Why stopped?

[Cvikli]

Hey,
I just checked this repo and it is pretty awesome.
Why did the development stop?

Why doesn't Google's team maximize its speed? The read/serialize speed is really close to JLD2 already.

The lib is like a left pearl.

[quinnj]

There is an updated implementation here: https://github.com/JuliaData/Arrow.jl/tree/main/src/FlatBuffers. The missing piece is integration with the official flatc compiler to produce Julia flatbuffer code. Currently, that is a manual process where I generated the flatbuffer code in go language and then translated to Julia as seen here: https://github.com/JuliaData/Arrow.jl/blob/main/src/metadata/Schema.jl. I have plans to eventually port the julia code back here and hopefully merge upstream, but just need to find the time.