We decided that the wiki should be publicly accessible and canceled the previously scheduled meeting for 10/2. Next meeting is 10/7.

We started an initial review of the persistent identifiers inventory and discussed some of the approaches that arXiv has taken:

• arXiv has persisted its URLs over time - the old ones continue to work
• arXiv authors have unique internal identifiers - which they expect to expose

We then expanded to a more general discussion:

• There needs to be an easy way to declare author equivalence - how should we relate to OCLC author identities?
• The solution must support multiple schemes for naming
• PIDs should be content negotiated
• VIVO supports both NetID and employee IDs - we will certainly need a PID for each NetID at a minimum

Should we just plan to support "string: arbitrary_garbage", where we could map old DNS names to the "string" part. This could potentially handle a number of existing id schemes in a somewhat uniform way.

Our overall approach will be to continue for a bit longer in exploring the breadth of the problem and looking at existing systems. Next we will brainstorm an initial strawman proposal for discussion.