]>
git.sur5r.net Git - i3/i3lock/log
Diogo Ferreira [Tue, 13 Sep 2016 19:59:54 +0000 (20:59 +0100)]
i3lock: Stop leaking the image_path dup. (#93)
koebi [Sun, 4 Sep 2016 16:22:49 +0000 (18:22 +0200)]
unlock automatically after authentication failure (#90)
This will make i3 unlock automatically after authentication failure and
correct input without pressing <Enter> again after the "wrong"-indicator
is gone.
Vasilij Schneidermann [Thu, 18 Aug 2016 18:53:19 +0000 (20:53 +0200)]
Set font face (#89)
eplanet [Sat, 13 Aug 2016 20:46:35 +0000 (22:46 +0200)]
Added version files for release tarball. Fixes #40 (#86)
Florian Zipperle [Wed, 20 Jul 2016 06:50:07 +0000 (08:50 +0200)]
Removed obsolete inactivity timeout (#84)
* Removed obsolete inactivity timeout
Michael Stapelberg [Mon, 20 Jun 2016 17:46:38 +0000 (19:46 +0200)]
i3lock.1: use signal names without SIG prefix (Thanks rodfersou)
see #32
Michael Stapelberg [Sat, 4 Jun 2016 17:26:57 +0000 (19:26 +0200)]
update CHANGELOG
Michael Stapelberg [Fri, 3 Jun 2016 15:46:29 +0000 (17:46 +0200)]
clang-format-3.5
Not sure why travis didn’t warn about this on the PR. The only reference
to the PR I can find is https://travis-ci.org/i3/i3lock/builds/
130470377
Michael Stapelberg [Fri, 3 Jun 2016 15:28:43 +0000 (17:28 +0200)]
add CONTRIBUTING.md
…so that people know we don’t want to extend i3lock’s image handling.
related to #81, #68, #31
Christoph Ruegge [Wed, 25 May 2016 20:21:12 +0000 (22:21 +0200)]
Add support for `xss-lock --transfer-sleep-lock' (#75)
Add support for `xss-lock --transfer-sleep-lock'
martin [Wed, 25 May 2016 20:19:17 +0000 (22:19 +0200)]
Flush xcb connection after opening fullscreen window (#74)
We need to ensure that the window handle is valid, i. e. the window is
actually created and accessible, before returning.
This is necessary because we immediately fork after returning, and the
child process opens its own X11 connection and expects the window handle
to be valid.
Fixes #46
Karl Tarbe [Sat, 30 Apr 2016 10:25:51 +0000 (13:25 +0300)]
Allow CTRL+J as enter and CTRL+H as backspace (#72)
fixes #71
Michael Stapelberg [Fri, 25 Dec 2015 21:10:06 +0000 (22:10 +0100)]
Show unlock indicator if password was entered during PAM verification
fixes #51
Michael Stapelberg [Fri, 25 Dec 2015 14:44:46 +0000 (15:44 +0100)]
Eat XKB_KEY_Delete and XKB_KEY_KP_Delete (Thanks bebehei)
fixes #50
Michael Stapelberg [Sat, 7 Nov 2015 13:51:02 +0000 (14:51 +0100)]
Remove libxkbfile dependency from README.md/comments
It has not been in use since commit
7eb43aeabd7a5eb71927520c427debb45eab05f2
Michael Stapelberg [Fri, 6 Nov 2015 08:01:49 +0000 (09:01 +0100)]
use trusty, install libxkbcommon-dev from wily
Michael Stapelberg [Fri, 6 Nov 2015 07:52:48 +0000 (08:52 +0100)]
travis: install clang-format-3.5 from llvm repository
Ubuntu utopic disappeared from archive.ubuntu.com, it’s EOL.
Michael Stapelberg [Thu, 5 Nov 2015 22:22:08 +0000 (23:22 +0100)]
fix pam_securetty: set PAM_TTY to getenv("DISPLAY")
fixes #43
Michael Stapelberg [Tue, 11 Aug 2015 19:50:42 +0000 (21:50 +0200)]
Remove DPMS support in favor of a wrapper script and xset(1).
fixes #32
Michael Stapelberg [Fri, 3 Jul 2015 06:30:44 +0000 (23:30 -0700)]
Merge pull request #24 from ixjlyons/timeout-doc
Indicate that the --inactivity-timeout option takes an argument.
