c seems less than clear to me.  Can autotalent be
packaged as is?  If it can be packaged, do we need to include the Pure Data
license file as a second source file in order to comply with that license? Or
does "If it were up to me, I'd say go do whatever you want with it" mean we
can do whatever we want with it?

Also, the patents in question appear to be U.S. Patents 4646256 and 4878187,
which, if I understand correctly, have expired.

Thanks for giving this your attention.

David Cornette

More information about the legal mailing list