4 ms·

This is nice, and yet, not much different from what the Haskell community was dealing with 5 years ago. They are also mostly programming concerns, as opposed to

This is nice, and yet, not much different from what the Haskell community was dealing with 5 years ago. They are also mostly programming concerns, as opposed to engineering concerns. How do you deal with simple things like exceptions and string interpolation? There are guides trying to explain monads, but no straightforward answer for dealing with these sorts of things, with building separate libraries / packages, setting up your own CI with your own stackage repo. And we haven’t gotten into production yet. The ecosystem doesn’t have these things built out, or, it’s just not documented/agreed upon so you’d effectively be starting from scratch. While it might be nice to use Haskell for math-y things, in the end, you’d have to be able to instrument it, monitor/log it, etc. That doesn’t appear to be in place either. When I worked with Haskell last, these were all blockers.

johnday 6y ago
As a PhD student, I don't have enough experience working with Haskell in an industrial context to comment on the engineering concerns. I believe you when you say that these things are all problems. However, I'm aware that at least some solutions are underway. For example, the shake library [1] is a nice way to set up competent, reproducible build systems. [1] https://hackage.haskell.org/package/shake https://hackage.haskell.org/package/shake
kccqzy 6y ago
> How do you deal with simple things like exceptions and string interpolation? While I agree string interpolation isn't convenient, exceptions in Haskell are great. You can throw exceptions in pure code (throw upon evaluation), in IO (throw when sequenced), catch them, inspect them by doing type down casting, etc, all without pulling a dependency. It's even better than Rust. (In Rust I had to use the anyhow crate to get back some basic features I thought should be present by default.)
st1ck 6y ago
Exceptions work rather similar to other languages, except that you can only catch them in IO code. So generally (IMO) it's better to reserve them for actually exceptional cases, and not throw them around like in Python. String interpolation can be done with external libs using quasiquotations. It's not very pretty this way, so typically string concatenation or printf is used. Of all the things, math-y is probably not the be best usecase for Haskell, when you can use Python, R, SageMath maybe, or Julia. Haskell is best for writing parsers, compilers, interpreters. Or if you need a rich static type system. If you throw in LiquidHaskell, you have pretty good tooling for writing mostly correct programs.

johnday 6y ago

As a PhD student, I don't have enough experience working with Haskell in an industrial context to comment on the engineering concerns. I believe you when you say that these things are all problems. However, I'm aware that at least some solutions are underway. For example, the shake library [1] is a nice way to set up competent, reproducible build systems. [1] https://hackage.haskell.org/package/shake https://hackage.haskell.org/package/shake

kccqzy 6y ago

> How do you deal with simple things like exceptions and string interpolation? While I agree string interpolation isn't convenient, exceptions in Haskell are great. You can throw exceptions in pure code (throw upon evaluation), in IO (throw when sequenced), catch them, inspect them by doing type down casting, etc, all without pulling a dependency. It's even better than Rust. (In Rust I had to use the anyhow crate to get back some basic features I thought should be present by default.)

st1ck 6y ago

Exceptions work rather similar to other languages, except that you can only catch them in IO code. So generally (IMO) it's better to reserve them for actually exceptional cases, and not throw them around like in Python. String interpolation can be done with external libs using quasiquotations. It's not very pretty this way, so typically string concatenation or printf is used. Of all the things, math-y is probably not the be best usecase for Haskell, when you can use Python, R, SageMath maybe, or Julia. Haskell is best for writing parsers, compilers, interpreters. Or if you need a rich static type system. If you throw in LiquidHaskell, you have pretty good tooling for writing mostly correct programs.