Conal Elliott

Thanks for that link. Since we're not using GHC's run-time system (including its heap layout strategies), I think it's only the issue of _unlifted_ types that's a problem here (so...

Simon PJ mentions this hierarchy in the discussion ["Concrete syntax for open type kind?"](http://haskell.1045720.n5.nabble.com/Concrete-syntax-for-open-type-kind-td5747748.html), saying "... they are a gross hack, with many messy consequences. Particularly the necessary sub-kinding, and...

> Both are in Core, for example unlifted tuples vs unboxed Int#s. And, IIUC, unboxed things are necessarily unlifted. The difficulty I'm having with `Int#` is that it's unlifted, not...

Addressed via `reifyMethod`, as part of commit https://github.com/conal/lambda-ccc/commit/8ea54bd22a192ef3e00138f71e0c2029aa22b913: given `reifyEP (m d)`, if `m` is a variable and d is a dictionary, then `anytd inline >>> simplify`.

Now I see that this strategy doesn't work so well when the dictionary has already been reified, which happens when it's defined in a module that's being reified (either the...

Oops. Make `unlessDict` check that the _result_ is a dictionary. Otherwise, we inline dictionary-constructing functions, which come from instances with constraints, i.e., compound dictionaries.

> Shouldn't it be u-prefix instead of u-suffix? If you're asking whether "u-prefix" would be a more accurate description than "u-prefix", I'd say not to my ears, since I'm really...

How to specify what type of primitives?

Suggestive of "vector space", since integral needs to scale and add. (Really, one needs some more structure to define integration.)

I don't know how integration over time can be meaningful without time deltas being the scalar field associated with the vector spaces being integrated, as I think you're saying here....