fix bug in hashing samples

[ericphanson]

The test I added fails on main:

julia> @test hash(samples) == hash(samples2)
Test Failed at REPL[23]:1
  Expression: hash(samples) == hash(samples2)
   Evaluated: 0x9bab1974999310ac == 0x877a3e1403901e84

but passes on this PR.

closes #150

[palday]

I think this makes sense, but I think @ararslan should weigh in on whether this will mess with anything in Julia internals relying on Base.hash

[kimlaberinto]

Looks good! And seems to make sense as it follows what is recommended in Base.hash's docstring to implement the 2-argument form and calling it recursively to to mix the hashes of the contents with each other.

(docstring from Julia 1.9.4)

  hash(x[, h::UInt]) -> UInt

  Compute an integer hash code such that isequal(x,y) implies hash(x)==hash(y). The optional second argument
  h is another hash code to be mixed with the result.

  New types should implement the 2-argument form, typically by calling the 2-argument hash method recursively
  in order to mix hashes of the contents with each other (and with h). Typically, any type that implements
  hash should also implement its own == (hence isequal) to guarantee the property mentioned above. Types
  supporting subtraction (operator -) should also implement widen, which is required to hash values inside
  heterogeneous arrays.

[ericphanson]

The CI failure looks unrelated but concerning

[palday]

I can replicate the CI failure on a local run and it also happens on main. Did an update in Arrow break something?

[palday]

Tests pass on main with Arrow set to 2.6 -- current Arrow release is 2.7.

[palday]

see #153.

This is definitely something that needs to be fixed in upstream Arrow, but we have a work around that's trivial

[ericphanson]

This is definitely something that needs to be fixed in upstream Arrow, but we have a work around that's trivial

Just to say, I don't think there was any bug in the Arrow release, but rather in Onda we were depending on Legolas generating a specific fromarrow method that we stopped generating after beacon-biosignals/Legolas.jl#106, since Samples has some custom serialization as it is not a Legolas @version. Resolved by #153 in any case.