On Wed, Sep 2, 2009 at 8:38 PM, graham <[email protected]> wrote:
> Hi, > > The problem with identifier systems is that they normally require a trusted persistent authority which manages the uniquification. It would be technically possible to canonicalize the bibliographic metadata (but you would have to ask the experts) and then create a GUID or other hash. This is technically not unique but very unlikely to collide. The real problem is the canonicalization I sympathize with the spirit of your wish. P. -- Peter Murray-Rust Reader in Molecular Informatics Unilever Centre, Dep. Of Chemistry University of Cambridge CB2 1EW, UK +44-1223-763069
_______________________________________________ okfn-discuss mailing list [email protected] http://lists.okfn.org/cgi-bin/mailman/listinfo/okfn-discuss
