[arjunnarayan]

protobufs still get encoded and decoded by each client when loaded into memory. arrow is a little bit more like "flatbuffers, but designed for common data-intensive columnar access patterns"

[hackcasual]

Arrow does actually use flatbuffers for metadata storage.

[poorman]

https://github.com/apache/arrow/tree/master/format

[hawk_]

is that at the cost of the ability to do schema evolution?