The implementation of <> for Max a and Min a guarantees commutativity as long as Ord class laws are satisfied.

Hi there

I'm wondering whether it's possible to make a new release to Hackage? (After #10, I'm guessing everything should just work?)

I have a few packages that depend (transitively) on commutative-semigroups: I'd love to be able to provide support for GHC 9.6.1 and base version 4.18. (Obviously, I can tell people to write --allow-newer as a workaround, so this is not urgent!)

Best wishes!

Jonathan

Off the top of my head:

• patch (so long as it has a Group class) (reflex-frp/patch#39)
• groups (Taneb/groups#14)
• group-theory (if required, or maybe the transitive re-export is fine)
• monoid-subclasses (blamario/monoid-subclasses#35
• monoidal-containers

blamario/monoid-subclasses#35 proposes to add a couple of orphan instances:

• instance Ord a => Commutative (Set.Set a)
• instance Commutative IntSet.IntSet

ISTM that monoid-subclasses is a bit of an odd place for them, and that we could depend on and provide instances for boot packages like containers without making this package feel too "heavy"

Ping @Taneb , as groups currently depends only on base, and he may have opinions.

Hi there

Would it be possible to publish a new (or updated) version of commutative-semigroups on Hackage, to support the GHC 9.8 series (base major version 4.19)?

Many thanks!

Jonathan

Context: the latest release of commutative-semigroups on Hackage (0.1.0.1) has the constraint base < 4.19.