[rabbitmq-discuss] Encoding message payload for multiple languages

Garrett Smith g at rre.tt
Thu Dec 31 19:17:33 GMT 2009


I'm curious what others here use for encoding message payload for
cross-language consumption.

I've used Thrift for quite some time, and it's quite nice I think, but
there's the annoying overhead of IDL maintenance, compiling and
distributing application specific client code, etc.

JSON is an obvious choice, but incurs a pretty high overhead for
encoding/decoding.

XML is similar to JSON in terms of pluses and minuses.

There's Protobuf, which seems like a possibly simpler alternative to
Thrift, but still requires IDL maintenance, etc.

Binary representation of Erlang terms is darned near perfect (fast and
flexible), but it's not as easy to consume from other language
environments.

CSV might be a good approach for some applications.

Obviously there's no "one way" here, but I'm curious what others are doing.

Garrett

P.S. Happy New Year!




More information about the rabbitmq-discuss mailing list