Proof was done by Hai & me

This fixes issue #44.

No longer `put box_own_prop γ P` in the invariant, it is persistent.

    The idea on magic wand is to use it for curried lemmas and use ⊢ for uncurried lemmas.

ProofMode intro patterns: accept _ as part of variable names

@robbertkrebbers beat this. ;)

See merge request !29

Use notation N @⊆ E to avoid ambiguity.

Since `nclose : namespace → coPset` is declared as a coercion, the notation `nclose N ⊆ E` was pretty printed as `N ⊆ E`. However, `N ⊆ E` could not be typechecked because type checking goes from left to right, and as such would look for an instance `SubsetEq namespace`, which causes the right hand side to be ill-typed.

See merge request !24

Fractional typeclass.

A typeclass for fractional assertions, that is assertions that depend on a fraction and that can be split.

This is used to derive generically a few other instances for framing , destructing, combining and spliting assertions of sums of fractions. I found it usefull when doing fraction-heavy proofs in LambdaRust.

The Right Way  To Do It would be to use a typeclass over the *predicate* itself. Unfortunately, the unification algorithm of typeclasses is not powerful enough to do the right beta-expansion that would expose the predicate applied to some fraction. Instead, the `Fractional` type class has as parameters both the predicate and the applied form that can be directly unified with the fractured assertion. Not very pretty.

I wonder whether I should split this into two type classes: the first one would depend only on the predicate and would actually state the fractionality of it, and the second would do the beta-expansion job. What do you think?

See merge request !23