Data.Effect.Concurrent.Parallel might not be the best naming choice

[jeukshi]

I'll preface this by saying that I was really, really on the fence on opening this issue, because, yea, arguing about naming things on the internet is kinda rude. Please accept my apologies, I mean no harm, and I certainly won't insist on anything!

I'm going by the distinction from the legendary book "Parallel and Concurrent Programming in Haskell", where parallel computations are meant to be deterministic. You might want to reserve that name for those if you ever elect to implement such abstraction. Especially, because, as it is right now, "Parallel" after "Concurrent" is somewhat confusing. Why not simply one or the other?

Also, I'm interested in adding parallel effect to Bluefin (interesting exercise for an effect system), and since no Haskell effect system that I know about has this, it is not a bad time to keep the names consistent!

[ymdfield]

Thank you for opening the issue! I wasn't aware of the meaning of parallel in that book. I think following the Haskell convention for Parallel and Concurrent is a good choice. I will think it over, but for now it may be best to rename the Parallel effect to something like Concurrent.

BTW the reason I named it the Parallel effect is that the underlying abstraction is the Parallel type class in Scala:
https://www.javadoc.io/doc/org.typelevel/cats-docs_2.13/latest/cats/Parallel.html
https://medium.com/@PerrottaFrancisco/learning-cats-effects-parallel-execution-f617f883e390
The Haskell port of this abstraction is named ParDual:
https://hackage.haskell.org/package/par-dual-0.1.0.0/docs/Control-ParDual-Class.html
In any case these are non deterministic, so according to our discussion they should be called concurrent rather than parallel.

The tricky part is that, apart from ordinary concurrency/parallelism, the Validation Applicative also falls under this abstraction. In that sense, this abstraction (which supplies a one‑to‑one correspondence between a monad and an applicative) is not covered by existing concepts and is therefore something somewhat new.

I'd like to use a term that's as close to the opposite of "Sequential" as possible. It seems that not only "Parallel" but also "Concurrent" can be considered antonyms of "Sequential," so I will probably rename it to Concurrent.