GNOME Bugzilla – Bug 104204
Seahorse crashes on startup
Last modified: 2018-08-03 00:03:43 UTC
Package: seahorse Severity: critical Version: 0.6.0 Synopsis: Seahorse crashes on startup Bugzilla-Product: seahorse Bugzilla-Component: general BugBuddy-GnomeVersion: 2.0 (2.0.3) Description: Description of Problem: toor@~]$ rpm -q seahorse seahorse-0.6.0-0 [toor@~]$ rpm -q gpgme gpgme-0.3.14-1 [toor@~]$ seahorse ** (seahorse:20921): CRITICAL **: file seahorse-context.c: line 108 (seahorse_context_init): assertion `err == GPGME_No_Error' failed Steps to reproduce the problem: 1. See above 2. 3. Actual Results: It crashes Expected Results: Should start up How often does this happen? Always Additional Information: RH 8.0 all upgrades applied Debugging Information: Backtrace was generated from '/usr/bin/seahorse' (no debugging symbols found)...[New Thread 8192 (LWP 20925)] 0x420ae169 in wait4 () from /lib/i686/libc.so.6
+ Trace 33037
Thread 1 (Thread 8192 (LWP 20925))
------- Bug moved to this database by unknown@bugzilla.gnome.org 2003-01-23 02:54 ------- Reassigning to the default owner of the component, seahorse-maint@bugzilla.gnome.org.
what version of gnupg do you have? if you have a 1.0 version, try upgrading to 1.2.
Yep that was it. After installation of gpg 1.2.1 it worked. Shouldn't there be a check in the source code? This shouldn't happen to a user :) /Richard
Well, this should be a dependecy of GPGME 0.3.14 package, or source itself when you compiled it. Anyway, I'll try to work on this later today.
It is actually GPGME that fails because of a check done in the Seahorse source code. But GPGME apparently doesn't check the version during installation, so hopefully jose can figure this out so it checks during seahorse configure.
*** Bug 104370 has been marked as a duplicate of this bug. ***
I'm not sure if checkig for GnuPG >= 1.2 belongs to Seahorse compilation scripts. I'm thinking on filing a bug on GPGME and also adding the check in the meantime. Could be Seahorse 0.6.1 released after this? There are a lot of bugs being filled about this problem.
Only 2 bugs, but since gpgme doesn't seem to be doing the check, seahorse should. If you can figure that out, I'll include it in 0.6.1. Hopefully the gpgme 0.4 series will do it.
*** Bug 105995 has been marked as a duplicate of this bug. ***
Created attachment 365374 [details] [review] x11: don't call gdk_display_get_name before gtk_init The x11 plugin calls gdk_display_get_name at load time to set the device name state. This no longer works, since gtk_init_check() happens later at open_device time, and GTK+ no longer allows gdk_display_get_name before gtk_init. This commit moves the gtk_init call earlier, to create_backend, so gdk_display_get_name works again.
Attachment 365374 [details] pushed as e953e47 - x11: don't call gdk_display_get_name before gtk_init
sorry guys, wrong bugzilla!
Closed as fixed, as GPG version checking was added to our build-scripts quite some time ago.