tint2 becomes unresponsive

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Run tint2 and use it for awhile
2. Happens eventually/intermittently
3.

What is the expected output? What do you see instead?
Tint2 should run normally but instead it becomes unresponsive after using it 
for awhile and have to be killed manually.  Could not determine a pattern for 
reproducing it consistently.

What version of the product are you using? On what operating system?
started happening in some release after r678, could not precise which.  
Currently using r692, but also happened in r687 and r690.

Please provide any additional information below.
I know this is a little vague, but it is the best I can report, sorry that I 
cannot provide more details.  Will keep trying to determine a pattern for 
reproducting.

Original issue reported on code.google.com by marcelof...@gmail.com on 14 Feb 2015 at 1:07

[GoogleCodeExporter]

Thanks for your report. We have not seen hangs recently, so we will need some 
help to figure out what is happening.

Answering a few more questions may help us:

Could you please provide a tint2rc, so we can try to reproduce it and also 
check if there is anything suspicious in the config?

When tint2 hangs, does CPU usage go to 100% on a core?

Which window manager (e.g. openbox, xfwm, metacity, mutter, kwin) or which 
desktop environment are you using?

Is it a single monitor or multi monitor setup?

Original comment by mrovi9...@gmail.com on 14 Feb 2015 at 10:32

• Changed state: NeedInfo
• Added labels: Type-Defect

[GoogleCodeExporter]

Could you please provide a tint2rc, so we can try to reproduce it and also 
check if there is anything suspicious in the config?
> Yes, will paste it in the next comment below

When tint2 hangs, does CPU usage go to 100% on a core?
> Not that I've noticed but will have to confirm the next time it happens

Which window manager (e.g. openbox, xfwm, metacity, mutter, kwin) or which 
desktop environment are you using?
> I'm using gnome-shell (Gnome3) on Fedora20.

Is it a single monitor or multi monitor setup?
> Single monitor

OBS: I use this tint2rc since 2012.  The only changes are occasional adding or 
removing a launcher here and there.  The freezing started only after the recent 
updates (that I could notice after r678).

Original comment by marcelof...@gmail.com on 14 Feb 2015 at 10:45

[GoogleCodeExporter]

The tint2rc:

# Tint2 config file
# Generated by tintwizard (http://code.google.com/p/tintwizard/)
# For information on manually configuring tint2 see 
http://code.google.com/p/tint2/wiki/Configure

# Background definitions
# ID 1
rounded = 7
border_width = 2
background_color = #000000 60
border_color = #FFFFFF 0

# ID 2
rounded = 5
border_width = 0
background_color = #FFFFFF 24
border_color = #FFFFFF 40

# ID 3
rounded = 5
border_width = 0
background_color = #FFFFFF 16
border_color = #FFFFFF 40

# ID 4
rounded = 5
border_width = 2
background_color = #FFFFFF 0
border_color = #FFFFFF 27

# ID 5
rounded = 5
border_width = 2
background_color = #FFFFFF 0
border_color = #FFFFFF 66

# ID 6
rounded = 5
border_width = 1
background_color = #000055 50
border_color = #FFFFFF 100

# ID 7
rounded = 5
border_width = 1
background_color = #FFFFFF 24
border_color = #FFFFFF 40

# ID 8
rounded = 5
border_width = 0
background_color = #FFFFFF 0
border_color = #FFFFFF 40

# Panel
panel_items = LCT
panel_monitor = all
#panel_position = bottom center horizontal
panel_position = bottom left horizontal
panel_size = 97% 070
panel_margin = 2 0
panel_padding = 6 4 6
panel_dock = 1
wm_menu = 0
panel_layer = normal
panel_background_id = 1

# Panel Autohide
autohide = 0
autohide_show_timeout = 0.3
autohide_hide_timeout = 2
autohide_height = 2
strut_policy = follow_size

# Taskbar
taskbar_mode = multi_desktop
#taskbar_mode = single_desktop
taskbar_padding = 2 3 2
taskbar_background_id = 4
taskbar_active_background_id = 5
taskbar_name = 0
taskbar_name_background_id = 3
taskbar_name_active_background_id = 2
taskbar_name_font = Visitor TT1 BRK 8
taskbar_name_font_color = #FFFFFF
taskbar_name_active_font_color = #FFFFFF

# Tasks
urgent_nb_of_blink = 8
task_icon = 1
task_text = 0
task_centered = 0
#task_maximum_size = 300 35
task_maximum_size = 49 35
#task_maximum_size = 10 10
task_padding = 2 2
task_background_id = 3
task_active_background_id = 7
task_urgent_background_id = 2
task_iconified_background_id = 3
task_tooltip = 1

