Licensing for vocabularies

[baskaufs]

The current default license for standards documents is specified as CC BY. However, I think this would probably be a mistake for vocabulary term lists. Do we really want to imply that every user of terms on the list must provide attribution? Shouldn't it be CC0? Is this important? If so, there probably needs to be a policy decision made on this by the Executive. See section 4.2 of the documentation spec

[ramorrismorris]

This sounds like an intellectual property(IP) attorney should weigh in on this issue and address such things as whether use of terms may be fair use of a copyrighted document that contains the definition of the terms. Also, there may be other kinds of IP that are relevant. The use of Adobe xmp by Audubon Core comes to mind in that Part1 of the XMP spec is copyrighted. I sure hope fair use is a simple question. I don't believe there is much merit to arguments of the form "it is illogical for this to be an issue" when uttered by a non-lawyer.

[jar398]

Thinh Nguyen, an attorney who worked for Science Commons, favors CC0 http://sciencecommons.org/resources/readingroom/ontology-copyright-licensing-considerations

Mike Linksvayer of Creative Commons also favors CC0, arguing that CC-BY is pointless http://gondwanaland.com/mlog/2013/11/25/upgrade-to-0/

John Wilbanks likes CC-BY for ontologies http://ontolog.cim3.net/file/work/IPR/OOR-IPR-01_IPR-landscape_2010-09-09/licensing-n-ontologies--JohnWilbanks-CC_20100909.pdf

Note that CC-BY version 3 is superior to version 4 in that it waives database rights (while retaining copyright).

The risk of CC-BY is that a potential user will be scared away because they may wrongly think an attribution requirement applies to a use of the terms. (It wouldn't, it would only apply to copying the whole ontology.)

Alan Ruttenberg favors 'URL as attribution' - the interoperability goal is to get everyone to write and use the same terms in the same way, so requesting the term URL be necessary and sufficient for attribution uses the CC-BY attribution requirement as a lever to help prevent people from changing the spelling of the terms (as a certain well-funded ontology warehouse did at one point).

Trademark protection + CC0 is an interesting alternative mentioned in Thinh's article. This was roughly the approach taken by the US government for the Ada programming language specification.

The CC-BY vs. CC0 argument has been stewing for many years in the OBO community with no resolution. Hilmar Lapp notes that CC-BY is not really tested or accepted for ontologies, and favors CC0. Hilmar probably knows more of the history of this topic than anyone else in TDWG circles, so it might be wise to consult him.

http://obo-discuss.2851485.n2.nabble.com/Ontology-licensing-td5914012.html https://github.com/JervenBolleman/FALDO/issues/7

I doubt you'll find a suitable attorney for hire, since you want someone who cares about the public interest, and lawyers who are able to conceive of interests other than private ones are very rare. You might find a volunteer though.

[baskaufs]

In order to keep this issue from blocking the completion of the Standards Documentation Specification draft, in Section 4.2 I have changed the "Type of value" text for "License" to "use a license type in accordance with current TDWG policy". Personally, I think vocabularies should be explicitly licensed as CC0, but it's not up to me to set this policy. If there is a policy on this by the time the specification works its way through the Standards Process, we can designate specific licenses in this section again.

In order to keep this issue on the table, I'm not going to close it. However, I'm going to remove it from blocking completion of the documentation spec draft.

[peterdesmet]

Hi all, didn't see this discussion. I've also been working quite a lot with licenses and suggested to the exec (2 years ago I think) to use CC0 for all TDWG standards. It was then decided to use CC-BY 4.0 for all TDWG content. One reason for this is that a discussion about licenses is almost always mixed with a discussion regarding receiving credit for your work (i.e. citations), even though both should be seen separately.

I still think CC0 would be the best option for standards and vocabulary terms in specific: I will raise this again with the exec (see also https://github.com/tdwg/infrastructure/issues/75).

Note on CC-BY 4.0: URL attribution is often enough, fair use doesn't apply in all jurisdictions + applying such a license doesn't guarantee people will cite the resource.

Also, @jar398 :

Note that CC-BY version 3 is superior to version 4 in that it waives database rights (while retaining copyright).

From my reading of CC-BY 4.0, you ARE granted to extract, reuse, reproduce, and Share all or a substantial portion of the contents of the database. So sui generis database rights cannot be invoked to restrict use?