We are using KOHA 3.0 ubuntu. It works fine with english records. The problem is that it does not retrive Urdu and Arabic records correctly
This is a very tricky problem related to encoding and z39.50. You have stated only the most basic summary of the problem. For a full analysis we would need to know: - What MARC "flavour" is your Koha catalog: MARC21 or UNIMARC? - Can you catalog original records (i.e., not downloaded) in Urdu or Arabic yourselves? This would demonstrate that Koha itself is setup OK. The rest of the questions would have to do with your z3950 target: - MARC21 or UNIMARC? - What encoding is the z39.50 connection: MARC8 or UTF8? If the target is publicly available, please supply the information we would need to test connecting to it. We might not be able to read Urdu, but we can still detect if the encoding is invalid. -- Joe Atzberger LibLime - Open Source Library Solutions