# Task Icons
task_icon_asb = 70 0 0
task_active_icon_asb = 100 0 0
task_urgent_icon_asb = 1000 0 0
task_iconified_icon_asb = 70 0 0

# Fonts
task_font = Sans 7
task_font_color = #FFFFFF 80
task_active_font_color = #FFFFFF 80
task_urgent_font_color = #FFFFFF 80
task_iconified_font_color = #FFFFFF 60
font_shadow = 0

# System Tray
systray = 0
systray_padding = 0 4 5
systray_sort = ascending
systray_background_id = 0
systray_icon_size = 16
systray_icon_asb = 70 0 0

# Tooltips
tooltip = 0
tooltip_padding = 2 2
tooltip_show_timeout = 0.7
tooltip_hide_timeout = 0.3
tooltip_background_id = 6
tooltip_font = Sans Italic 10
tooltip_font_color = #FFFFFF 100

# Mouse
mouse_left = toggle_iconify
mouse_middle = close
mouse_right = none
mouse_scroll_up = toggle
mouse_scroll_down = iconify

# Battery
battery = 0
battery_low_status = 10
battery_low_cmd = notify-send "battery low"
battery_hide = 98
bat1_font = sans 8
bat2_font = sans 6
battery_font_color = #FFFFFF 60
battery_padding = 1 0
battery_background_id = 6

# Clock
#time1_format = %l:%M
time1_format = |||
time1_font = Visitor TT1 BRK 30
#clock_font_color = #ffffff 76
clock_font_color = #ffffff 80
#clock_font_color = #00ff00 80
#clock_padding = 8 2
clock_padding = 5 0
#clock_background_id = 8
clock_background_id = 8
clock_rclick_command = sh -c "xset r on"
clock_lclick_command = sh -c "killall -15 conky ; sleep 0.1 ; killall -15 tint2 
; tint2 & sleep 0.1 ; ~/Conky/conky.sh &"

# Launcher
launcher_icon_theme = Adwaita
launcher_padding = 5 3 2
launcher_background_id = 7
launcher_icon_size = 24
launcher_item_app = 
/home/marcelo/.config/tint2/launchers/cxmenu-cxoffice-cxsetup.desktop
launcher_item_app = 
/home/marcelo/.config/tint2/launchers/cxmenu-cxoffice-Microsoft+Office+Excel+200
7.desktop
#launcher_item_app = /usr/share/applications/vmware-player.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/vmware-player.desktop
launcher_item_app = 
/home/marcelo/.config/tint2/launchers/cxmenu-cxoffice-Microsoft+Office+Word+2007
.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/dosbox.desktop
#launcher_item_app = 
/home/marcelo/.config/tint2/launchers/cxmenu-cxoffice-Microsoft+Office+Outlook+2
007.desktop
launcher_item_app = /usr/share/applications/screenruler.desktop
launcher_item_app = /usr/share/applications/arduino.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/lightscribe.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/amulegui.sh.desktop
launcher_item_app = /usr/share/applications/skype.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/domination.sh.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/fedora-kid3.desktop
launcher_item_app = /usr/share/applications/gnomine.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/xsane.desktop
launcher_item_app = /usr/share/applications/xmahjongg.desktop
launcher_item_app = /usr/share/applications/fslint.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/xmms.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/chrome-sling.desktop
launcher_item_app = /usr/share/applications/livna-xv.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/barscan.sh.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/hp48.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/nedit.desktop
#launcher_item_app = 
/home/marcelo/.config/tint2/launchers/chrome-linktrade.desktop
#launcher_item_app = /home/marcelo/.config/tint2/launchers/acddialog.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/chrome-icloud.desktop
launcher_item_app = 
/home/marcelo/.config/tint2/launchers/g-large-terminal.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/localevolution.desktop
#launcher_item_app = /home/marcelo/.config/tint2/launchers/chrome-plex.desktop
launcher_item_app = /usr/share/applications/plexhometheater.desktop
launcher_item_app = /usr/share/applications/firefox.desktop
launcher_item_app = /usr/share/applications/putty.desktop
#launcher_item_app = /home/marcelo/.config/tint2/launchers/opera-browser.desktop
launcher_item_app = /usr/local/share/applications/google-chrome.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/putty-jupiter.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/xlock.desktop
launcher_item_app = /home/marcelo/.config/tint2/launchers/g-terminal.desktop

# End of config

Original comment by marcelof...@gmail.com on 14 Feb 2015 at 10:46

[GoogleCodeExporter]

OBS: The "time1_format" line above have been manually edited before pasting it 
here because I use a unicode character and I didn't know if it would paste 
correcly in this page.  It is actually a recycle symbol that can be obtained in 
bash by running 'printf "\u267B\n"'

Original comment by marcelof...@gmail.com on 14 Feb 2015 at 11:01

[GoogleCodeExporter]

