Would this be workable if we added a coercible constraint to it or does that still hit the same sortah issue? I’m a bit too scatterbrained today to experiment locally. On Thu, May 27, 2021 at 1:45 PM Oleg Grenrus wrote:

I don't think this as good idea, as that definition doesn't work with GND.

{-# LANGUAGE RankNTypes, GeneralizedNewtypeDeriving, ConstrainedClassMethods #-}

import Data.Bits import Data.Word

class Bits a => Bits' a where withFiniteBits :: Bits a => a -> (FiniteBits a => r) -> r -> r

instance Bits' Word8 where withFiniteBits _ x _ = x

newtype W = W Word8 deriving (Eq, Show, Bits, Bits', FiniteBits)

fails with

Bi.hs:12:47: error: • Couldn't match type ‘Word8’ with ‘W’ arising from the coercion of the method ‘withFiniteBits’ from type ‘forall r. Bits Word8 => Word8 -> (FiniteBits Word8 => r) -> r -> r’ to type ‘forall r. Bits W => W -> (FiniteBits W => r) -> r -> r’ • When deriving the instance for (Bits' W) | 12 | newtype W = W Word8 deriving (Eq, Show, Bits, Bits', FiniteBits)

Note, FiniteBits has nominal role, so cannot be coerced to FiniteBits W.

If CLC decides this is still fine, then I'd suggest to not have any migration period as adding method is breaking change for GND users, so rather break everyone at once, maybe even by moving bitSizeMaybe out of the class.

- Oleg On 27.5.2021 20.22, Edward Kmett wrote:

This does seem like a strict improvement over the status quo.

Users can then conditionally get access to count(Trailing|Leading)Zeros by refining the type information available to them using this combinator, which offers a bunch of bit twiddling usecases.

-Edward

On Tue, May 18, 2021 at 9:30 AM Zemyla wrote:

...

Every Bits instance has to know that whether it's a FiniteBits instance as well, due to bitSizeMaybe. Therefore, it should also be able to tell a program that it is in fact a FiniteBits instance. There should be a function added to Data.Bits.Bits:

withFiniteBits :: Bits a => a -> (FiniteBits a => r) -> r -> r

The default should be withFiniteBits _ _ x = x, at least for the next several versions. bitSizeMaybe can be defined as

bitSizeMaybe x = withFiniteBits x (Just (finiteBitSize x)) Nothing

once everyone is on board with properly defining the value. _______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

_______________________________________________ Libraries mailing listLibraries@haskell.orghttp://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

Yeah. Is there any other way we could make it work or Does it turn into the same issue of coercions in ghc don’t have the right fancy to support this improvement nicely? On Fri, May 28, 2021 at 2:45 AM Edward Kmett wrote:

Destroying GND for Bits _is_ a pretty huge downside. =/

On Thu, May 27, 2021 at 10:45 AM Oleg Grenrus wrote:

...

I don't think this as good idea, as that definition doesn't work with GND.

{-# LANGUAGE RankNTypes, GeneralizedNewtypeDeriving, ConstrainedClassMethods #-}

import Data.Bits import Data.Word

class Bits a => Bits' a where withFiniteBits :: Bits a => a -> (FiniteBits a => r) -> r -> r

instance Bits' Word8 where withFiniteBits _ x _ = x

newtype W = W Word8 deriving (Eq, Show, Bits, Bits', FiniteBits)

fails with

Bi.hs:12:47: error: • Couldn't match type ‘Word8’ with ‘W’ arising from the coercion of the method ‘withFiniteBits’ from type ‘forall r. Bits Word8 => Word8 -> (FiniteBits Word8 => r) -> r -> r’ to type ‘forall r. Bits W => W -> (FiniteBits W => r) -> r -> r’ • When deriving the instance for (Bits' W) | 12 | newtype W = W Word8 deriving (Eq, Show, Bits, Bits', FiniteBits)

Note, FiniteBits has nominal role, so cannot be coerced to FiniteBits W.

If CLC decides this is still fine, then I'd suggest to not have any migration period as adding method is breaking change for GND users, so rather break everyone at once, maybe even by moving bitSizeMaybe out of the class.

- Oleg On 27.5.2021 20.22, Edward Kmett wrote:

This does seem like a strict improvement over the status quo.

Users can then conditionally get access to count(Trailing|Leading)Zeros by refining the type information available to them using this combinator, which offers a bunch of bit twiddling usecases.

-Edward

On Tue, May 18, 2021 at 9:30 AM Zemyla wrote:

...

Every Bits instance has to know that whether it's a FiniteBits instance as well, due to bitSizeMaybe. Therefore, it should also be able to tell a program that it is in fact a FiniteBits instance. There should be a function added to Data.Bits.Bits:

withFiniteBits :: Bits a => a -> (FiniteBits a => r) -> r -> r

The default should be withFiniteBits _ _ x = x, at least for the next several versions. bitSizeMaybe can be defined as

bitSizeMaybe x = withFiniteBits x (Just (finiteBitSize x)) Nothing

once everyone is on board with properly defining the value. _______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

_______________________________________________ Libraries mailing listLibraries@haskell.orghttp://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries

_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries