4

Is anyone aware of (work towards) a license with similar freedom-preserving restrictions as the AGPL, which can be applied to datasets?

Some background:

The GPL puts restrictions on the legally allowed uses of source code to protect your freedom, that is, the users keep control over the program. The source code cannot evolve into having control over you.

I wonder if the same things exists for datasets.

For example, if you base a service on such licensed dataset and say your service improves the dataset, you must contribute this back. If you enrich a dataset, you must license the enrichment under same license, etc.

  • I've put this question on hold because it doesn't contain sufficient detail for us to recommend you the license you want. See How do I ask for a license recommendation? for more detail. – ArtOfCode Oct 20 '15 at 17:28
  • If I read the question correctly, they want copyleft, and oblige full access to the dataset when it can be interacted with. – Martijn Oct 20 '15 at 17:42
  • Thanks for the comments, let me read up. I realise I am not exactly sure either, it might be more subtle and maybe CC covers the case already. – wires Oct 20 '15 at 17:57
  • Interesting word, oblige didn't know about it @Martijn, thanks! I've updated the question, hope this is clearer, I could expand more if needed. Just ask – wires Oct 20 '15 at 18:09
  • @wires nice edits - reopened – ArtOfCode Oct 20 '15 at 18:48
4

Datasets are difficult.

They pose no originality, and this is handled differently in different jurisdictions. Some jurisdictions have sui generis database rights, which are handled similarly to, but distinct from copyright.

In the United States for example, databases are not under any protection at all. No license can be attached to it, because it holds nothing that can be licensed in the first place.

The EU does recognize database rights, with a term of 15 years after first publication.

Different countries have different laws, and it will be almost impossible to write a license that would be effective across different jurisdictions.

Having specific license terms that interact with database rights and copyright would be an even more difficult task. It would, in theory, probably be possible to draft a license that places restrictions on software that interacts with datasets, or places license restrictions on datasets that incorporate your data set that would be valid for a specific country, or for the EU, but as far as I know, this hasn't been tried.

Open datasets are generally handled under a CC Zero public domain dedication. Despite there being plenty of problems with CC Zero, it least it's somewhat portable, in that it means roughly the same across jurisdictions.

  • Databases can be original, but then it's really the contents of the database which are copyrighted. For example, Wikipedia's database contains articles under CC-BY-SA 3. – Kevin Oct 22 '15 at 11:51
  • 1
    A reasonable argument can be made that the selection of articles also is a creative work. As are things like top 100 best lists. – Martijn Oct 22 '15 at 15:55
3

Does the Open Database License (ODbL) solve a part of your problem?

If you alter or build upon our data, you may distribute the result only under the same licence.

It is in use by OpenStreetMap, and a.o. ensures coverage of the world gets more granular over time.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.