Licensing issues

[VarLad]

This issue has been created to discuss the license of this library further going on. No new PRs will be made by the author of this issue, before this issue is resolved

cc: @ChrisRackauckas

[ChrisRackauckas]

What licensing issues could there be? If I am understanding correctly, the code you have contributed here is not a copy of some other code and is an implementation of the methods from the papers. Can you please confirm this?

[VarLad]

Yes.

But the algorithm published in the paper is an IP of Qu&Co, was my concern.

[VarLad]

The concern was also raised by one of the authors of the paper @vincentelfving

[ChrisRackauckas]

That is unrelated to software licensing. Since you did an independent implementation of the method from paper, you have the copyright ownership of the software, which can then be assigned to SciML and licensed however.

Now, there can be an issue of use if Qu&Co have a patent on certain applications of software (since software itself cannot be patented, it would be the use in some way or some application). Software licensing is copyright which is separate from patents and in no way can disallow you from assigning copyright ownership of your independently created software, and from licensing that how you wish. However, this may mean that certain users may need to be aware that if they use this software for some applications, they may be violating possible patents.

Certain open source licenses have patent grants, though those only apply to the patents of the copyright holders and contributors (because of course you cannot grant a patent that you do not own 😅). That said, the MIT license is not even a license with a patent grant, so (a) it makes no mention that you or any contributor is granting patent licenses to any user, and (b) currently you and all other contributors have no patents on the subject to grant. That said, what may come up is that maybe someone from Qu&Co may some day be interested in contributing. If that is the case, we could add a note to the license to make it explicit that no implicit patent licenses are granted to the user, or only patents owned by SciML are granted to the user (of which there are none), and it's up to the user to determine if they have the rights to use the software for their application beyond the copyright.

But since there is no contributor with a patent for which implicit patent assignment would be relevant, at least right now this is a non-issue. And because you created the software, there is definitely no "licensing issues".

Some basics you may want to read:

• https://opensource.guide/legal/
• https://google.github.io/opencasebook/patents/#:~:text=A%20patent%20gives%20its%20owner,and%20selling%20the%20claimed%20invention.&text=In%20contrast%2C%20open%20source%20licenses,distribute%2C%20and%20use%20the%20software.

[VarLad]

Resolved! Thanks to everyone involved for their time!