These days I am of the opinion that we don’t need a real IDL. Instead one should choose a “primary language” and define the schema in that language. All other language bindings are generated from that canonical language. Then for wire format you’d just pick something workable from pb json msgpack etc. This replaces all the Google codegen intermediate layer, and for the primary language it naturally becomes very ergonomic.

People who want protobuf but better often settle for something like cap’n proto https://capnproto.org/

didn't the capnproto author end up creating protobuf2? or am I mixing things up.

That's me. Other way around -- proto2 came first.

Hence the “something like”. There are a few of these sorts of things. I’m not really in the market for them so I don’t try to keep up.

In my experience, during the phase where you scale up and move from interpreted to compiled code with microservices, someone inevitably benchmarks protobuf vs JSON, and protos come out ahead.

Also, using it as a schema doesn’t totally suck. There are others but proto is… fine.

I have lots of opinions, but don’t feel like spinning up to a full rant here :-)

Cargo-cult mentality.

If protobuf hasn’t been created by Google, nobody would use it.

It’s actually not well designed for small requests, and doesn’t even deal with versioning all that well. It’s more verbose than a C struct yet not self-describing either.

To me, it fails to excel at any one thing yet people blindly use it anyway.

If one really considers the details, Protobuf makes ASN.1 actually start to look good!