Excuse the cross-posting, please. We are setting up a pilot archive. We know that over time we are very likely to re-implement the archive with the current (DSpace) or other software. Each re-implementation means that the string to access a given asset changes, and would involve some reworking (or a long redirect table/complex redirect algorithm) between the front end interfaces and the archival objects.
If we were publishers, I'd simply register a Digital Object Identifier (DOI) for each item (although, given enough items, those charges add up) and move on, and write our interfaces to access objects via DOI. Is there an equivalent to DOIs for non-text digital objects? What are other institutions doing, or is this a rare enough problem that people have dealt with it with the aforementioned rewrites on a case-by-case basis? Thanks, Ari Davidow
