Hi, 2011/2/10 Frédéric DEMIANS <f.demi...@tamil.fr>: >> [2] >> http://dilettantes.code4lib.org/blog/2010/09/a-proposal-to-serialize-marc-in-json/ >> [3] http://www.oclc.org/developer/groups/marc-json-format-specification >> [4] >> http://robotlibrarian.billdueber.com/marc-hash-a-proposed-format-for-jsonyamlwhatever-compatible-marc-records/ > Why did you choose [2] JSON schema rather than one of the two others?
Because it's more concise than what the OCLC spec proposes, can be more easily handled by JSONPath expressions, and has gained the most traction thus far. Furthermore, the OCLC spec embeds an assumption about the ordering of control fields that is not met by some records found in the wild, most notably the FMT fields that are sometimes found in records coming from ALEPH 500 systems. As far as Bill Dueber's MARC hash proposal is concerned, he now favors the MARC-in-JSON spec. [1] All of that said, I'm certainly not wedded to having just one MARC::File::JSONxxx module. While I think Ross' version is the best of the lot, if OCLC has (or will have) web services that emit OCLC MARC-JSON, obviously it would be useful to support it. [1] http://www.mail-archive.com/perl4lib%40perl.org/msg01490.html Regards, Galen -- Galen Charlton gmcha...@gmail.com