Re: [general] regarding permission

On 5/20/20 3:41 PM, Mudit Sharma wrote:
> Dear All,
> 
> I am interested in designing a mathematical computation package for
> Octave. Hence I have collected a lot of reading material for this purpose.
> 
> However I am a bit confused where :
> I want to know if I can use the formulas or some algorithms present in
> the text ( Springer books, research papers ) in my code and then present
> it in Octave as open source  ( cite them ), is this method correct or do
> I need any kind of permission from the publisher or the authors ?
> 
> I want to ask the if the experienced authors of libraries/packages know 
> all the maths/algorithms when they designed their libraries/packages or
> they took help from other sources like me ?
> 
> Any response is appreciated !
> 
> Best wishes, Mudit


Dear Mudit,

Thank you for considering Octave to create your package.  As you
correctly see, none of those many mathematical software libraries
available today invented all their contained algorithms (sorting,
matrix-multiplication, etc.) from scratch.  Same for Octave in many cases.

First let me answer the question regarding using "ready-to-use" code
from a publication.  This strongly depends on the journal where it was
published.  For example, many algorithms are published in "ACM
Transactions on Mathematical Software (TOMS)".  They have a copyright
notice [1] which prohibits commercial use without permission by the ACM.
 In this case, using that "ready-to-use" code in your package will
automatically prohibit its usage for commercial applications as well.
This makes it impossible for Octave to use these algorithms.

==> Watch out for the journals copyright notices.

Second answer is regarding formulas, abstract algorithms, or ideas from
a publication.  To my humble knowledge, converting them into code is
less restrictive, as you are authoring/creating new work.  Of course you
should not copy large text sections of the publication into your code
comments.  There may be other exceptions, but I am no lawyer =(  In any
case you should definitely cite those papers, as it is good academic
practice.  Experienced users can easily get the idea of your
implementation, and you credit original work.

In case of doubt, it might be useful to contact the original authors.
Maybe they have a proof-of-concept implementation floating around
somewhere, ask them to publish it (on their own homepage, GitLab, ...)
under the terms of the GPLv3 [2] and you are on the safe side.

HTH,
Kai

[1] https://www.acm.org/publications/policies/software-copyright-notice
[2] https://www.gnu.org/licenses/gpl-3.0.en.html