Kenneth Lyons [Thu, 2 Jul 2015 21:44:06 +0000 (14:44 -0700)]
Indicate that the --inactivity-timeout option takes an argument.
Michael Stapelberg [Wed, 20 May 2015 07:14:39 +0000 (09:14 +0200)]
update CHANGELOG
Michael Stapelberg [Wed, 20 May 2015 07:00:19 +0000 (09:00 +0200)]
makefile: s/README/README.md/
Michael Stapelberg [Mon, 27 Apr 2015 16:44:57 +0000 (09:44 -0700)]
Merge pull request #14 from ony/master
Make pkg-config configurable
Nikolay Orlyuk [Sun, 26 Apr 2015 18:57:47 +0000 (21:57 +0300)]
Make pkg-config configurable
Michael Stapelberg [Fri, 24 Apr 2015 22:27:06 +0000 (00:27 +0200)]
Merge pull request #13 from Airblader/feature-12
Only redraw the screen if the unlock indicator is actually used.
Ingo Bürk [Fri, 24 Apr 2015 21:36:24 +0000 (23:36 +0200)]
Only redraw the screen if the unlock indicator is actually used.
fixes #12
Michael Stapelberg [Tue, 21 Apr 2015 20:51:58 +0000 (22:51 +0200)]
Merge pull request #11 from Deiz/update-notices
Update copyright notices and get rid of ranges
Deiz [Tue, 21 Apr 2015 15:47:51 +0000 (11:47 -0400)]
Update copyright notices and get rid of ranges
The script used to make these changes can be found at:
https://gist.github.com/Deiz/
32322020f76d23e2bf8f
Michael Stapelberg [Sun, 29 Mar 2015 20:59:39 +0000 (22:59 +0200)]
Merge pull request #10 from Deiz/modifiers
List generic pressed modifiers on failure
Deiz [Thu, 26 Mar 2015 07:06:18 +0000 (03:06 -0400)]
List generic pressed modifiers on failure
Alexandre Boeglin [Wed, 11 Feb 2015 19:52:07 +0000 (20:52 +0100)]
Add capslock and numlock indicators
When the unlock indicator warns the user that a password was typed
wrong, it now also reports the state of the capslock and numlock
modifiers.
Signed-off-by: Alexandre Boeglin <alex@boeglin.org>
Michael Stapelberg [Fri, 27 Mar 2015 07:31:58 +0000 (08:31 +0100)]
Merge pull request #9 from qfire/master
refresh pam credentials on successful authentication
James Cassidy [Thu, 26 Mar 2015 13:26:48 +0000 (09:26 -0400)]
refresh pam credentials on successful authentication
Michael Stapelberg [Tue, 17 Mar 2015 08:00:15 +0000 (09:00 +0100)]
disable clang temporarily
Michael Stapelberg [Tue, 17 Mar 2015 07:50:50 +0000 (08:50 +0100)]
commit missing .clang-format
Michael Stapelberg [Tue, 17 Mar 2015 07:42:49 +0000 (08:42 +0100)]
add travis config
Michael Stapelberg [Mon, 16 Mar 2015 17:47:16 +0000 (18:47 +0100)]
clang-format -i **/*.[ch]
From here on, we’ll use clang-format to automatically format the source.
This has worked well in i3, so we are introducing it for i3lock.
Michael Stapelberg [Thu, 12 Feb 2015 22:08:35 +0000 (23:08 +0100)]
Merge pull request #6 from danielotero/master
Add support for Compose and dead-keys with libxkbcommon
Daniel Otero [Thu, 12 Feb 2015 20:38:37 +0000 (21:38 +0100)]
Update the key symbol if the composing ends.
Daniel Otero [Thu, 12 Feb 2015 10:22:58 +0000 (11:22 +0100)]
Avoid handle a keysym if the compose state gets cancelled
This is the approach taken by libX11, and feels more consistent.
