Hello Matt. On Sun, Aug 30, 2009 at 08:07:05PM -0700, Matt Taggart wrote: > Hi Klaus, > > The bug #509554 "pinentry-curses destroys terminal settings" is a bug you > filed against version 0.7.2-3 of pinentry-curses, which was the version > included in etch and is quite old.
I've been having the same issue forever with regular use of the
gnupg.vim script.
I'm running latest Debian testing packages.
>
> It's a little unclear how to repeat it, I just trying running it and doing
> a CONFIRM to get it to prompt me, then after that I ran vim and I could
> still move the cursor. I tried in both 0.7.2-3 (etch, where you reported
> the problem) and 0.7.5-2.1 (lenny) and it worked, so I guess I'm not sure
> how to repeat it.
>
> Can you still repeat the problem with a newer pinentry-curses?
With proper GnuPG configuration, try using the attached script to
reproduce instantly. You don't even need to enter a passphrase,
cancelling will do just fine.
The key command being:
vi -c ':r!gpg --decrypt encrypted.gpg'
Once inside Vim you will find that terminal settings are broken (easy to
test with up/down arrows). Running reset (:!reset) should fix things. On
my system, "stty sane" does not seem to be needed.
Another side effect of this bug is that, with appropriate t_ti and t_te
values for termcap switching, when Vim exits, the content of the session
is left on the screen. This leads me to think that pinentry-curses
messes with the terminal state or at least switches out of termcap mode,
leaving Vim in an inconsistent state.
Cheers,
Jean-Sébastien Trottier
>
> Thanks,
>
> --
> Matt Taggart
> [email protected]
test-vim-debian-bug-509554.sh
Description: Bourne shell script
signature.asc
Description: Digital signature
