Eric Hanson

I think this might be https://github.com/JuliaData/Arrow.jl/issues/211 (but see also https://github.com/JuliaData/Arrow.jl/issues/90#issuecomment-914870675 for plans to soon switch the metadata system)

> writing custom metadata that supports reading a vector of data into an array structure? LegolasFlux does this: https://github.com/beacon-biosignals/LegolasFlux.jl/blob/aaa94fdb050e0b5333725d7d968b58b080b67df2/src/LegolasFlux.jl#L16 I’m not really sure the practical difference between something like that...

I ran into this with LegolasFlux; there I worked around it with a manual "FlatArray" type: https://github.com/beacon-biosignals/LegolasFlux.jl/blob/80569ab63a8248a8a063c76e0bbf701f4ada9bd4/src/LegolasFlux.jl#L15-L36 This is also coming up in Lighthouse.jl: https://github.com/beacon-biosignals/Lighthouse.jl/pull/45/files#diff-790bf7705c3c1ba479ac3f9931bc82ec801c90f412d218ab4f9f88020fc128beR1-R16

I've made a PR in https://github.com/tpgillam/arrow-julia/pull/1 to try to make this change backwards compatible.

Could be something to work on at the JuliaCon workshop [Package development: improving engineering quality & latency ](https://pretalx.com/juliacon2021/talk/review/HFFUMRJE8RGQPWZS9QRTVE9NEWCUMH3N)!

I would try to see if you get the error still with `DataFrame(Arrow.Table(input); copycols=true)`

Ooh, good point, maybe https://github.com/JuliaData/Arrow.jl/issues/237 is related. Maybe we just shouldn’t have an `@async` there? Since `ntasks=1`.

Well, it’s still spawning more than one task which seems wrong for `ntasks=1`, though the docstring does specify “threaded tasks” so I guess it’s accurate. Just feels like the `ntasks`...

It looks to me like this was fixed (or at least worked-around) in https://github.com/JuliaPlots/PlotlyJS.jl/commit/c4418ccc2537d1954913af351da9fbd71a55175b, there just hasn't been a release since then

I realized that we don't care which registry we get a repo from if two registries have the *same* repo; no need to prompt or error in that case (which...