All right, thanks for the quick reply. I will run tint2 with this config for a 
while and will let you know. Nice config BTW :)

Original comment by mrovi9...@gmail.com on 14 Feb 2015 at 11:16

[GoogleCodeExporter]

I've been working all day using your config file and no hang happened yet :(

Another option would be for you to compile tint2 in debug mode and get a stack 
trace and/or a core dump and send it to us. This is a lot of trouble, but if 
you are willing to do it, here is how: 
https://code.google.com/p/tint2/wiki/Debug

Please keep in mind that a core dump of tint2 will contain all its memory 
contents, including the list of windows you have open and their titles (from 
the taskbar). Make sure there is nothing sensitive in there before posting it 
on the Internet.

Original comment by mrovi9...@gmail.com on 14 Feb 2015 at 11:43

[GoogleCodeExporter]

Yes, didn't happen here since I reported it either, although before it happened 
some times.  Must be a weird combination of factors.  Will continue monitoring 
and trying to determine a pattern for reproducing.

Thanks.

Original comment by marcelof...@gmail.com on 16 Feb 2015 at 5:14

[GoogleCodeExporter]

Ok, happened again.  No specific pattern yet, but no CPU core spinning.

I noticed when I clicked in a launcher and nothing happened (meaning it may 
have happened before or in the exact moment).

A theory:  I've seen in the changelogs that the clock engine has been updated 
to make the seconds to update in some conditions where it didn't.  (For a long 
time now) I use the clock section just to print some characters (not a clock).  
Could it be related?

Original comment by marcelof...@gmail.com on 18 Feb 2015 at 12:14

[GoogleCodeExporter]

I don't think it is related to the clock change. Actually I looked over the 
last few commits and I did not see anything that could cause problems.

If there is no CPU spinning, it might mean that tint2 stops interacting with X 
correctly. That could happen if tint2 is doing something wrong or if another 
component of the graphics stack is broken. Did you upgrade anything else 
recently? Maybe X, your desktop environment or your compositor?

You could also start tint2 like this to log its output to a file:

(tint2 &>> ~/tint2.log &)

After it hangs, check the last few lines and see if there is anything unusual 
(an error message). But I think this problem will not be solved without 
debugging.

Original comment by mrovi9...@gmail.com on 20 Feb 2015 at 10:21

[GoogleCodeExporter]

Hi.  Follows the last lines of the tint2.log (as instructed above) from a hang 
that just happened.  (Some of the messages seem to be from chrome, but the 
d-bus one seems very suspect.  Note though that after after killing tint2 and 
relaunching it, the new instance works just fine):

--- CUT HERE ---
[WARNING:flash/platform/pepper/pep_module.cpp(63)] SANDBOXED
java version "1.7.0_75"
OpenJDK Runtime Environment (fedora-2.5.4.2.fc20-x86_64 u75-b13)
OpenJDK 64-Bit Server VM (build 24.75-b04, mixed mode)
[8433:8433:0221/110732:ERROR:gconf_listener.cc(116)] Error with gconf key 
'/apps/metacity/general': Configuration server couldn't be contacted: D-BUS 
error: The GConf daemon is currently shutting down.
ATTENTION: default value of option force_s3tc_enable overridden by environment.
[WARNING:flash/platform/pepper/pep_module.cpp(63)] SANDBOXED
Fontconfig error: Cannot load default config file

Original comment by marcelof...@gmail.com on 21 Feb 2015 at 1:35

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

One additional note:  This time it seems that it happened just after I close an 
applicaiton (x48 an HP48 calculator emulator).  I say this because I closed it 
and the closing was not reflected in tint2 (i.e., its icon stayed there), so I 
noticed it hanged.

Original comment by marcelof...@gmail.com on 21 Feb 2015 at 1:41

[GoogleCodeExporter]

Thanks, unfortunately I think none of the messages are displayed by tint2. They 
are displayed by programs started through the launcher.

The GConf message is most likely displayed by Chrome ( 
https://chromium.googlesource.com/chromium/src/+/master/chrome/browser/ui/libgtk
2ui/gconf_listener.cc#116 ). I don't know why "metacity" shows up, since it is 
a Gnome 2 component which should not exist in Gnome 3.

Anyways... I am currently downloading Fedora 20, will install it in a VM, 
hopefully I will manage to reproduce the bug.

Original comment by mrovi9...@gmail.com on 21 Feb 2015 at 6:44

[GoogleCodeExporter]

I've got it running. Will let you know how it goes.

Original comment by mrovi9...@gmail.com on 21 Feb 2015 at 7:35

[GoogleCodeExporter]

Hi.

