Antoine Beaupré wrote: > > The right fix would be to open an issue with gpg to introduce a > > --no-interactive switch which you could then use to override the gpg.conf > > option I think. If you're actually doing this, you could also try to rename > > this misleading option in the same patch. ;) > > I would think that --batch would imply --no-interactive, of course. :p > > Would you mind opening an issue upstream? Otherwise I can just track this > here. > > > I would not mind at all since I don't use monkeysign. I don't even remember > > what was missing or if it just was this bug that kept me from using it. Hell > > I don't know if the bug is still occuring after 3 years. > > Understood. > > The only thing left is whether we forward it to the upstream GPG bugtrackers > or assign it to the gnupg package in the Debian BTS. > > We would need a backtrace first, of course, in any case.
Heyho Antoine, as far as I remember it was not a crash as in SIGSEGV or similar but rather a controlled generic error case of monkeysign. I'm very short on spare time right now, so I won't be of much help investigating this further, but if you're on short time as well, I'll keep this bug on my radar and return to it later (I'd probably ask the gnupg guys about the flag naming and the inverse version of the flag on the mailing list first). --Markus
