hallo,
I did some more debugging, because Debian unstable + psi + gnupg 1.4-3 works on my workstation, but not my laptop.
After some digging, I found the difference:
the workstation has a small and clean gpg keyring, containing only my jabber key and the keys of some jabber-friends - the laptop contains my email-key and all keys of persons which sent signed-mails through mailinglists I'm on, over 16mb accumulated throughout the years - and naturally, many long expired keys and signatures.
On the non-working startup, psi calls gpg 6 times:
# gpg --version 0 # gpg --local-user 0x81E518CA --clearsign --list-options no-show-policy-urls --armor --textmode --output - - 0 # gpg --no-tty --fixed-list-mode --with-colons --list-secret-keys 0 # gpg --no-tty --list-secret-keys 0 # gpg --no-tty --fixed-list-mode --with-colons --list-public-keys 2 # gpg --no-tty --list-public-keys 2
The number is the exit-code of gpg. On the working setup, all calls return with an exit-code of zero.
As I discovered with other gpg scripts, gpg 1.4 finishes with code 2 (contrary to the documentation), when it encounters some key-problems, like:
[EMAIL PROTECTED]:~/p$ gpg --no-tty --fixed-list-mode --with-colons --list-public-keys >/dev/null
gpg: public key 9BC57565 is 23952030 seconds newer than the signature
gpg: public key 9BC57565 is 23952462 seconds newer than the signature
gpg: public key CB62BDA1 is 3167 seconds newer than the signature
gpg: public key CB62BDA1 is 3119 seconds newer than the signature
gpg: public key 2731EAB5 is 1902 seconds newer than the signature
gpg: buffer shorter than attribute subpacket
gpg: buffer shorter than attribute subpacket
gpg: buffer shorter than attribute subpacket
gpg: buffer shorter than attribute subpacket
gpg: buffer shorter than attribute subpacket
gpg: public key 35013FAB is 3270 seconds newer than the signature
gpg: public key 35013FAB is 3289 seconds newer than the signature
gpg: public key 35013FAB is 3243 seconds newer than the signature
So, it looks like exit-code 2 is a mere warning.
The following small patch to psi gets me gnupg support back up&running:
--- psi-0.9.3/cutestuff/openpgp/gpgop.cpp.before-haegar 2005-04-11
23:32:12.866642824 +0200
+++ psi-0.9.3/cutestuff/openpgp/gpgop.cpp 2005-04-11 23:32:12.866642824
+0200
@@ -494,7 +494,7 @@
finished(true);
}
else if(d->op == SecretKeys || d->op == PublicKeys) {
- if(!clean || code != 0) {
+ if(!clean || code < 0 || code > 2) {
doFail();
return;
}c'ya sven
--
The Internet treats censorship as a routing problem, and routes around it. (John Gilmore on http://www.cygnus.com/~gnu/)
-- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
