Matt Renaud

Ohhh, sorry, I misunderstood. I'll rename (and rework the benchmarks) as soon as I get a chance. I was thinking: - `alterF_IdentityNoRewrite` (using TestIdentity) - `alterF_IdentityRewrite` (using Identity) - `alterF_ConstNoRewrite`...

Sorry for the delay, I only had a few minutes here and there over the weekend to look at this stuff. I've re-organized the benchmarks as discussed and fixed the...

I originally had a todo to see if `alterF` should be marked `INLINABLE [1]` or `NOINLINE [1]` but it somehow got clobbered (I went with `NOINLINE` to be safe). The...

I ran it several more times and changed the order of benchmarks but the results remain about the same.

Woah woah woah, I ran it a few more times for good measure and got the below results which indicate that the rewrite is now slower than the non-rewritten version...

Done. I've left a few of the tests and benchmarks in case we want to revisit this in the future.

Updated PR description with new benchmark results and note on what GHC version used. Is rewording an already pushed commit allowed? I thought it may mess up history or something.

Sooo, these benchmarks tell a different story than the other ones. TL;DR; rewrite rules perform better in casses when the element being altered is present, but worse when the element...

Hey @DrNico, did you get around to sending a proposal to the libraries list? I noticed in PR #291 that the current suggested name is `lookupEntry` which SGTM. Also, could...

Could you also add a similar function to `Data.IntSet` and `Data.IntMap`?