Does tint2 waits for the windows of children it spawns from its launchers?  
This is why I'm asking:

It happened again when I was using the x48 emulator (this doesn't happen 
always, but happening twice after using the emulator made it more likely to be 
related).

Please note that this doesn't mean the emulator is the culprit as I explain 
below, but may help detecting where the issue is, just an hyposthesis, because 
I use wmctrl many times in scripts when launching it.

I have written a scheme of scripts to avoid the to be launched twice and reuse 
the existing instance if one is already runnging.  This is something I used for 
years already and uses wmctrl to check if an existing instance is running.  If 
there is an already running instance, I use wmctrl again to bring it to the 
current workspace, otherwise I launch it and use wmctrl to turn on the "always 
on top" flag of its window.  Follows the script and the .desktop launcher:

----- The launcer (hp48.desktop): -----
#!/usr/bin/env xdg-open

[Desktop Entry]
Encoding=UTF-8
Version=1.0
Type=Application
Terminal=false
Name[C]=HP48
Exec=/home/marcelo/bin/x48ontop
Name=HP48
Icon[C]=/home/marcelo/GnomeIcons/x48b.jpg
Icon=/home/marcelo/GnomeIcons/x48b.jpg
Name[en_US]=HP48
Icon[en_US]=/home/marcelo/GnomeIcons/x48b.jpg
----- END OF LAUNCHER ------

----- The x48ontop script: -----
#!/bin/sh

wmctrl -xR x48.X48

if [ $? == 1 ]; then
    x48 +xshm -title "HP48 Calculator" &
    sleep 1
    wmctrl -xr x48.X48 -b add,above
fi
----- END OF SCRIPT ------

For the most close scenario to reproduce, the emulator itself can be downloaded 
here: http://sourceforge.net/projects/x48.berlios/

OBS: This is just an hypothesis as this is not a way to reproduce the issue 
100% or even 50% of the times, but the last two times it happened it was after 
I launched the emulator and before clicking in tint2 again.

Original comment by marcelof...@gmail.com on 25 Feb 2015 at 12:08

[GoogleCodeExporter]

Thanks, I will try this during the weekend.

I kept the Fedora vm running until today but apart from some graphical 
corruptions around the mouse cursor, no hang occurred.

However I did not try running x48 and I did not use the machine intensively, 
just for browsing and starting apps randomly from time to time.

In the meantime I have two suggestions:

1. Compile tint2 without startup notification support. You need to pass the 
following parameters to cmake:

cmake -DENABLE_SN=OFF ..

2. Disable transparency. For this you would have to add to tint2rc the line:

disable_transparency = 1

Original comment by mrovi9...@gmail.com on 26 Feb 2015 at 6:59

[GoogleCodeExporter]

Thank you.  Will try disabling the Startup Notification support and disabling 
the transparency this weekend.

Please notice that as I explained, I don't think that x48 is the culprit, but 
possible one of the use cases of wmctrl that I use in its start scripts from 
the launcher I created.  Unfortunately the issue didn't happen since the last 
report.

Thanks

Original comment by marcelof...@gmail.com on 26 Feb 2015 at 11:47

[GoogleCodeExporter]

Just a follow-up:

It did happen twice again since last report.  Both times when using the x48 
emulator and the wmctrl scritps used to start it.

Will start using tint2 with the startup notifications disabled today.

Thanks

Original comment by marcelof...@gmail.com on 1 Mar 2015 at 1:15

[GoogleCodeExporter]

Thanks for the info, that's a good lead. Unfortunately I have very little free 
time at the moment :( so it will be a few days until I can look at this.

Original comment by mrovi9...@gmail.com on 2 Mar 2015 at 5:21

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Follow-up:

Today it completes a week I'm using r727 compiled using `cmake -DENABLE_SN=OFF 
..` and the issue didn't happen any single time, even with frequent use of the 
scenarios it was using to happen.

OBS: It occurred me that the issue may be related to the fact I have the x48 
emulator running with its window always on top (so it should affect any other 
application with such feature turned on).  So, maybe, when a new window happens 
to open below the emulator, tint2 may not be receiving or handling the 
notification correctly (just another theory as I could not find yet a way of 
deterministically reproducing the issue).  Will insist a bit further with the 
`cmake -DENABLE_SN=OFF ..` version and if this doesn't happen again get back 
testing with normal compilation.

Thanks

Original comment by marcelof...@gmail.com on 8 Mar 2015 at 1:05

[GoogleCodeExporter]

I found this open issue when searching just now, since I've recently started to 
see hangs like this more often, as well. I remember this happening a year or 
two back, as well, but it only happened once. Now it happens nearly every day

It does *appear* random, and I cannot (as yet) connect it to any particular 
action I am doing.