Daniel Otero [Thu, 12 Feb 2015 10:12:38 +0000 (11:12 +0100)]
Remove xkb_context initialization code from load_compose_table
xkb_context is guaranteed to be initializated from load_keymap.
Daniel Otero [Wed, 11 Feb 2015 23:37:23 +0000 (00:37 +0100)]
Add support for Compose and dead-keys with libxkbcommon
Michael Stapelberg [Tue, 6 Jan 2015 07:34:35 +0000 (08:34 +0100)]
Merge pull request #3 from stibi/whoami
getting current user with whoami like function rather than from env variable
Martin Stiborsky [Mon, 5 Jan 2015 23:02:23 +0000 (00:02 +0100)]
getpwuid call moved
Martin Stiborsky [Mon, 5 Jan 2015 21:54:32 +0000 (22:54 +0100)]
null check for getpwuid and pw_name
Martin Stiborsky [Mon, 5 Jan 2015 18:54:48 +0000 (19:54 +0100)]
removed the whoami function, replaced with an inline getpwuid call
Martin Stiborsky [Mon, 5 Jan 2015 10:46:11 +0000 (11:46 +0100)]
getting current user with whoami like function rather than from env variable
Michael Stapelberg [Sat, 27 Dec 2014 21:21:33 +0000 (22:21 +0100)]
Merge pull request #1 from koebi/master
added debug-flag to manpage
koebi [Sat, 27 Dec 2014 21:19:07 +0000 (22:19 +0100)]
debugging outputs the password, so the manpage should mention that
koebi [Sat, 27 Dec 2014 20:54:41 +0000 (21:54 +0100)]
added debug-flag to manpage
Michael Stapelberg [Fri, 26 Dec 2014 20:03:12 +0000 (21:03 +0100)]
README is actually in markdown, so tell github that
Michael Stapelberg [Fri, 26 Dec 2014 20:02:58 +0000 (21:02 +0100)]
make github the canonical i3lock repository
Neodyblue [Mon, 13 Oct 2014 10:39:13 +0000 (11:39 +0100)]
Ignore password validation is pam is in wrong state
This add a protection against wrong password spamming
koebi [Thu, 24 Jul 2014 20:32:07 +0000 (22:32 +0200)]
implemented logging the number of failed attempts
This closes #1243
Michael Stapelberg [Thu, 21 Aug 2014 07:34:03 +0000 (09:34 +0200)]
Die when the X11 connection breaks during runtime (Thanks Eduan)
Michael Stapelberg [Fri, 18 Jul 2014 18:21:09 +0000 (20:21 +0200)]
update the changelog/readme for 2.6
Michael Stapelberg [Sat, 18 Jan 2014 20:12:59 +0000 (21:12 +0100)]
remove obsolete xkb_state_update_key() calls (Thanks Ran)
Michael Stapelberg [Sat, 18 Jan 2014 18:40:11 +0000 (19:40 +0100)]
listen for XKB events instead of the MappingNotify event (Thanks Ran)
Michael Stapelberg [Sat, 18 Jan 2014 18:20:58 +0000 (19:20 +0100)]
little cleanups thanks to Ran Benita’s review
Michael Stapelberg [Sat, 18 Jan 2014 16:13:59 +0000 (17:13 +0100)]
use libxkbcommon-x11 instead of libX11
Michael Stapelberg [Fri, 2 May 2014 17:57:22 +0000 (19:57 +0200)]
Scale the unlock indicator (for retina displays)
Vytautas Šaltenis [Sun, 23 Mar 2014 17:14:09 +0000 (19:14 +0200)]
Cleanup scattered timer calls to use macros
Vytautas Šaltenis [Mon, 10 Mar 2014 20:02:27 +0000 (22:02 +0200)]
Ignore empty passwd after repeated Enter keypress
When no ignore_empty_password flag is specified, but Enter key is held
down for some time, i3lock enters a finite, but hours-long loop
verifying empty password. Don't do that, skip empty password
verification if nothing was pressed after last Enter keypress.
Again, making the software cat-proof.
