News for hackers. Stuff that matters.
Re:Recursive Program Synthesis using Paramorphisms [pdf] by bubblyworld on Saturday July 13, 2024 @03:14AM (#40952389)
Mmm, yeah on reading the paper more closely it seems like they don't make essential use of the concept really. It's more of a descriptor of the mechanics of their combinator language.

Personally I'm fine with that - I have no issue with CS people borrowing terminology from category theory to describe stuff, so long as the analogy is reasonable. But I get where you're coming from - having learned some CT in (grad) school, I used to find Haskell wikis confusing as heck because they use CT terminology but nobody clarifies what category they're actually working in!

Re:Recursive Program Synthesis using Paramorphisms [pdf] by tome on Saturday July 13, 2024 @03:31AM (#40952460)
> I used to find Haskell wikis confusing as heck because they use CT terminology but nobody clarifies what category they're actually working in!

"Hask", the (almost) category where the objects are types and the morphisms from A to B is the type A -> B.

Re:Recursive Program Synthesis using Paramorphisms [pdf] by bubblyworld on Saturday July 13, 2024 @04:14AM (#40952609)
Thanks, I know that now, but it's the "almost" part that trips up people like me! And it's not helped by the fact that the same sources will say something like "a monad is a monoid in the category of endofunctors", yet if you take Hask seriously they should also be objects in _that_ category, but there's no discussion of an reflective functor or anything like that.