Closed gallais closed 3 years ago
In a different draft branch I re-did it, but I kept the ~~
syntax uniform, and changed the justification syntax ...
.
Any thoughts about this?
If you think this might be better, I'll wait until you finish with the certification PR and systematically change the syntax in both PRs.
Looks like the refactored refactoring introduces stuff the simplifier doesn't simplify any more?
The expected
file looks extremely dodgy
Oh, right. I forgot to copy the expected tex
Looks OK now, doesn't it?
I suppose it is in the nature of commutativity that you can either apply it forward or backward and it doesn't matter. I'm guessing you flipped a direction somewhere during the refactoring hence the changes?
Ah, probably. Esp. if it used an _
to work out the terms it should commute.
Generated proofs will be hard enough to read without adding an extra layer of cruft by writing e.g.
<~ x ...( (cast model).equivalence.symmetry ? ? $ prf )
when we could write~: x ...( prf )
with aSetoid
-specific syntax for reasoning.May as well refactor the library to use this notation too.