Verifying serialisation streams when used for messaging
- by Nick
I want to add distributed messaging to some applications. One target would be embedded and the other applications would be running on Windows. I'm thinking of having a connection manager, to which all of the applications would connect and communicate.
I only need some simple kind of RPC mechanism. I've looked at Google Protocol Buffers, but I don't really like the idea of code generation. I'd rather use serialisation and somehow verify the format of the serialised stream. I.e. Perhaps send metadata about what the stream would contain so that each endpoint can verify it is reading the correct version. Does anyone know of any examples of this?
Perhaps the message format could be read from a file which all the applications read to verify the stream format?