Btw. had to revert your unique ptr change ;-) I didn't want to raise the c++ level just for that.

ikloecker I have just added the ki18n main code to pinentry-qt as qti18n.cpp this fixes it for me. I have commented out everything but the base catalog.

Saw this again and the commit was not in the Stable 2.2 branch. I have cherry picked it. This should resolve this issue.

Yes I think changing the textinteraction flags for these labels would be fine. But as this is only for one customer we should probably add some config like "no links". I think the about dialog things are more problematic as they come from Frameworks.

Oh, this is something we should fix anyway because users when evaluating Kleopatra and making configuration changes regularly run "gpgconf --kill all" anyway. Could it be that the SCD DEVINFO --watch fails because the gpg-agent is not yet started again?

I am tending towards wontfix. The reason is here that the sender attempts to send HTML with inline pgp. Which is not supported. Then that HTML apparently tries to be mutlipart/related which is not supported for inline PGP. Then it would require us to correct a wrongly sent content type of the inline attachment so that outlook does not interpret it as a png. And in that Format it could even be that Attachment1.pgp is not encrypted but instead png data, as the content type indicates.

I have just checked both the installation script, which still installs gpgme-json.exe and the gpg4win-4 installer downloaded from gpg4win.org gpgme-json.exe is properly installed under <instdir>\bin gpgme-json.exe and under bin_64

Because, as a user, what do you do if you see "invalid object" you think that something is wrong with your data instead of trying to type the passphrase again.

As I understand it after the p12 decryption the output is just tried to be imported. With the wrong passphrase this is just garbage and can lead to different errors.

I've uploaded my testcerts to: https://heinecke.or.at/div/testzertifikate.tar.gz.gpg

Nice. The way with KS_GET and assuan is much better then I thought about using a temporary home and a --locate-key

While testing I noticed that another requirement was to hide the advanced button. I have added this myself.

Ingo: Exactly we have the problem that we don't compile build tools before building for the target. So we take the build tooling like kconfig_compiler from the system we compile on. This means that we compile with the tooling from debian buster. Except for Qt which handles stuff like that directly and builds for example moc and the other tools correcly for the build system first.

Ui, Kudos to you for seeing that possibility in Boost Graph and applying that. *Thumbs up*

Quick gen key is only used for the keygen in GpgOL and KMail. Kleopatra itself uses the old batch generate interface.