On Feb 4, 2013, at 11:52 AM, Greg Logan <greg.lo...@usask.ca> wrote: > In regards to the rest of your email, are you proposing this for 1.4, or > 1.5? I was unable to attend your session, so I'm a bit out of the loop!
Well, I would consider it a bug that Matterhorn ingest treats series catalog different than episode catalog. I think this is because series catalog has to be processed in a different place than all the other catalogs since it has to exist before the zipped ingest file can be processed. The other catalogs are processed in the same code loop, later on. I am concerned that what is a bug for us may be a feature for others. However for us, yes! we'd love it to go into 1.4. It means we do not have to be concerned about capture agents properly marshaling and unmarshaling the series catalog. > … worth its salt should > be unpacking them and ingesting them unmolested. There should be no > modification of the data unless the agent has taken new input (from a > command console, or whatever). It most certainly should not be losing > data just because it was passed through … In all fairness, we have tried to pass series elements directly into Matterhorn through the series REST endpoint. Including dublin core elements outside of the 28 dublin core elements identified in Matterhorn dublin core class code. The hope was that these elements would hibernate unmolested and buried in the catalo, unused by Matterhorn, but could be parsed by a third party when the whole catalog was retrieved. However, elements outside those 28 were stripped from the series catalog when Matterhorn wrote the catalog to disc. A few months ago, we experienced the same ummarshalling/marshalling behavior with Galicaster. Where Galicaster parses the series and episode catalogs, probably to use the data their interfaces and allow it to be modified. But then they re-marshals the catalogs from those objects. Their objects have less than the 28 dublin core supported by Matterhorn. The series gets updated by Matterhorn, but not the episode, so we have data loss in the series. We have not had any issues with NCast ingests. I don't know if NCast unmarsalls the series catalog or if it bother's to repackage it for the return journey, which is not necessary if the series already exists in Matterhorn. I hope Galicaster and Matterhorn both uncover a creative technique for marshaling & marshaling catalogs that contain foreign data ; ) In the meantime, those 4 lines of code from MH-9281 would help. - Karen _______________________________________________ Matterhorn mailing list Matterhorn@opencastproject.org http://lists.opencastproject.org/mailman/listinfo/matterhorn To unsubscribe please email matterhorn-unsubscr...@opencastproject.org _______________________________________________
