GNU bug report logs - #49281
Add dynaconf

Previous Next

Package: guix-patches;

Reported by: paul <goodoldpaul <at> autistici.org>

Date: Tue, 29 Jun 2021 22:39:02 UTC

Severity: normal

Tags: patch

Merged with 49829, 49830, 49831, 49832, 49833, 49834, 49835

Done: Efraim Flashner <efraim <at> flashner.co.il>

Bug is archived. No further changes may be made.

Full log


Message #32 received at 49281 <at> debbugs.gnu.org (full text, mbox):

From: paul <goodoldpaul <at> autistici.org>
To: Sarah Morgensen <iskarian <at> mgsn.dev>
Cc: 49281 <at> debbugs.gnu.org
Subject: Re: bug#49281: Add dynaconf
Date: Mon, 2 Aug 2021 20:13:20 +0200
Dear Sarah,

thank you for your suggestions :D , I believe I addressed most of them.

On 7/23/21 8:14 AM, Sarah Morgensen wrote:
> * Unvendoring or otherwise removing files from sources is typically done
>    with a snippet in the origin rather than a patch, as it's much smaller
>    and doesn't break when updating. It might look like (untested):
>
> (origin
>    ...
>    (modules '((guix build utils)))
>    (snippet
>      '(begin
>        ;; Remove vendored dependencies
>        (let ((unvendor '("click" "dotenv" "ruamel" "toml")))
>          (with-directory-excursion "dynaconf/vendor"
>            (for-each delete-file-recursively unvendor))
>          (with-directory-excursion "dynaconf/vendor_src"
>            (for-each delete-file-recursively unvendor))))))
>
>    You'll still have to have the edits to dynaconf as a patch, of course.
It make much more sense, now the patch just changes the imports and the 
actual removal is up to the snippet.
> * You've still included a python-box package despite none of the
>    packages in your patch using it.
Yes I included it while unvendoring, I figured since the tests pass it 
would still make sense to upstream it. Should I remove it?
> * pep8-naming has released 12.0.0, and tests pass :)
Fixed, thanks !
> * Some of your patches no longer apply on master, and you should rebase
>    them before sending a revised patchset. Consider using the `--base`
>    option with format-patch, which helps git know what commit the patch
>    is based on when applying.

I rebased and I'll send the patches with 
--base=f12a35cfa22092a7e3157c94abfef8335f86ac1c .

Thank you for your help!


Cheers,

giacomo






This bug report was last modified 3 years and 171 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.