Note that my setup is a little simpler. I'm running Arch Linux with FVWM on 
Intel graphics, and I am usually running a single monitor. I've seen it happen 
both on my laptop's screen and an external monitor.

The last time it happened was yesterday, and I noticed because the clock had 
stopped updating. When it hangs, I cannot launch any new programs from tint2, 
and of course the items in the taskbar are stale. One thing that might be 
helpful: I have a vague feeling it may have happened a couple of times when I 
had closed a window. I then noticed the window's taskbar item was still there 
(and then I noticed the hang).

Original comment by j...@wildlava.com on 14 Mar 2015 at 4:24

[GoogleCodeExporter]

Thanks Joe.

If you are willing to compile from source, I suggest building with "cmake 
-DENABLE_SN=OFF" as I have some suspicions that is the culprit.

Unfortunately I have not been able to reproduce the bug, but I am also using an 
older system. It could be that some newer version of X or a library we link to 
interacts badly with tint2. In the end I might have to install the latest 
Fedora or Arch and see what is going on, but this will have to wait as I am 
swamped with work at the moment.

Original comment by mrovi9...@gmail.com on 14 Mar 2015 at 4:48

[GoogleCodeExporter]

Hi,

I cannot say for sure Joe's hangs causes are the same as mine or not, but for 
me the issue didn't happen any single time after compiling with "cmake 
-DENABLE_SN=OFF".  It didn't happen for 2 weeks now and it used to happen 3 or 
4 times per week with my regular utilization.

mrovi, did you have the chance to play with always-on-top windows laying around 
while opening other windows, etc?  As I reported, it seems to always happen 
when using x48 always-on-top, although I don't have yet an exact step-by-step 
to reproduce the issue.

Will recompile it with startup notifications enabled again as a double-check.  
Last time I changed two factors at once (i.e., changed to r727 and disabled the 
SN, so this is a matter of using the scientific method to isolate the issue 
changing just one factor at a time).

Will keep updating the results here.

Thanks.

Original comment by marcelof...@gmail.com on 14 Mar 2015 at 10:47

[GoogleCodeExporter]

Well....

Compiled just a few minutes ago with startup notifications enabled and... 
TADA!!!  Already happened.

Just opened mate-terminal (mate is a spinoff of gnome) followed by opening the 
x48.  Tried to reproduce many times after that but it just happened once, but 
it is enough to reinfoce that, at least in my case, it for sure has to do with 
the "-DENABLE_SN=OFF/ON".  (please note that as reported above in the begining 
of the thread, I use a sequence of scripts involving wmctrl to put x48 on top 
and other tricks).

Thanks

Original comment by marcelof...@gmail.com on 14 Mar 2015 at 11:03

[GoogleCodeExporter]

Again...

Steps to reproduce (not 100% guaranteed):
1 - Move to an empty workspace
2 - Open Mate-Terminal (probabbly will work with gnome-terminal)
3 - Open x48 (using the launcher and scripts provided in the posts above in 
this thread).

That's it.  Happened twice in 20 or 30 attempts.

Original comment by marcelof...@gmail.com on 14 Mar 2015 at 11:07

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Update:
Step 2 is actually unnecessary.

Step 1: Move to an empty workspace
Repeate the below steps many times:
  Step 2a: Launch x48 (with the aforementioned scripts)
  Step 2b: If it didn't hang so far, close it and go back to step 2a

At some point it will open x48 and hang tint2 (must be compiled with SN enabled)

Original comment by marcelof...@gmail.com on 14 Mar 2015 at 11:34

[GoogleCodeExporter]

I did some experiments to make it simpler to reproduce (and confimed x48 is not 
the culprit).

Below is a script that uses xclock instead and equally triggers the issue.  I 
believe it can help (a .desktop launcher that runs this script must be used by 
tint2):

---- CUT HERE ----
earth:~> cat bin/xclockontop 
#!/bin/sh

wmctrl -xR xclock

if [ $? == 1 ]; then
    xclock &
    sleep 1
    wmctrl -xr xclock -b add,above
fi
---- CUT HERE ----

What does the script do:
1) It tries to bring an already running instance of xclock to the current 
workplace "wmctrl -xR xclock";
2) If it fails (no xclock already running), it lanches an instance "xclock &" 
and waits a second "sleep 1";
3) It adds the running xclock to the 'above' list (i.e., makes xclock 
always-on-top).

Steps to reproduce the issue are the same as described on #28:

Step 1: Move to an empty workspace
Repeate the below steps many times:
  Step 2a: Launch xclock (using the script above)
  Step 2b: If it didn't hang so far, close it and repeat step 2a

At some point it will lauch xclock and hang tint2.  Tested many times with 
Startup notifications disabled and it really seems to only happen with it 
enabled.

