Open
Description
From @mikesol in Discord:
Would it make sense to revamp
Semiring
so that it had only the operatorsadd
andmul
and then useDioid
forzero
andone
? There is some precedent for that: https://en.m.wikipedia.org/wiki/Semiring. It would create a nice symmetry between semigroup/monoid/group and semiring/dioid/ring. It'd be a breaking change for folks usingzero
andone
, but otherwise I think it wouldn't be too disruptive. Also, it would allow for the generalization of semiring to an analogous biproduct class, which now isn't possible becausezero
andone
necessarily leave out all but one of the types in the biproduct or biCCC class. Thoughts?
Metadata
Metadata
Assignees
Labels
No labels