Why MIRIAM URIs are URNs and not URLs?
Initially, MIRIAM Resources was managing (up to) two official URIs per data type: one URN and one URL. These URIs were based on some existing identifiers (for example LSIDs) and some domain names. This can still be observed in the list of deprecated URIs we maintain for the data types. However, it quickly appears that there were various issues with those unconsistent URIs. Moreover users where confused and lost with the difference between the URL as an identifier and the URLs as the resources’ physical locations. During the Super-hackathon about Standards and ontologies for Systems Biology in Okinawa, Japan (January 2008), we presented the set of issues with those URIs. A session was held and we discussed various possible solutions. The consencus lead to the current URN scheme. The main reasons for this choice were: • Maintaining various URIs per data type is counter productive and defeats the purpose of an identifier. • For a location independent identifier, a URN makes more sense: a