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 12:07

[GoogleCodeExporter]

s/workplace/workspace/ (in the post above... :) )

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 12:09

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Now it's quite easy to reproduce the issue and requires just xclock and a 
little persistence... :)

Just see post #29

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 12:17

[GoogleCodeExporter]

Thanks to both of you for the debugging effort. OK, I now have tint2 built with 
"-DENABLE_SN=OFF", and I'll give it a try. Note that the version I have built 
is Arch's latest official, which is version 0.11-8. Perhaps this is quite old, 
but is it still a valid test?

Original comment by j...@wildlava.com on 15 Mar 2015 at 4:26

[GoogleCodeExporter]

OK, that older version was too old (did not understand my config file), so I 
got the Arch AUR package tint2-svn (which is the one I had installed before), 
did the same "-DENABLE_SN=OFF" tweak to the package script, and installed it. 
It works - so let's see how it goes! It's running [subversion] v727.

Original comment by j...@wildlava.com on 15 Mar 2015 at 4:54

[GoogleCodeExporter]

Thanks Marcelo.

Cannot reproduce it :(

I made a recording (attached), could you please take a look and tell me if I am 
doing something wrong. If not, the problem could be that I am running Fedora on 
a virtual machine which is a bit slow. Maybe it is a timing issue and happens 
only if the system is fast.

In any case, I am planning to disable the SN code. Hopefully it also solves 
Joe's problem.

Original comment by mrovi9...@gmail.com on 15 Mar 2015 at 10:55

Attachments:

• fedora-2015-03-15T10-43-31-220064000Z.webm

[GoogleCodeExporter]

Hi mrovi,

I checked the recording.  You are doing nothing wrong, this should be exactly 
the way to reproduce it and you were persistent enough.  The only noticeable 
difference is that I use a static number of workspaces, instead of dinamically 
allocated.  So, in my case, tint2 doesn't keep changing from 2 to 3 WS every 
time xclock opens or closes, but stays with solid 8 WS all the time (not sure 
if that really makes much difference, though).

My curiosity:  Since I'm happily using tint2 with SN disabled as a workaround 
for 2 weeks now, I could only notice one tiny difference in behavior that is 
the absence of the busy mouse pointer when I click a launcher.  Is there any 
other significant advantage in enabling the SN feature?

Thank you

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 12:01

[GoogleCodeExporter]

No, that is the only thing SN does. IMHO it is more annoying than useful, but 
other users like it.

Original comment by mrovi9...@gmail.com on 15 Mar 2015 at 1:24

[GoogleCodeExporter]

Got it, thanks.

Maybe an option would be to change it from a compile time decision to a 
configuration parameter.  I understand this would be a low priority change 
compared to the existing list anyway.  In any case, due to the bug we are 
discussing, I would advocate for the default compile-time option to be disabled 
instead of enabled.

Thanks,

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 1:33

[GoogleCodeExporter]

It just hanged :D

For now the solution is to use it with SN disabled at compile time. Thank you 
for all your help in tracking this problem. 

Including stacktrace for further reference:

