[bug#44518] [PATCH] gnu: Add matterhorn.

["Romain GARBAGE" <romain.garbage@inria.fr>]

Hi Ricardo,

On Sun Jul 7, 2024 at 1:07 AM CEST, Ricardo Wurmus wrote:
> Hi Romain,
>
> > * gnu/packages/haskell-apps.scm (matterhorn): New variable.
> > * gnu/packages/haskell-crypto.scm (ghc-crypton, ghc-crypton-x509,
> > ghc-crypton-x509-store, ghc-crypton-x509-system,
> > ghc-crypton-x509-validation, ghc-tls-1.9): New variables.
> > * gnu/packages/haskell-web.scm (ghc-crypton-connection,
> > ghc-mattermost-api, ghc-mattermost-api-qc, ghc-modern-uri): New variables.
> > * gnu/packages/haskell-xyz.scm (ghc-aspell-pipe, ghc-bimap,
> > ghc-checkers, ghc-monad-parallel, ghc-text-zipper-13, ghc-stm-delay,
> > ghc-unique, ghc-unix-compat-7, ghc-vty-6, ghc-vty-crossplatform,
> > ghc-vty-unix, ghc-hclip, ghc-brick, ghc-brick-skylighting): New variables.
>
> thank you for the patch!  This looks all pretty good.

Thank you for your review. I will send a v2 fixing the comments I don't answer
below.

> There are a few minor quibbles:
>
[...]
> * ghc-tls-1.9, ghc-unix-compat-7, ghc-vty-6, ghc-text-zipper-13: Is it
>   desirable to have these packages instead of upgrading the existing
>   packages?  Is there pending work on the Haskell team branch(es) that
>   would make the use of these versioned variants obsolete?

I don't have myself an answer for the first question: I didn't try to upgrade
any of these packages locally as I didn't want to "break things in Guix".
I guess I could give it a try before sending the v2 if you think it is
better. WDYT?

As for the 2nd question, I checked the haskell-team branch and didn't
see any update of these packages (I might not have checked the right
thing though).

-- 
Romain