Hacker News new | ask | show | jobs
by taeric 261 days ago
If algebraic effects is not about how to manipulate effects using algebra, then I will flat out say that it is poorly named. Which, fair that that is already established and I don't expect some rando on the internet's objection to matter on that.

The paper you linked confuses me on this stance, as it explicitly goes through the effort of defining operations as distinct from the values they produce. And it then shows how they can be manipulated using operations. To quote: "Think of a value as an inert datum that needs no further computation, such as a boolean constant, a numeral, or a λ-abstraction. An operation takes a parameter p, for instance the memory location to be read, or the string to be printed, and a continuation κ, which is a suspended computation expecting the result of the operation, for instance the contents of the memory location that has been read" I read that as explicitly setting up how to use algebraic constructs on operations. You are taking the stance that that is not what they are doing?

That all said, I fully sympathize that you are likely hitting a blind spot of mine. I'm already fine reading up more on this paper later, so I'm not wanting to burden you on trying to cover it again.