Thread 1 (Thread 0x7ffff7fc89c0 (LWP 26037)):
#0  __lll_lock_wait () at 
../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:132
No locals.
#1  0x00007ffff4780065 in _L_lock_858 () from 
/lib/x86_64-linux-gnu/libpthread.so.0
No symbol table info available.
#2  0x00007ffff477feba in __pthread_mutex_lock (mutex=0x62a000126218) at 
pthread_mutex_lock.c:61
        __PRETTY_FUNCTION__ = "__pthread_mutex_lock"
        type = 1204760
#3  0x00007ffff386b134 in xcb_send_request () from 
/usr/lib/x86_64-linux-gnu/libxcb.so.1
No symbol table info available.
#4  0x00007ffff386ee01 in xcb_create_window () from 
/usr/lib/x86_64-linux-gnu/libxcb.so.1
No symbol table info available.
#5  0x00007ffff49989a5 in sn_internal_broadcast_xmessage () from 
/usr/lib/x86_64-linux-gnu/libstartup-notification-1.so.0
No symbol table info available.
#6  0x00007ffff4996955 in sn_launcher_context_complete () from 
/usr/lib/x86_64-linux-gnu/libstartup-notification-1.so.0
No symbol table info available.
#7  0x000000000041f5aa in sigchld_handler (sig=17) at 
~/src/tint2-svn/src/tint.c:157
        ctx = 0x60b00007b820
        pid = 16687
#8  <signal handler called>
No symbol table info available.
#9  0x00007ffff44a320b in __libc_writev (fd=<optimized out>, 
vector=0x7fffffffdd50, count=3) at ../sysdeps/unix/sysv/linux/writev.c:51
        resultvar = 68
        result = <optimized out>
#10 0x00007ffff6f2997f in writev () from /usr/lib/x86_64-linux-gnu/libasan.so.1
No symbol table info available.
#11 0x00007ffff386a965 in ?? () from /usr/lib/x86_64-linux-gnu/libxcb.so.1
No symbol table info available.
#12 0x00007ffff386addf in ?? () from /usr/lib/x86_64-linux-gnu/libxcb.so.1
No symbol table info available.
#13 0x00007ffff386ae64 in xcb_writev () from 
/usr/lib/x86_64-linux-gnu/libxcb.so.1
No symbol table info available.
#14 0x00007ffff63f3977 in _XSend () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#15 0x00007ffff63f3eab in _XReply () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#16 0x00007ffff63ef99d in XSync () from /usr/lib/x86_64-linux-gnu/libX11.so.6
No symbol table info available.
#17 0x00007ffff531b4f1 in ?? () from /usr/lib/libImlib2.so.1
No symbol table info available.
#18 0x00007ffff52f605a in imlib_render_image_on_drawable () from 
/usr/lib/libImlib2.so.1
No symbol table info available.
#19 0x0000000000432a52 in draw_launcher_icon (obj=0x610000028540, 
c=0x7ffff5d559e0) at ~/src/tint2-svn/src/launcher/launcher.c:352
        launcherIcon = 0x610000028540
        icon_scaled = 0x60b0001c1360
#20 0x000000000044dcd6 in draw (a=0x610000028540) at 
~/src/tint2-svn/src/util/area.c:372
        cs = 0x616000152480
        c = 0x7ffff5d559e0
#21 0x000000000044c371 in refresh (a=0x610000028540) at 
~/src/tint2-svn/src/util/area.c:217
        l = 0x0
#22 0x000000000044c6a1 in refresh (a=0x61b000014750) at 
~/src/tint2-svn/src/util/area.c:227
        l = 0x6150001dbd00
#23 0x000000000044c6a1 in refresh (a=0x61b000014280) at 
~/src/tint2-svn/src/util/area.c:227
        l = 0x615000258180
#24 0x000000000044b7da in rendering (obj=0x61b000014280) at 
~/src/tint2-svn/src/util/area.c:107
        panel = 0x61b000014280
#25 0x00000000004272dc in main (argc=1, argv=0x7fffffffe818) at 
~/src/tint2-svn/src/tint.c:1089
        e = {type = 14, xany = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197}, xkey = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, root = 62, subwindow = 580263522, time = 4535485464577, x = 0, y = 522, x_root = 15118610, y_root = 0, state = 0, keycode = 0, same_screen = 1}, xbutton = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, root = 62, subwindow = 580263522, time = 4535485464577, x = 0, y = 522, x_root = 15118610, y_root = 0, state = 0, button = 0, same_screen = 1}, xmotion = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, root = 62, subwindow = 580263522, time = 4535485464577, x = 0, y = 522, x_root = 15118610, y_root = 0, state = 0, is_hint = 0 '\000', same_screen = 1}, xcrossing = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, root = 62, subwindow = 580263522, time = 4535485464577, x = 0, y = 522, x_root = 15118610, y_root = 0, mode = 0, detail = 0, same_screen = 1, focus = 0, state = 3200171710}, xfocus = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, mode = 62, detail = 0}, xexpose = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, x = 62, y = 0, width = 580263522, height = 0, count = 1}, xgraphicsexpose = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, drawable = 44040197, x = 62, y = 0, width = 580263522, height = 0, count = 1, major_code = 1056, minor_code = 0}, xnoexpose = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, drawable = 44040197, major_code = 62, minor_code = 0}, xvisibility = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, state = 62}, xcreatewindow = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, parent = 44040197, window = 62, x = 580263522, y = 0, width = 1, height = 1056, border_width = 0, override_redirect = 522}, xdestroywindow = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62}, xunmap = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, from_configure = 580263522}, xmap = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, override_redirect = 580263522}, xmaprequest = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, parent = 44040197, window = 62}, xreparent = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, parent = 580263522, x = 1, y = 1056, override_redirect = 0}, xconfigure = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, x = 580263522, y = 0, width = 1, height = 1056, border_width = 0, above = 15118610, override_redirect = 0}, xgravity = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, x = 580263522, y = 0}, xresizerequest = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, width = 62, height = 0}, xconfigurerequest = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, parent = 44040197, window = 62, x = 580263522, y = 0, width = 1, height = 1056, border_width = 0, above = 15118610, detail = 0, value_mask = 1}, xcirculate = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, event = 44040197, window = 62, place = 580263522}, xcirculaterequest = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, parent = 44040197, window = 62, place = 580263522}, xproperty = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, atom = 62, time = 580263522, state = 1}, xselectionclear = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, selection = 62, time = 580263522}, xselectionrequest = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, owner = 44040197, requestor = 62, selection = 580263522, target = 4535485464577, property = 2241972928512, time = 15118610}, xselection = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, requestor = 44040197, selection = 62, target = 580263522, property = 4535485464577, time = 2241972928512}, xcolormap = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, colormap = 62, new = 580263522, state = 0}, xclient = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, message_type = 62, format = 580263522, data = {b = "\001\000\000\000 \004\000\000\000\000\000\000\n\002\000\000\022\261", <incomplete sequence \346>, s = {1, 0, 1056, 0, 0, 0, 522, 0, -20206, 230}, l = {4535485464577, 2241972928512, 15118610, 0, 1}}}, xmapping = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, request = 62, first_keycode = 0, count = 580263522}, xerror = {type = 14, display = 0x100c1d, resourceid = 13744632836034396160, serial = 107820862593280, error_code = 5 '\005', request_code = 0 '\000', minor_code = 160 '\240'}, xkeymap = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, window = 44040197, key_vector = ">\000\000\000\000\000\000\000b\036\226\"\000\000\000\000\001\000\000\000 \004\000\000\000\000\000\000\n\002\000"}, xgeneric = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, extension = 44040197, evtype = 0}, xcookie = {type = 14, serial = 1051677, send_event = 0, display = 0x62100036d900, extension = 44040197, evtype = 0, cookie = 62, data = 0x22961e62}, pad = {-4702111237675155442, 1051677, -4702111237675155456, 107820862593280, 44040197, 62, 580263522, 4535485464577, 2241972928512, 15118610, 0, 1, -4702111234474983746 <repeats 12 times>}}
        ev = 0x7fffffffe520
        fdset = {__fds_bits = {512, 0 <repeats 15 times>}}
        x11_fd = 9
        i = 0
        panel = 0x61b000014280
        it = 0x0
        timeout = 0x66dda0
        hidden_dnd = 0
        damage_event = 104
        damage_error = 171

