--- Begin Message ---
Subject: |
GnuPG 2.2.30 cannot do symmetric encryption |
Date: |
Tue, 14 Dec 2021 11:46:12 -0500 |
Our GnuPG package is version 2.2.30, which includes this bug:
https://dev.gnupg.org/T5577
The effect of this bug is that symmetric encryption / decryption does
not work. The bug was fixed in 2.2.31 and 2.3.3.
Changing GnuPG will cause 2406 rebuilds. I think that's suboptimal but
it's the situation.
There is a gnupg-2.2.32 package, but it's hidden because it would break
emacs-pinentry:
https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/gnupg.scm?id=861ee6d908fefc47d765f81e33cdf6f84b6d50eb#n350
Here is a patch that unhides gnupg-2.2.32, updates it to 2.2.33, and
makes emacs-pinentry use that package variant.
If emacs-pinentry cannot use a current GnuPG, what should we do? And
maybe we don't need emacs-pinentry anymore?:
https://emacs.stackexchange.com/a/64721
0001-gnu-GnuPG-Update-gnupg-2.2.32-to-2.2.33.patch
Description: Text document
signature.asc
Description: PGP signature
--- End Message ---
--- Begin Message ---
Subject: |
Re: bug#52483: GnuPG 2.2.30 cannot do symmetric encryption |
Date: |
Sun, 26 Dec 2021 17:46:41 -0500 |
On Sun, Dec 19, 2021 at 03:36:25PM -0500, Leo Famulari wrote:
> On Sun, Dec 19, 2021 at 03:25:42PM -0500, Maxim Cournoyer wrote:
> > > I tested with GnuPG 2.2.23 by building all packages that depend directly
> > > on GnuPG. There were no new failures on x86_64-linux when using GnuPG
> > > 2.2.32.
>
> I see that my message was confusing, with mixed up version numbers.
> There's no regressions on x86_64 with 2.2.32. Since it fixes this bug,
> I'd like to make it available soon. Maybe we can unhide the gnupg-2.2.32
> variable, and remove emacs-pinentry, since everyone is saying that it's
> no longer useful.
I pushed commit d03aa942d, which makes GnuPG 2.2.32 available in the UI.
I didn't remove emacs-pinentry... that's something that Emacs users
should decide how to handle.
--- End Message ---