Vytautas Šaltenis [Tue, 18 Mar 2014 20:40:17 +0000 (22:40 +0200)]
Fix typo in manual
Vytautas Šaltenis [Thu, 6 Mar 2014 21:16:43 +0000 (23:16 +0200)]
Fix typo
Vytautas Šaltenis [Thu, 6 Mar 2014 21:06:28 +0000 (23:06 +0200)]
Discard half-entered passwd after some inactivity
When you start typing a password and then wander off, let's discard that
attempt after a few minutes of inactivity (and turn off the monitors if
dpms is on).
This is a real actual case of making the software cat-proof: when my cat
treads over the keyboard at night, I don't want the monitors to shine
for the rest of the night.
Vytautas Šaltenis [Thu, 6 Mar 2014 20:55:37 +0000 (22:55 +0200)]
Delay to turn off the screen after wrong passwd
Also add flag for number of seconds to wait for it.
When user cancels the password or enters a wrong one, i3lock will wait
for some time before putting the monitors back to sleep. By default it's
30 seconds, but this flag allows to control that.
TheKK [Sun, 16 Mar 2014 13:54:42 +0000 (21:54 +0800)]
Replace '1' with "EXIT_FAILURE"
Vytautas Šaltenis [Thu, 6 Mar 2014 20:22:02 +0000 (22:22 +0200)]
Extract start/stop timer code to reusable funcs
Vytautas Šaltenis [Sun, 2 Mar 2014 21:53:42 +0000 (23:53 +0200)]
Extract if(dpms) calls to wrapper functions
Philippe Virouleau [Sun, 10 Nov 2013 22:12:46 +0000 (23:12 +0100)]
Turn on the screen on successful authentication
As described in ticket #1114, the screen may be left turned off on successful
authentication. This commit fixes this behaviour by turning the screen back on
after the authentication.
Fixes #1114
Michael Stapelberg [Sun, 10 Nov 2013 19:40:52 +0000 (20:40 +0100)]
Re-raise i3lock when obscured in a separate process
This is cleaner than commit
a305e62 (running authentication in a
separate process) because we don’t lose the feature that X11 will buffer
KeyPresses for us while the process blocks on PAM. See
http://cr.i3wm.org/patch/305 for the amount of code (and new known bugs)
that would have to be introduced to make
a305e62 work.
Michael Stapelberg [Sun, 10 Nov 2013 19:03:53 +0000 (20:03 +0100)]
Revert "Run authentification in different process"
This reverts commit
a305e622a6cad3f9ba267d6810fee54b7c7c86c3 .
In the next commit, the same functionality will be achieved in a cleaner
way.
Michael Stapelberg [Sun, 10 Nov 2013 19:03:48 +0000 (20:03 +0100)]
Revert "Bug fix: Call clear_input() when the password is wrong."
This reverts commit
33d953fbdf4f8630e919643974f32ce6793c7c2b ,
in preparation for reverting commit
a305e62
oblique [Thu, 17 Oct 2013 10:35:33 +0000 (13:35 +0300)]
Bug fix: Call clear_input() when the password is wrong.
clear_input() calls start_clear_indicator_timeout() to clear the
indicator after 1 second. This had as a result the screen to be cleared
when the "verifying…" message was showed, so the "wrong!" message didn't
show up when the password was wrong.
First bad commit:
a305e62
koebi [Sat, 7 Sep 2013 17:41:16 +0000 (19:41 +0200)]
Run authentification in different process
When having i3bar in “hide” mode, it could previously be brought up
while the authentication was running. This is not 100% fixed with this
change, but a lot better, as the normal “raise i3lock to the top”
behavior is not blocked now anymore.
fixes #895
Vincent Bernat [Sat, 31 Aug 2013 12:00:16 +0000 (14:00 +0200)]
Small improvements to manual page.
Use a bulleted list for the IMPROVEMENTS section. Fix some
grammar. Remove the use of "fuck" which may offend some people.
Michael Stapelberg [Sun, 1 Sep 2013 08:01:38 +0000 (10:01 +0200)]
Bugfix: update the initial keyboard modifier state (Thanks lee, Ran)
This should make the numpad work correctly.
Michael Stapelberg [Sat, 31 Aug 2013 18:42:06 +0000 (20:42 +0200)]
Add -e to --help
Vincent Bernat [Mon, 26 Aug 2013 21:39:38 +0000 (23:39 +0200)]
Add `-e` option to not validate empty password.
