FlatBuffers VS Protobuf

According to the official website:

> Protobuf is intentionally designed to NOT require any parsing at all.

As others have mentioned, this is simply not the case, and the VARINT encoding is a trivial counterexample.

It is this required decoding/parsing that (largely) distinguishes protobuf from Google's flatbuffers:

https://github.com/google/flatbuffers

https://flatbuffers.dev/

Cap'n Proto (developed by Kenton Varda, the former Google engineer who, while at Google, re-wrote/refactored Google's protobuf to later open source it as the library we all know today) is another example of zero-copy (de)serialization.

I don't work at Cloudflare but follow their work and occasionally work on performance sensitive projects.

If I had to guess, they looked at the landscape a bit like I do and regarded Cap'n Proto, flatbuffers, SBE, etc. as being in one category apart from other data formats like Avro, protobuf, and the like.

So once you're committed to record'ish shaped (rather than columnar like Parquet) data that has an upfront parse time of zero (nominally, there could be marshalling if you transmogrify the field values on read), the list gets pretty short.

https://capnproto.org/news/2014-06-17-capnproto-flatbuffers-... goes into some of the trade-offs here.

Cap'n Proto was originally made for https://sandstorm.io/. That work (which Kenton has presumably done at Cloudflare since he's been employed there) eventually turned into Cloudflare workers.

Another consideration: https://github.com/google/flatbuffers/issues/2#issuecomment-...

Much more in the discussion of this particular issue onGitHub: flatbuffers:Reverse engineering #4258

Related but not Rust-specific: FlatBuffers, Cap'n Proto.

If json files are a concern because of space, you can always look into something like protobuffers or flatbuffers. But whatever you use, you should try to find a solution where you don't have to think about the actual serialization/deserialization of your objects, and can just concentrate on the data.

Or, as an alternative to protobuffers, there's also flatbuffers, which is lighter weight and needs less memory: https://google.github.io/flatbuffers/

Note: If you want to use the Google Protobuf compiler for code generation, download it from GitHub based on your OS. Follow this link for code generation instructions.

Unary length prefix is a solid technique. I would be careful with the performance claims, though. Sometimes what the machine can do is surprising and the performance of a system that looks like it needs to loop and branch is faster than you expected. The protobuf project, unsurprisingly for a project of its age, has several different varint parsing strategies. One is described at https://github.com/protocolbuffers/protobuf/blob/main/src/go... and another at https://github.com/protocolbuffers/protobuf/blob/main/src/go...

See protobuf field presence:

https://github.com/protocolbuffers/protobuf/blob/main/docs/f...

This is a language-agnostic serialization library that is much faster than JSON and provides way more compact representation. There are several implementations with different features and trade offs. μpb is a very lightweight C++ implemantion that uses arena allocations.

I have raised an issue requesting the addition of optional scalar types, but it is not planned to be supported by protoc-gen-kotlin.

protobuf: go get -u google.golang.org/protobuf/proto

It's because they changed the versioning format: https://github.com/protocolbuffers/protobuf/releases?page=5

But I suppose old version still receive bugfixes.

For at least 4 years protobuf has had decent support for self-describing messages (very similar to avro) as well as reflection

https://github.com/protocolbuffers/protobuf/blob/main/src/go...

Xgooglers trying to make do on the cheap will just create a Union of all their messages and include the message def in a self-describing message pattern. Super-sensitive network I/O can elide the message def (empty buffer) and any for RecordIO clone well file compression takes care of the definition.

Definitely useful to be able to dig out old defs but protobuf maintainers have surprisingly added useful features so you don’t have to.

Bonus points tho for extracting the protobuf defs that e.g. Apple bakes into their binaries.