Wer eine #Arch-basierte #Distribution verwendet, sollte mit Problemen rechnen und diese beheben können. Entweder mit eigenem Wissen oder mithilfe der #Community.
Wer eine #Rolling #Distribution verwendet, muss früher oder später mit Problemen rechnen. Das gilt auch für das kuratiert-rollende #Manjaro. Es können jahrelang überhaupt keine Schwierigkeiten auftreten, oder es kommt auf einmal knüppeldick.
After thinking about it a bunch, I have decided that I'll refactor my cryptographic deadhand to use python-gnupg until the sequoia-sop python bindings are released.
The official #GPGME bindings are just too damn broken to be of any real use – and I think that says *a lot*.
Honestly, I'm not at all sure how people can release something like that as "production grade" (for security-critical tooling, no less) and not feel deeply ashamed.
Further #GPGME python binding adventures.
gpg.Context.key_import, unlike what the docs claim, doesn't raise any Exception if you pass in the wrong kind of data, but rather just returns the string "ERROR" and neither return values nor what type of data the function expects are documented.
I continue to be amazed by how #GnuPG manages to be broken for pretty much every. single. thing.
I've been trying for an hour to get the python bindings of #gpgme to generate a key and it just keeps failing with the most generic error message imaginable: gpgme_op_createkey: GPGME: General error
Literally the only information I get out of that is that the key creation failed – no duh… 🙄
GNU/Linux.ch
David Runge
Begasus
Rafael Sadowski
smjn
Ben
phryk 🏴