How ready is Linear Haskell for real world use cases? about linear-base HOT 3 CLOSED

I'd say that linear GHC is somewhat ready. Linear types work fine, but let and where bindings are never linear, which can be rather painful. This is something that I've fixed in a merge request, which has been reviewed, just needs a little bit of clean up probably. It'll be merged soon, which means that GHC 9.10 will be in a much better place for linear types. One thing that I don't really have a path to fixing, though, is the rather unhelpful error messages you get from linearity mismatches.

Linear types can still be useful to you today, though. If they bring strong benefits to you, it may be worth the inconvenience.

Linear-base is going slowly largely because there isn't much that I both want to do and know how to do.

Is there any way to help for a layman like me?

I don't have anything in particular in mind. But if you decide to use the linear GHC and linear-base and give feedback, this would already be a ton of help. It may also be a path to deeper contributions 🙂 .

Are there any plans for integrating directly with rust's affine types too?

There aren't. I'm not even quite sure what it'd mean, to be honest. Having affine types in addition to linear types would not be much of a technical challenge (the way we've built the formalism and the implementation actually makes it rather easy). It's just that ownership à la Rust and affine types aren't really the same thing. So we'd have to understand what kind of integration we're looking at before we can start really discussing what it'd imply.

Is there a chat or mailing list I could join?

Not as of yet, but I've opened the discussions on Github, don't hesitate to have chats there.

from linear-base.

Let me close as I believe the question has been answered.

from linear-base.

As a follow up, you were asking for chats: we now have a Discord: https://discord.gg/7yg5GxzvDJ .

from linear-base.