Publishing datasets with multiple licenses

I’m moving the conversation here from the old google-group mailing list (Thanks Kelly for reminding me of this forum!) to see if there are any other comments on the subject.
Thanks for your suggestions.

When publishing a dataset that contains multiple objects/files some of which are code using OSI licenses and some of which are datafiles with CC0 waivers and some images with CC-BY licenses how should one add that information to the datacite metadata?

Would it be appropriate to add a license.txt file to the dataset listing all the different licenses of the various objects and then add this to the DataCite metadata:

Open Access

bespoke licence file

or is there a better way to do it?

If this is covered somewhere in the documentation please point me to it as I couldn’t find it when I was looking.


Kelly replied with:
Hi Chris,

Great question! I would encourage you to post this to the PID Forum—you might get more responses there in the DataCite
Chat Room (We’re more
active over there lately!)

I think including a license.txt file with the dataset is a good idea. In
addition, you can repeat the rights property with the different licenses
(OSI, CC0, CC-BY, etc) too, if you want those licenses to be visible in the
metadata. However, as this does not allow for distinguishing which
license(s) apply to which file(s), having the license.txt file is still key.