Notation abuse is just a parametric typeclass instance
March 6, 2026 at 11:33 AM
Notation abuse is just a parametric typeclass instance
don’t say that shit to a haskeller they HATE people thinking rust has a Real typeclass system
October 9, 2025 at 4:05 PM
don’t say that shit to a haskeller they HATE people thinking rust has a Real typeclass system
the `Ord` typeclass in Haskell represents types that have been ordained
October 16, 2025 at 11:36 PM
the `Ord` typeclass in Haskell represents types that have been ordained
instancier la typeclass des foncteurs)
February 21, 2025 at 9:56 AM
instancier la typeclass des foncteurs)
haskell should really have a second ord typeclass
April 26, 2025 at 12:30 AM
haskell should really have a second ord typeclass
The "Representable" typeclass in Haskell should have been called "NotIndexOutOfBoundsWorryable" instead.
hackage.haskell.org/package/adju...
hackage.haskell.org/package/adju...
Data.Functor.Rep
hackage.haskell.org
January 2, 2025 at 7:59 PM
The "Representable" typeclass in Haskell should have been called "NotIndexOutOfBoundsWorryable" instead.
hackage.haskell.org/package/adju...
hackage.haskell.org/package/adju...
is this why Rust stole Haskell’s typeclass system? who can say!
October 9, 2025 at 4:03 PM
is this why Rust stole Haskell’s typeclass system? who can say!
i’m explaining the Scala programming language
November 17, 2024 at 4:18 PM
i’m explaining the Scala programming language
ℹ️ GHC 9.11 just released, shipping a new language feature for expressing negative typeclass instance assertions. This often pleases the typechecker when OverlappingInstances is in use. For instance:
{-# LANGUAGE DecidableNoninstances #-}
newtype Pred a = Pred (a → Bool) deriding Functor
{-# LANGUAGE DecidableNoninstances #-}
newtype Pred a = Pred (a → Bool) deriding Functor
July 13, 2025 at 5:54 PM
ℹ️ GHC 9.11 just released, shipping a new language feature for expressing negative typeclass instance assertions. This often pleases the typechecker when OverlappingInstances is in use. For instance:
{-# LANGUAGE DecidableNoninstances #-}
newtype Pred a = Pred (a → Bool) deriding Functor
{-# LANGUAGE DecidableNoninstances #-}
newtype Pred a = Pred (a → Bool) deriding Functor
Reading the typeclass for functors gave me the clearest intuition for functors, tempted to change my category theory reading club to a category theory with haskell reading club
August 20, 2026 at 3:13 AM
Reading the typeclass for functors gave me the clearest intuition for functors, tempted to change my category theory reading club to a category theory with haskell reading club
what even is a typeclass bro
August 20, 2025 at 8:39 PM
what even is a typeclass bro
An introduction to typeclass metaprogramming. ~ Alexis King. lexi-lambda.github.io/blog/2021/03... #Haskell #FunctionalProgramming
April 7, 2025 at 6:38 AM
An introduction to typeclass metaprogramming. ~ Alexis King. lexi-lambda.github.io/blog/2021/03... #Haskell #FunctionalProgramming
probably Haskell
In Haskell when we say that some type constructor "T" "is a Monad" that's actually a shorthand for saying that T "implements Monad" or (even longer) that "T implements the Monad typeclass"
Any type constructor "T" that implements the Monad typeclass "is a Monad"
In Haskell when we say that some type constructor "T" "is a Monad" that's actually a shorthand for saying that T "implements Monad" or (even longer) that "T implements the Monad typeclass"
Any type constructor "T" that implements the Monad typeclass "is a Monad"
November 10, 2025 at 3:49 PM
probably Haskell
In Haskell when we say that some type constructor "T" "is a Monad" that's actually a shorthand for saying that T "implements Monad" or (even longer) that "T implements the Monad typeclass"
Any type constructor "T" that implements the Monad typeclass "is a Monad"
In Haskell when we say that some type constructor "T" "is a Monad" that's actually a shorthand for saying that T "implements Monad" or (even longer) that "T implements the Monad typeclass"
Any type constructor "T" that implements the Monad typeclass "is a Monad"
i just now created a c# project (cause that's the one language i both do not loathe and have enough experience in to feel confident writing production code) and then remembered i wouldn't be able to make cborable a typeclass and deleted the folder
May 20, 2025 at 11:56 PM
i just now created a c# project (cause that's the one language i both do not loathe and have enough experience in to feel confident writing production code) and then remembered i wouldn't be able to make cborable a typeclass and deleted the folder
Have you ever wondered how to build automatic derivation? If you want to know more about this mechanism in Scala 3, I have written something about it.
gilwath.com/articles/sca...
