guix-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#37435: [bug#37434] [PATCH 1/2] gnu: ghc-megaparsec: Update to 7.0.5.


From: Eric Bavier
Subject: bug#37435: [bug#37434] [PATCH 1/2] gnu: ghc-megaparsec: Update to 7.0.5.
Date: Tue, 24 Sep 2019 22:20:28 -0400 (EDT)

----- On Sep 24, 2019, at 12:08 PM, Ludovic Courtès address@hidden wrote:

> Hi Eric,
> 
> Eric Bavier <address@hidden> skribis:
> 
>> From 832529b71c2c5c7e6aa6e88a206c0db65e8a0c2e Mon Sep 17 00:00:00 2001
>> From: Eric Bavier <address@hidden>
>> Date: Mon, 16 Sep 2019 22:20:23 -0500
>> Subject: [PATCH] gnu: idris: Update to 1.3.2.
>>
>> * gnu/packages/idris.scm (ghc-network-2.8, ghc-megaparsec-7): New variables.
>> (idris): Update to 1.3.2.
>> [source]: Remove patch.
>> [inputs]: Use ghc-network-2.8 and ghc-megaparsec-7.
>> * gnu/packages/patches/idris-test-no-node.patch: Delete.
>> * gnu/local.mk (dist_patch_DATA): Remove it.
> 
> [...]
>> +;;; Until we update our default GHC to >=8.6 we cannot update our 
>> ghc-network
>> +;;; package, since the 'cabal-install' package that supports the current
>> +;;; 'Cabal' module requires 'network==2.6.*'.  Providing a local ghc-network
>> +;;; works in this case since we use it only for a standalone program.
>> +(define ghc-network-2.8
>> +  (package
>> +    (inherit ghc-network)
> 
> You should only inherit from packages defined in the same module,
> otherwise we risk hitting the dreaded circular dependency mess (actually
> this particular case is fine because idris.scm is a “leaf”, but still,
> let’s not take this risk.)
> 
> So I recommend moving these two ghc-* packages next to their friends.

Makes sense.  Done.

> With this change the patch LGTM.

Pushed with that change and other small adjustments for latest master in commit 
4780db2c93.

-- 
`~Eric





reply via email to

[Prev in Thread] Current Thread [Next in Thread]