Original comment by mrovi9...@gmail.com on 15 Mar 2015 at 4:29

• Changed state: Accepted

[GoogleCodeExporter]

Hi mrovi, I'm happy you could reproduce the issue.

I'm back using it with SN disabled.  Please let me know if you need any 
additional info I could provide or additional experimentation that may help 
tracing it down.

Thank you.

Original comment by marcelof...@gmail.com on 15 Mar 2015 at 6:33

[GoogleCodeExporter]

Original comment by mrovi9...@gmail.com on 19 Mar 2015 at 2:42

[GoogleCodeExporter]

Should be fixed in r735, which also enables SN again. But if it still causes 
problems I will disable it for good.

Original comment by mrovi9...@gmail.com on 20 Mar 2015 at 10:50

• Changed state: Fixed

[GoogleCodeExporter]

Hi mrovi,

I tested it intensively and indeed it seems to be solved.  In any case, I'll 
continue using it with SN enabled for a week for confirmation.

I have to say that now that I learned it is possible to disable it, I prefer 
tint2 without the SN.

I'm fine with the fact that this is currently a compile option as I already 
have to use the svn version (I use launchers, which are not available in 0.11), 
but I'd advocate that it become a configuration option when 0.12 gets launched, 
when I intend to use an official package of the Fedora distribution when it 
happens in the future.

Thanks for fixing it.  Tint2 is a great software.

My best,
Marcelo

Original comment by marcelof...@gmail.com on 21 Mar 2015 at 1:58

[GoogleCodeExporter]

I agree, I don't like the "busy" mouse cursor either. I will make an option for 
it.

Thanks for all your feedback!

Original comment by mrovi9...@gmail.com on 21 Mar 2015 at 9:34

[GoogleCodeExporter]

Added config option startup_notifications = 0/1 in r737.

Original comment by mrovi9...@gmail.com on 21 Mar 2015 at 1:40

[GoogleCodeExporter]

Thanks!!  Already using it.

Original comment by marcelof...@gmail.com on 21 Mar 2015 at 2:07

[GoogleCodeExporter]

This is excellent. Thanks for being so responsive - tint2 indeed is a great 
panel; I've tried many, but no others had the same configurability, clean/crisp 
look, and lightness.

Original comment by j...@wildlava.com on 21 Mar 2015 at 2:25

[GoogleCodeExporter]

As promised, I tested SN enabled for a bit more than a week after the fix and 
confirmed the issue didn't happen any single time.  For me this is fixed.

Will now disable SN as I found it to be nicer for my taste.

Thanks

Original comment by marcelof...@gmail.com on 29 Mar 2015 at 7:24

[GoogleCodeExporter]

Thanks for the update.

Original comment by mrovi9...@gmail.com on 29 Mar 2015 at 8:24