This week, I spend more time on OAI-PMH. I try to figure out what is the best way to harvest the records which we have deposited on DSpace. It seems that some libraries have tried to use software XML Harvester to harvest non-MARC records. This commercial software mostly has been integrated into ILS, and facilitates to harvest and convert Non-MARC records to MARC records.
Free software MarcEdit can also harvest NON-MARC records, but at some point, it has some problems. I am still exploring to make it work at my library. Hopefully, I can post the process later.