Hi Jan,
mkAlgTyCon is good for making an algebraic datatype -- like Maybe or Bool or
Either. It's not what you want for RNil, which is really a data constructor.
Instead, rowKindCon should be the algebraic datatype, and RNil should be one of
its constructors. Then you get the RNil TyCon via
I would suggest that it's easier to define a normal type constructor
and data cons and then promote them.
This is how the runtime rep polymorphism stuff works for instance so
you can look there for inspiration (it's where I look to work out how
to implement multiplicity polymorphism).
Matt
On
I did, see here:
https://gitlab.haskell.org/jvanbruegge/ghc/blob/73b383275f3d497338ca50a3a7934445c3858450/compiler/prelude/TysWiredIn.hs#L235
Am 03.04.19 um 16:28 schrieb Simon Peyton Jones:
>
> Make sure you add the new wired in this to `TysWiredIn.wiredInTyCons`
>
>
>
> *From:*ghc-devs *On
Make sure you add the new wired in this to `TysWiredIn.wiredInTyCons`
From: ghc-devs On Behalf Of Jan van Brügge
Sent: 03 April 2019 12:06
To: ghc-devs@haskell.org
Subject: Defining a wired-in type of a different kind
Hi,
when trying to get familiar with the GHC code base for my Bachelor's
Yeah, sorry, it is here:
https://gitlab.haskell.org/jvanbruegge/ghc/commit/73b383275f3d497338ca50a3a7934445c3858450
Am 03.04.19 um 13:11 schrieb Ben Gamari:
> On April 3, 2019 7:06:11 AM EDT, "Jan van Brügge" wrote:
>> Hi,
>>
>> when trying to get familiar with the GHC code base for my
On April 3, 2019 7:06:11 AM EDT, "Jan van Brügge" wrote:
>Hi,
>
>when trying to get familiar with the GHC code base for my Bachelor's
>thesis. I followed the GHC Wiki, especially the case study about the
>bool type.
>Now I wanted to add a new kind and a new type inhabiting this kind
>(without
Hi,
when trying to get familiar with the GHC code base for my Bachelor's
thesis. I followed the GHC Wiki, especially the case study about the
bool type.
Now I wanted to add a new kind and a new type inhabiting this kind
(without having to expose a data constructor, so without datatype
promotion).