License is missing

[jnnr]

This repo needs a license. I suppose we take the same license as for most other repos in oemof, GPL-3.0?

[p-snft]

Pick the most permissive, every author agrees to. Re-licensing to a less permissive one is easy.

[uvchik]

I would prefer MIT. See oemof/oemof#50.

[jnnr]

I am ok with MIT.

[jnnr]

@uvchik ✔️
@ckaldemeyer ✔️
@FranziPl
@jakob-wo ✔️
@jnnr ✔️
@Pyosch ✔️

[6] All
[5] MIT: @uvchik, @jnnr, @Pyosch, @ckaldemeyer, @jakob-wo

[uvchik]

I edited the list and added @Pyosch because of the pending PR #7

[jakob-wo]

Re-licensing to a less permissive one is easy.

Why is that?
I imagine it to be difficult to re-license a repo after having agreed on a permissive one. On one hand I expect heavy discussions if some contributers want a roll-back after some time. And on the other hand every one who starts including MIT-code now will be in trouble if later versions are released under a more restricted license. Don't you think so?

I am still not sure whether we should go along with MIT or GPL.

[p-snft]

Re-licensing to a less permissive one is easy.

Why is that?

It is possible to license additional contributions to a MIT codebase under e.g. LGPL and then to (re)distribute the new collection (as a whole) under LGPL. However, a given license cannot be revoked. So, whoever got the MIT-Licensed code can continue to use it. A well-known example is Cedega: The software marketed a Software based on a MIT-licensed codebase without upstreaming the improvements. Consequently, the upstream project (Wine) changed to LGPL.

[uvchik]

It is not allowed or at least a grey zone to use solph, feedinlib, windpowerlib in other programs that do have a more permissive licence than the GPLv3. Therefore it is not useful to have a restrictive licence in a basic library. The GPLv3 could be useful for program which is not intended to be integrated by others.

I is my goal that oemof can be used as easy as possible to create energy models and to publish these models under any licence and not only the GPLv3.

[p-snft]

Adding to that: The Free Software Foundation, that authored the GPL, advices not to use it for libraries. Instead, they created the aforementioned LGPL, which allows free usage of the library (as a whole) also for e.g. MIT-licensed models. When you, @jakob-wo, dislike the idea that someone applies changes to oemof(-thermal) but does not grant the same freedoms as we do, the LGPL is what you want. However, while the use of oemof(-thermal) would be possible, this would still limit usability of oemof(-thermal) code fragments in other scientific projects.

[jakob-wo]

Thank you both, @p-snft and @uvchik, for your statements and explanations.

I do support to license this repo under MIT.

[p-snft]

@FranziPl: I misinterpreted your consent in oemof/oemof#50 as one to thermal as well. I hope, this was Ok.