On Mon, Mar 21, 2011 at 10:02 AM, martin <mar...@ics.forth.gr> wrote: > Dear All, > > Your comments on http://www.cidoc-crm.org/URIs_and_Linked_Open_Data.html > will be most welcome! > > Best, > > Martin > --
Dear Martin, I agree with Maximillian that overlapping URIs are inevitable for many reasons. At ISAW, we've already minted URIs for objects in our exhibitions that have never been so identified, and seemed unlikely to be so identified, by their curators. There are entries for objects in Wikipedia and in derived linked datasets like Dbpedia which are already widely used as vernacular identifiers for the physical objects [1]. The profusion of URIs has already happened. That said, it would indeed be a shame if museums lost out on the network effects of using the same URIs to refer to objects. There are some options for reconciling (owl:sameAs) and converging (HTTP redirects) and so the mess you refer to can be unpicked to some degree, but yes, it's best to start out emphasizing the benefits of reuse. On #8, I think that the web can withstand some URI overlap and indirection. If an object passes permanently from one institution to another, a new URI could be warranted. Information fetched from the previous URI will be out of date. The previous curators could keep the old URI and redirect to the new one as a part of the process of transferring the object. If this is exactly what you want to avoid, it might be worth considering the digital object identifier (DOI) scheme. I hope the proposal goes over well with institutions, thanks for writing it. [1] http://dbpedia.org/page/Liberty_Bell Best regards, -- Sean Gillies Programmer Institute for the Study of the Ancient World New York University