On Mon, May 20, 2013 at 1:41 PM, J. McRee Elrod <[log in to unmask]> wrote:
>Inadvertently creating two Instances of the same item, one with the 10
>digit ISBN, and one with the 13 digit one, would be *very* frustrating.
Whatever problems Bibframe is going to have with ISBNs, this is unlikely to be one of them. It's trivial for a computer to calculate that a 13-digit and 10-digit ISBN are the same. In WorldCat, all the ISBNs are paired. If you enter a 13-digit ISBN in Connexion, it automatically adds the 10-digit one and vice versa. If you enter a numerically invalid ISBN, it automatically moves it to $z. So in Roy's list, the records with two 020s really only have one ISBN with both versions recorded. The records with odd numbers of ISBNs presumably represent some failure of OCLC's validation routine.