Diana's General Comments
- Standards: UW is always concerned about adhering to national standards. The MARC21 standard describes the 001 field: "The control number assigned by the organization creating, using, or distributing the record. The MARC code for the organization is contained in field 003 (Control Number Identifier)." I don't know that the 003 always reflects the source of the 001 in vendor records or not. This is also something to watch out for if we take a control number from another field, like an 035 or a 9xx. Something to keep in mind: we will essentially be creating our own local standard by choosing a control number and an alpha prefix (if the vendor does not supply one) and putting these in the 001s.
- Catalog interface environment: We are creating this 001 practice to ensure record merger in Summit. It should not adversely affect our local catalogs. But we need to keep in mind this is a Summit-specific solution. MARC records are increasingly appearing in multiple environments. Since these are non-OCLC records with somewhat strict licensing conditions, it remains unclear if they will ever be found in Google or added to OCLC group catalogs, etc. UW did contact OCLC and they could not foresee a problem with our plan to use the 001 field with alpha prefixes for non-OCLC records, as long as the 001 numbers were clearly distinguished from OCLC numbers. But who knows--we can't predict the future.
- Reversibility: Because we can't predict the future, the solution we choose should be reversible without too much cost or effort. I think this argues for trying to use the control numbers and any alpha prefixes supplied by the vendors as is whenever possible (hopefully they will already be found in 001s). The least editing, the better, not only for reasons of reversibility, but also local libraries' technical capacities (see below). It might also be more likely that other libraries outside of Summit would be using the same vendor numbers as is. (In fact, the library community as a whole might like to lobby vendors for unique, stable control numbers that would be the same for all customers and yet would also be unique across record sets and across different vendors. A pipedream?)
- Technical capacities: We are all limited by INN-Reach software, which can only reliably match on an 001 right now. There is also the question of the load capacity for local to central updates (the UW would have to update tens of thousands of records for this project, and we are also in the midst of a MARC holdings conversion project that will entail tens of thousands of more updates). The pipeline to Summit might be a real roadblock for the larger libraries when it comes to retrospective cleanup of non-OCLC record 001s. Is there some adjustment that can be made to the Summit "pipeline"? There is also the question of local libraries' capacity to edit and customize vendor records, so that they can comply with whatever standard we create. I like the idea of support for loader editing through the Alliance. Also, member libraries could help provide advice for MARCedit.
- Coordination/communication: I think it would be best to have this plan coordinated somehow through the central ORBIS offices, as Nancy described. A central office/person to keep track of the control numbers and prefixes for each record set (could also be documented on the Web). Also each set that is new to the Summit system should be reviewed and vetted by someone, at least in relation to control numbers. Which number to choose from the records? Are the 001s OK as is, or do they need to be edited? Someone to contact the vendor and establish if the numbers are truly unique and stable and so on.