According to the conformance suite, there are a handful of cases for which JSON serialization should fail:
= Duration value in memory is outside of valid range for JSON = Timestamp value in memory is outside of valid range for JSON = Fieldmask value in memory is not representable in JSON This would seem to violate one of the stated goals of proto3 — to simplify implementations. In particular, requiring that JSON serialization be able to fail requires additional boilerplate for every single call to a JSON serialization method. This seems somewhat burdensome just to handle the remote possibility that someone might try to accidentally process a date thousands of years from now. (It’s especially burdensome for those users who aren’t actually using any of these types so have to write boilerplate error-handling logic that will never be called.) It would be far simpler to instead truncate Durations and Timestamps to the representable range and to redefine Fieldmask to drop unrepresentable fields. This would eliminate a lot of boilerplate from implementations (and users) of the JSON code. Cheers, Tim -- You received this message because you are subscribed to the Google Groups "Protocol Buffers" group. To unsubscribe from this group and stop receiving emails from it, send an email to protobuf+unsubscr...@googlegroups.com. To post to this group, send email to protobuf@googlegroups.com. Visit this group at https://groups.google.com/group/protobuf. For more options, visit https://groups.google.com/d/optout.
