The important part of copyright is originality (not creativity). The test would be whether there was more than one way to express the information and whether there was originality in the expression chosen.
A court looking at the copyright status of the timezone database would look for originality in the selection of the identifiers. Why America/New_York and not America/NY or America/NYC or even America/East_Coast? Someone made a decision to choose one form over another so there is originality, hence it is copyrightable. The same applies for other systems such as Dewey Decimal classification or the Getty Thesaurus of Placenames.
That said, it seems the problem is not in the copyright status of the timezone database but in the fact the compilers used copyrighted work to derive their information. The assumption seems to be that makes it a derived work. I'm not so sure about that.
The existence of timezones are uncopyrightable facts. The boundaries of those timezones are uncopyrightable facts. The naming of the timezones is copyrightable by the database compilers not by the atlas owners unless the compilers copied the timezone names from the atlas. Anyone got a copy to check that?
A court looking at the copyright status of the timezone database would look for originality in the selection of the identifiers. Why America/New_York and not America/NY or America/NYC or even America/East_Coast? Someone made a decision to choose one form over another so there is originality, hence it is copyrightable. The same applies for other systems such as Dewey Decimal classification or the Getty Thesaurus of Placenames.
That said, it seems the problem is not in the copyright status of the timezone database but in the fact the compilers used copyrighted work to derive their information. The assumption seems to be that makes it a derived work. I'm not so sure about that.
The existence of timezones are uncopyrightable facts. The boundaries of those timezones are uncopyrightable facts. The naming of the timezones is copyrightable by the database compilers not by the atlas owners unless the compilers copied the timezone names from the atlas. Anyone got a copy to check that?