Conversation

Haven’t thought about it much yet but possibly via dependency on (some portion of) sig or m? I would tend to prefer to have this be in the carrier’s purview rather than the effect’s, however, and that’s already trivial: add a constraint to the Algebra instance.
1
Replying to
as it happens, they can— & demonstrated that last year (actually ages ago but that’s when it was published). the trouble as I recall is that the effect uses the higher-order parameter `m` invariantly so it doesn’t place nicely with other effects.
1
2