Reconvert a batch of 2000 MARC records to RDF, in order to replace a previously converted but otherwise unedited linked data-based record with a new graph based on an enriched MARC record, and with a new reconciliation process downstream.. This is TB1a with conversions early and often. The important implication here is the behavior of the converter when something is being reconverted. Is data clobbered? If data is missing in new record, do you do a delete? Do you save everything and version it? Do you just update changed fields? Do you let people choose with infinite flexibility and no constraints?
Reconvert a batch of 2000 MARC records to RDF, in order to replace a previously converted but otherwise unedited linked data-based record with a new graph based on an enriched MARC record, and with a new reconciliation process downstream.. This is TB1a with conversions early and often. The important implication here is the behavior of the converter when something is being reconverted. Is data clobbered? If data is missing in new record, do you do a delete? Do you save everything and version it? Do you just update changed fields? Do you let people choose with infinite flexibility and no constraints?