You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
to pun naming, but fundeps have a less systematic theory/support in core ghc, and I'm told that compile (well type checking particularly) times can be dramatically improved by a moving to type families.
Should consider migrating to
classArraymarrelwheretypeIxmarr::Nat-- might define Ix type family out of band of the class -- so i can pun it between arrays and formatsread:: (rnk~Ixmarr) =>ShaperankInt->marr (PrimStatem) el->mel
need to see how more complex algorithms wind up looking wrt types, and balancing usability/complexity
The text was updated successfully, but these errors were encountered:
currently, i do stuff like
to pun naming, but fundeps have a less systematic theory/support in core ghc, and I'm told that compile (well type checking particularly) times can be dramatically improved by a moving to type families.
Should consider migrating to
need to see how more complex algorithms wind up looking wrt types, and balancing usability/complexity
The text was updated successfully, but these errors were encountered: