Thanks for reporting. We usually test by moving the <keygrip>.key files around ;-)

For this I think a better solution might be if we would remove keys that are already in a group from the list of "available" keys.

By this I mean that if you edit a group and move keys down then they should be removed from the upper list of available keys so that it is easily visible which keys in your keyring are not part of a specific group.

done

works

very nice

This is now ready for testing. The import result dialog and the import error dialog now have an additional "Show Audit Log" button.

Actually we should switch from putenv to SetEnvironmentVariable et al. because that avoids problems wit different Windows libc versions, for example in DLLs.

Fixed in rG8e8971403f75: w32: Fix gnupg_unsetenv..

Sure, but this will need adaption in FIPS mode as it fails with:

There are other uses of MD5 and thus we can't disable it. For example gpgsm also lists the MD5 fingerprint of certificates because they are still in use at some places.

Done (STABLE-BRANCH-2-2.40 for now)

Yes, I'll do that. Thanks for the reminder.

Well, the modern way, recommended by the FSFE, for license notices in source files is SPDX instead of verbose license notices. https://reuse.software/

In T6285#165459, @gniibe wrote:

Now, the use of AM_PATH_GPGME_PTHREAD shows warning. Also I update the documentation.

Pushed the change.

Now, the use of AM_PATH_GPGME_PTHREAD shows warning. Also I update the documentation.