[Haskell-cafe] [ANN] ghc-typelits-knownnat-0.2: Solving GHCs KnownNat constraints

[Christiaan Baaij]

Yes, clash is the motivating use case. I wanted to simplify the API of the prelude for the upcoming 1.0 release. Especially to reduce monstrosities such as this: http://hackage.haskell.org/package/clash-prelude-0.10.11/docs/CLaSH-Sized-Fixed.html#t:ENumSFixedC

Cheers,

Christiaan

> On 17 Aug 2016, at 19:50, Will Yager <will.yager at gmail.com> wrote:
> 
> Ah. I believe I have run into such issues with Clash, requiring me to add extraneous constraints over the results of addition. I assume this is the motivating use case here?
> 
> Great work on this and Typelits-natnormalise!
> 
> Will
> 
>>> On Aug 17, 2016, at 06:11, Christiaan Baaij <christiaan.baaij at gmail.com> wrote:
>> 
>> If you have ever worked with `GHC.TypeLits`, then you have probably encountered an error very similar to:
>> 
>>>   • Could not deduce (KnownNat (n + 2))
>>>       arising from a use of ‘natVal’
>>>     from the context: KnownNat n
>>