[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Usage of MersenneTwister.h in CRAN bipartite (Was:r-cran-bipartite_2.15-1_amd64.changes REJECTED)

Dear Andreas,

TL;DR: I think it's fine, and I will continue claiming it is compatible with GPL.

Well, all I found, on the internet, is that this software is used all over the place, with various licenses, and I managed to track down "Richard Wagner": he died recently (https://www.legacy.com/obituaries/telegram/obituary.aspx?pid=196507520). I have no idea what that means for his code legacy, however.
I found this webpage (https://public.kitware.com/pub/itk/Examples/src/Core/Common/MersenneTwisterRandomIntegerGenerator/Documentation.html), which to me suggests that the code is CC-BY-.. licensed (see bottom of the page), although they reproduce exactly the same license statement. It is hard for me to see what that actually means, legally.

Also, did you notice that at the very bottom of the file is says that the license (in version 1.0) was changed to BSD?
// v1.0 - Revised seeding to match 26 Jan 2002 update of Nishimura and Matsumoto
<snip>
//      - Changed license from GNU LGPL to BSD

The audio-software Essentia also uses exactly this file and is Affero GPLv3-licensed (https://essentia.upf.edu/licensing_information.html)
(So does "open source" SUMO, but it does not provide any license information: https://sumo.dlr.de/daily/doxygen/index.html)

So, in summary, I suggest that this file is "permissively" licensed, and therefore "safe" to distribute also for Debian. I am, of course, ignorant both of the legal details and the strictness of Debian engineering. And of course that everybody else doesn't care doesn't mean it is correct to do so.

Cheers,
Carsten


On 7. Nov 2020, at 19:24, Andreas Tille <andreas@fam-tille.de> wrote:

Dear Carsten,

thanks a lot for your quick response.

On Sat, Nov 07, 2020 at 07:20:47PM +0100, carsten.dormann@biom.uni-freiburg.dewrote:
To be honest, I had scanned the licence statement in that file, but not noticed the notice.

The code in bipartite is based on code by Aaron Clauset, who himself used the Mersenne Twister to avoid the relatively short random number period. We (Rouven Strauss and I) did not touch that bit of code at all; our work was solely on the bit produced by Aaron. His code is available publicly here: http://tuvalu.santafe.edu/~aaronc/hierarchy/<http://tuvalu.santafe.edu/~aaronc/hierarchy/> 

I never wrote to any of the emails listed. It is indeed somewhat awkward and I have no idea what to do about it.

Wikipedia does not square with the notice you found: https://en.wikipedia.org/wiki/Mersenne_Twister<https://en.wikipedia.org/wiki/Mersenne_Twister>, calling it "permissively licensed".

Given Debian's merit of being very careful with such licenses, I can only ask you for suggestions: is there anything I can do, you think, to push the Mersenne Twister towards a GPL-like license?

I think the only way is to convince the copyright holders to drop that
paragraph and use one of the known licenses (not necessarily GPL even if
this would probably a nicely fitting choice for your own code). 

Kind regards

     Andreas.

On 7. Nov 2020, at 17:59, Andreas Tille <andreas@fam-tille.de> wrote:

Hi,

I intend to package bipartite for Debian since some R package needs
it to run its test suite.  Our ftpmaster who is checking licensing
issues stumbled upon the file src\MersenneTwister.h.  It contains
an additional clause to the license which says:

// The original code included the following notice:
//
//     When you use this, send an email to: matumoto@math.keio.ac.jp
//     with an appropriate reference to your work.
//
// It would be nice to CC: rjwagner@writeme.com and Cokus@math.washington.edu
// when you write.

Would you please clarify how far the "original code" is changed?  Its
also hard to interpret what the "send an email to" might mean.  Should
every user who is using the bipartite package send an e-mail to the
authors?  That's really hard to realise since we can not guarantee that
every user will read the copyright statement of all software that is
installed.  So some clarification what that clause might mean would be
appreciated.

Kind regards

   Andreas.

----- Forwarded message from Joerg Jaspert <ftpmaster@ftp-master.debian.org> -----

Date: Fri, 06 Nov 2020 23:00:08 +0000
From: Joerg Jaspert <ftpmaster@ftp-master.debian.org>
To: Andreas Tille <tille@debian.org>, Debian R Packages Maintainers <r-pkg-team@alioth-lists.debian.net>
Subject: r-cran-bipartite_2.15-1_amd64.changes REJECTED


Hi Maintainer,

rejected, sorry, but src\MersenneTwister.h is not GPL.
Also, how much is that code there based on "the original code"? See
the paragraph directly below the license. The requirement there is
*NOT* nice for inclusion in Debian.

--
bye Joerg



===

Please feel free to respond to this email if you don't understand why
your files were rejected, or if you upload new files which address our
concerns.


_______________________________________________
R-pkg-team mailing list
R-pkg-team@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/r-pkg-team

----- End forwarded message -----

-- 
http://fam-tille.de

— — — — — — 
Carsten Dormann

Professor of Biometry and Environmental System Analysis
Faculty of Environment and Natural Resources
University of Freiburg
Tennenbacher Str. 4
79106 Freiburg

Tel.: 0049 (0)761 203-3750
Fax.: 0049 (0)761 203 3751
Email: carsten.dormann@biom.uni-freiburg.de


-- 
http://fam-tille.de

— — — — — — 
Carsten Dormann

Professor of Biometry and Environmental System Analysis
Faculty of Environment and Natural Resources
University of Freiburg
Tennenbacher Str. 4
79106 Freiburg

Tel.: 0049 (0)761 203-3750

Reply to: