Add use guidelines

[znicholls]

Adds usage guidelines to OpenSCM

• Implementation finished (N/A)
• Tests added (N/A)
• Documentation added (N/A)
• Example added (in the documentation, to an existing notebook, or in a new notebook) (N/A)
• Description in CHANGELOG.rst added (single line such as: (`#XX <https://github.com/openclimatedata/openscm/pull/XX>`_) Added feature which does something)

[rgieseke]

Good idea to have some "expectation setting" in the Readme/Docs!

My 2-cent:

AGPL is not a "permissive" license in the usual usage in the OS community but a "copyleft" license i think (and MIT/BSD are permissive).

I strongly dislike co-authorships suggestions in other projects so would refrain here, too (though i haven't been much involved recently i probably should count much on this issue) - i'd think it's fine to notice that you're open to collaboration, but the license allows anyone to use it according to the license.

In OpenSCM's case it's also the case that it builds on other people's work heavily, so it's even more difficult whom to include. For PRIMAP-hist we went with

 When using this dataset, or one of its updates, please cite the DOI of the precise version of the dataset used and also the data description article to which this dataset is supplement. Please consider also citing relevant original sources when using the PRIMAP-hist dataset. References to the original sources can be found in the mansucript and on the dataset download page. 

So something like this might be appropriate for the SCMs shipping with OpenSCM in the future. I dislike lots of things about IIASA's SSP databases but i think them providing clear citation instructions and downloadable citation data is pretty great!

[swillner]

I'm also not too much a fan of such notices. Maybe a short notice suffices that "we are open about collaboration"? I suppose, a nice CITATION file (with a specific formatted citation to copy-and-paste) would be of more value.

[znicholls]

Ok. So how about we go for:

• notice in similar vein to PRIMAP-HIST (i.e. please consider citations of us and original source, open to collaboration)
• CITATIONS file, in which we have citations for each model plus OpenSCM

?

[swillner]

That would be awesome ;)

[swillner]

Sorry, I missed your understanding of the CITATION file :/ It's supposed to include the way we would like OpenSCM to be cited (hence the singular).
Nevertheless, we should add a cite method to the adapter including those model citation.

[swillner]

... I like the new encouraging colaboration paragraph! Can you sort the maintainers alphabetically again, please?

[znicholls]

Nevertheless, we should add a cite method to the adapter including those model citation.

Should we also provide a file which has all the citations in one (pulling citations from all over the place is one of the things I hate most when writing papers)?

[znicholls]

Can you sort the maintainers alphabetically again, please?

Done

[swillner]

Nevertheless, we should add a cite method to the adapter including those model citation.

Should we also provide a file which has all the citations in one (pulling citations from all over the place is one of the things I hate most when writing papers)?

Maybe a function that collects all cite values from all available adapters?

[znicholls]

Maybe a function that collects all cite values from all available adapters?

Just feels unnecessarily complex, why not just have a file?

[rgieseke]

I think R has some standards to get a citation but i'm not aware of any default ways in Python. I think having a file (or just in the Readme) should be easier.

[swillner]

Ok, then maybe rename CITATIONS to CITATION and add a way to cite OpenSCM on top with an explanatory text and then give the other citations?

[znicholls]

Ok, then maybe rename CITATIONS to CITATION and add a way to cite OpenSCM on top with an explanatory text and then give the other citations?

Like so? Given we don't have a citation yet this is the best we can do right?

[swillner]

Nice! Just two minor suggestions