When the XF86ScreenSaver key is used to put a laptop to sleep (or to
trigger the screensaver), the key may "bounce" on resume. This is
annoying as i3lock will try to validate several empty passwords and
wait several seconds before accepting a legit password.
Some users may want to validate an empty password: PAM may rely on
other sources to unlock the screen, like the presence of a token or
the proximity of some Bluetooth device. Hence, we don't forbid this
possibility and provide an command-line option for users not willing
to validate empty passwords.
Bas Pape [Tue, 30 Jul 2013 18:03:19 +0000 (20:03 +0200)]
Use cairo_status_to_string for descriptive errors.
Michael Stapelberg [Sun, 9 Jun 2013 09:44:57 +0000 (11:44 +0200)]
add xcb-atom to pkg-config calls
Michael Stapelberg [Sun, 9 Jun 2013 09:13:39 +0000 (11:13 +0200)]
update changelog for 2.5
Axel Wagner [Tue, 9 Apr 2013 16:31:25 +0000 (18:31 +0200)]
Set window name
Sebastian Ullrich [Thu, 28 Mar 2013 23:46:22 +0000 (00:46 +0100)]
Fix centering the indicator in the no-xinerama case
Michael Stapelberg [Fri, 29 Mar 2013 09:13:01 +0000 (10:13 +0100)]
fix ctrl-u handling
Sebastian Ullrich [Thu, 28 Mar 2013 23:43:43 +0000 (00:43 +0100)]
Fix resetting the password after failed authentications
This has been broken by commit 661344.
Philip Dexter [Sat, 23 Mar 2013 16:33:56 +0000 (12:33 -0400)]
Add ctrl+u password reset
Baptiste Daroussin [Wed, 20 Mar 2013 16:48:21 +0000 (17:48 +0100)]
Use ev_loop_fork after fork, this fix forking on kqueue based OSes
Michael Stapelberg [Sun, 10 Feb 2013 14:18:54 +0000 (15:18 +0100)]
README: refer to cr.i3wm.org for patches
Quentin Glidic [Thu, 7 Feb 2013 16:14:53 +0000 (17:14 +0100)]
Limit the debug print to the actual password size
Signed-off-by: Quentin Glidic <sardemff7+git@sardemff7.net>
Vivien Didelot [Thu, 6 Dec 2012 05:38:45 +0000 (00:38 -0500)]
RGB color notation is 3-byte hexadecimal
Signed-off-by: Vivien Didelot <vivien.didelot@gmail.com>
Moritz Augsburger [Sat, 3 Nov 2012 10:32:07 +0000 (10:32 +0000)]
Allow passwordless PAM conversations (e.g. fingerprint)
Moritz Augsburger [Sat, 3 Nov 2012 10:31:29 +0000 (10:31 +0000)]
Allow XKB_KEY_XF86ScreenSaver as synonm for enter
This keysym is generated on convertible tablets by pressing a hardware
lock/unlock button.
Michael Stapelberg [Fri, 26 Oct 2012 18:42:26 +0000 (20:42 +0200)]
remove support for NOLIBCAIRO, cairo-xcb is widespread by now
Michael Stapelberg [Fri, 26 Oct 2012 18:40:20 +0000 (20:40 +0200)]
README: remove unhelpful comment about pkg-config
Michael Stapelberg [Fri, 26 Oct 2012 18:24:55 +0000 (20:24 +0200)]
use libxkbcommon for input handling
Thanks to Ran Benita and Daniel Stone (the libxkbcommon authors) for
answering my questions and reviewing this code.
With this commit, input handling should be more correct with using less
code (in i3lock, that is).
Michael Stapelberg [Fri, 26 Oct 2012 18:21:53 +0000 (20:21 +0200)]
bugfix: free(reply) to avoid memory leak
Michael Stapelberg [Mon, 15 Oct 2012 20:39:49 +0000 (22:39 +0200)]
s/pam_message/pam_response/ (Thanks Tucos)
The previous code didn’t cause any issues since both types (currently!)
have the same size.