Christian Buhtz

OK, than it makes sense to me. This PR will go into 1.6.0 release and is by definition quit unstable. So bringing it into the wild will give us proof.

I am confused. Still failing on TravisCI. ``` test/test_sshtools.py::General::test_unlockSshAgent PASSED [ 67%] test/test_sshtools.py::General::test_unlockSshAgentKeyWithPassword Enter password for SSH private key profile "Main profile": ```

> I'm guessing from the logs that Travis doesn't have any secret services running? Yes, I do think the same. While reading the test code I do understand it on...

OK. And specific to this test case, who should provide the password for that key? backintime-askpass?

Mhm... OK. I think we should take care of our resources. It might not be ideal but also it is no big deal just skipping that test on Travis. ```sh...

> Did you still want to include `SSH_ASKPASS_REQUIRE=force` to force SSH to use `backintime-askpass` when `DISPLAY` is unset? 😄 I am not sure. But you made this modification because of...

> Some progress... Ah! I ❤️ that game!

I dropped the `make install` from travis config script for some stupid and today unknown reason. 😭 ``` script: # compile all files - ensure that syntax is correct -...

Great. The askpass workaround is OK for me. Can not imagine any security implications. So, do you think we are ready for merging?

I did some further line-by-line analysis of the code in serviceHelper.py and from the python-dbus docs. The goal was to identify python-dbus-docs related lines and find out if I can...