The required libgpg-error 1.38 has now been released.

There are likely some bugs in the new code and I also want to do some improvements; see rGb4f1159a5bd7. But things should basically work as before and thus I set this again to testing

i'd be unlikely to ship anything as /etc/gnupg/gpg.conf or /etc/gnupg/dirmngr.conf just because of the mess that admins have to deal with when shipped config files change.

Arggh, gpgconf uses its own option parser so adding the global config file there will require some extra work.

@dkg You might find this interesting. Debian could do stuff in /etc/gnupg/gpg.conf or /etc/gnupg/dirmngr.conf without patching GnuPG to change some defaults.

All done in master with the latest libgpg-error (see T4859). There is always a global configure file in /etc/gnupg (or whatever "gpgconf --list-dirs sysconfdir" prints). The name of the configure file is the same as the user config file (gpg.conf, gpgsm.conf, gpg-agent.conf, ...) but for gpg.conf no versioned config names are used.

Okay, we now have global conf files in master. The extra flags to ignore or force certain options will be added to libgpg-error.

This is fixed now.

This is fixed in Gpg4win-3.1.11

We now have a decent error message for this.

Maybe it is the same issue described in https://dev.gnupg.org/T4717 ?

I think we can fix it by removing the smime attachment from OOM, because we still have it in MAPI, we just never cared that it was also in OOM (where only our decrypted attachments belong) because it was hidden.

In T2300#90092, @aheinecke wrote:

Ah nevermind. I think myself that this is nobug and current behavior is correct.

To implement / test the "not literally RFC compliant but in practice better" behavior let us call this now a wish and feature request as there are certificates in the wild other then intevation's and customers in large institutions run into that.