Le Sat, Dec 29, 2012 at 08:42:41PM +0100, Steffen Möller a écrit : > > The GO license I read as "scientific integrity". Yes, as a consequence you > cannot modify the database at will or it is not GO any more and you cite it > where you use it. IIRC some two GO terms or so I had at some point suggested > myself, so changes _can_ be made and one is even helped to get it done > consistently. For other views on the world, have all the freedom of the > world to start your own ontologies, and many are doing so.
Hi Steffen, The problem would definitely be solved if license terms would allow to modify the data provided that the name is changed, in line with DFSG#4, but I could not find such a permission on GO's website. I considered proposing this in the message I sent to GO, but in the end, this is a practice that I do not want to recommend, as things would become very impractical if such clauses would become the norm for for software and data (this is why I am also very sceptical with trademark restrictions). I think that clauses requiring "proeminent notices" are a good compromise, and data integrity checks (signed checksums, etc.) are the best way to prevent potential problems. But I expect quite a variety of opinons on the subject, even within Debian Med. We should better speak of one voice, not only with GO, but also with the other ontology providers, so to the other developers: please express yourself if you wish, so that we can build a consensus. Cheers, -- Charles -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
