Yes. Not uncommon for keystone papers to be written by people whose pedigree is essentially "conference attendee." They're just typically fairly unknown (because only professionals have time to build a profile)

Necessary, or subset is not a well-typed relation.

Either that, internalization, or some absurd but technically correct characterization like "the domain of an injective function into a type is a subtype"

Philosophical buy-in I think. It's not clear to most that they *shouldn't* ship their own layer.

I've been putting off looking into this, but is there an essential difficulty in typing then in a Hoare / separation logic?

Far more difficult seems making them remotely ergonomic.

It really could have gone to characterization of inductive types

Is it? It's still my go-to reference for at least the linguistic aspects.

Probably the part where you don't have to bring your own monad batteries.

Type theory is infamous for this, where they'll say something to the effect of "we anticipate no essential difficulty in showing this" precisely because the result would be a 100-page paper devoid of novelty.

it's footgunning yourself, but mostly for psycholinguistic reasons. you can certainly learn to decipher text before you learn anything about speaking.

if there's a je ne sais quoi, I'm still looking for it 17 years in

the record is 302*

*achieved offline by a rather sophisticated heuristic search

If you get the trans girl runaround, if the world decides to run you ragged as one of us and you don't bail, honestly I feel like you're closer to me than if you're some sort of quisling ass brianna wu loser. solidarity does sort of make us one.

the true tragedy of the doll starts when you leave that house. when all the people pleasing bullshit you were taught as a survival skill starts hurting people, people you like, people you care about. people who care about you!!

who wouldn't dream of broken clockwork.