There was a good bit of cruft I'm pretty sure wasn't needed.
However, inlining and such may change, and this should be
investigated and dealt with. Also, for "expensive" functors,
it may pay to play tricks to fuse fmaps. I think this can be
done by using clonePrism (a . b) rather than just a . b in
various places.
There was a good bit of cruft I'm pretty sure wasn't needed. However, inlining and such may change, and this should be investigated and dealt with. Also, for "expensive" functors, it may pay to play tricks to fuse
fmaps. I think this can be done by usingclonePrism (a . b)rather than justa . bin various places.