Tony Morris

@andreypopp @substack except they're not monads, like semicolons are not monads, but they could be. I wonder what the benefits would be.

@puffnfresh @substack I would argue that semicolons are monads, well, not ; chars themselves but operational semantics of stmt execution

@puffnfresh @substack we could say that there's an isomorphism between semicolons and monads :-)

@andreypopp @substack statements can be embedded, it's not an isomorphism. You really don't get the benefits (i.e. abstracting over monads).

@puffnfresh @substack 3) being religious about algebraic structure? WTF?! there are a lot of them!

@andreypopp @substack I don't understand anything but point #3. People think monads are magic, but just another M[A] => M[B] is not magic.

@puffnfresh @substack prove that "operational semantics of statement execution" can't be modelled as a monad [...]

@andreypopp @substack you can model a language as a monad. Prove that you can abstract over statements without monads. Not at all the same.