#scala #scala3 #typeclass
gilwath.com/articles/sca...
#scala #scala3 #typeclass
Automatic Typeclass Derivation in Scala 3
Building a Debuggable typeclass from scratch and teaching the Scala 3 compiler to derive it for any case class or enum, using Mirror and inline metaprogramming.
gilwath.com
June 4, 2026 at 7:39 AM
Have you ever wondered how to build automatic derivation? If you want to know more about this mechanism in Scala 3, I have written something about it.
gilwath.com/articles/sca...
#scala #scala3 #typeclass
gilwath.com/articles/sca...
#scala #scala3 #typeclass
This seems like the silliest way to encode a partial order
November 24, 2025 at 3:22 PM
This seems like the silliest way to encode a partial order
GHC already does non deterministic typeclass resolution lmao
September 2, 2026 at 12:43 AM
GHC already does non deterministic typeclass resolution lmao
Me writing PureScript: Ah yes, every case must be covered, I need to create a typeclass here to push the logic into the types, fuck if I'm ever going to understand this code in the future but at least it's type safe!
Me writing Typescript: what?.the?.fuck?.fuckit ?? "whatever who gives a shit"
Me writing Typescript: what?.the?.fuck?.fuckit ?? "whatever who gives a shit"
March 23, 2025 at 12:25 PM
Me writing PureScript: Ah yes, every case must be covered, I need to create a typeclass here to push the logic into the types, fuck if I'm ever going to understand this code in the future but at least it's type safe!
Me writing Typescript: what?.the?.fuck?.fuckit ?? "whatever who gives a shit"
Me writing Typescript: what?.the?.fuck?.fuckit ?? "whatever who gives a shit"
you should also check out 'unliftio' hackage.haskell.org/package/unli...
unliftio
The MonadUnliftIO typeclass for unlifting monads to IO (batteries included)
hackage.haskell.org
March 14, 2026 at 8:55 AM
you should also check out 'unliftio' hackage.haskell.org/package/unli...
so you can say “here’s what it means to be a sorting algorithm” and mark it as typeclass, and write code against it. and then separately you can have different instances of that typeclass that actually provide the code, and control which one it’s resolved to
August 18, 2025 at 3:22 PM
so you can say “here’s what it means to be a sorting algorithm” and mark it as typeclass, and write code against it. and then separately you can have different instances of that typeclass that actually provide the code, and control which one it’s resolved to
there’s an interesting application of symbolic logic solvers to typeclass-style systems in PLs, and obviously it crops up a lot in proof assistants
January 11, 2026 at 9:25 PM
there’s an interesting application of symbolic logic solvers to typeclass-style systems in PLs, and obviously it crops up a lot in proof assistants
ah of course. 10 is defined relative to the radix typeclass/implicit variable
August 21, 2025 at 7:12 AM
ah of course. 10 is defined relative to the radix typeclass/implicit variable
A very educational post on current state of Scala typeclass derivation approaches, why it's bad, and how we can do better.
kubuszok.com//2025/sanely...
kubuszok.com//2025/sanely...
kubuszok.com
November 4, 2025 at 9:01 PM
A very educational post on current state of Scala typeclass derivation approaches, why it's bad, and how we can do better.
kubuszok.com//2025/sanely...
kubuszok.com//2025/sanely...
how would you document the `Monad` typeclass overall? How much should go their, vs being assumed FP knowledge?
January 3, 2024 at 1:55 AM
how would you document the `Monad` typeclass overall? How much should go their, vs being assumed FP knowledge?