With the release of AvroTurf 1.0 and this PR https://github.com/dasch/avro_turf/pull/111 we should now be able to cache schemas across the full execution rather than having one schema cached per producer/consumer. We should be able to have a singleton encoder/decoder now.
The previous problem is that we'd have multiple schemas defining the same subschema (e.g. "Status") and AvroTurf would crash.
With the release of AvroTurf 1.0 and this PR https://github.com/dasch/avro_turf/pull/111 we should now be able to cache schemas across the full execution rather than having one schema cached per producer/consumer. We should be able to have a singleton encoder/decoder now.
The previous problem is that we'd have multiple schemas defining the same subschema (e.g. "Status") and AvroTurf would crash.