You are not logged in.
knopwob wrote:For the moment dunst replaces newlines with spaces.
That's the odd thing, it doesn't replace a newline with a space on my system (en_GB locale), but with a curious character which comprises a V on top of a T!
I pushed the changes just an hour ago.. Does it still occur after an update?
Offline
Thanks a lot for this program, I really like it.
I just noticed a small glitch: I have bound win+space to close messages. If I press it without any message being displayed, the next message that arrives is not shown.
Also it seems that timeouts <1 second are not supported? If I send a message with notify-send -t 500 "message" it will be shown with timeout 0 which is not the desired result IMO.
My System: Dell XPS 13 | i7-7560U | 16GB RAM | 512GB SSD | FHD Screen | Arch Linux
My Workstation/Server: Supermicro X11SSZ-F | Xeon E3-1245 v6 | 64GB RAM | 1TB SSD Raid 1 + 6TB HDD ZFS Raid Z1 | Proxmox VE
My Stuff at Github: github
My Homepage: Seiichiros HP
Offline
ninian wrote:knopwob wrote:For the moment dunst replaces newlines with spaces.
That's the odd thing, it doesn't replace a newline with a space on my system (en_GB locale), but with a curious character which comprises a V on top of a T!
I pushed the changes just an hour ago.. Does it still occur after an update?
Hi, I just tried again (dunst-git-20110927-1) and it seems okay now ... thank you!
Offline
Thanks a lot for this program, I really like it.
I just noticed a small glitch: I have bound win+space to close messages. If I press it without any message being displayed, the next message that arrives is not shown.
fixed
Also it seems that timeouts <1 second are not supported? If I send a message with notify-send -t 500 "message" it will be shown with timeout 0 which is not the desired result IMO.
The clock I check the timeouts against has only a resolution of a second. With the last update, the timeout is properly rounded to the nearest second and timeouts > 0 and < 1 always result in a timeout of 1 second.
Offline
seiichiro0185 wrote:Thanks a lot for this program, I really like it.
I just noticed a small glitch: I have bound win+space to close messages. If I press it without any message being displayed, the next message that arrives is not shown.
fixed
seiichiro0185 wrote:Also it seems that timeouts <1 second are not supported? If I send a message with notify-send -t 500 "message" it will be shown with timeout 0 which is not the desired result IMO.
The clock I check the timeouts against has only a resolution of a second. With the last update, the timeout is properly rounded to the nearest second and timeouts > 0 and < 1 always result in a timeout of 1 second.
Thanks a lot for the quick fix, both work as expected now
My System: Dell XPS 13 | i7-7560U | 16GB RAM | 512GB SSD | FHD Screen | Arch Linux
My Workstation/Server: Supermicro X11SSZ-F | Xeon E3-1245 v6 | 64GB RAM | 1TB SSD Raid 1 + 6TB HDD ZFS Raid Z1 | Proxmox VE
My Stuff at Github: github
My Homepage: Seiichiros HP
Offline
First, nice app.
It would be nice to hide the messages from the append function, other than using >/dev/null. Also, seems like when dunst is running and I'm trying to exit pidgin using the ctrl+q shortcut pidgin simply freeze, but after I kill dunst it close right after so it must be related to dunst. It's very annoying to kill and start dunst every time I want to quit pidgin.
Arch64/DWM || My Dropbox referral link
Offline
First, nice app.
It would be nice to hide the messages from the append function, other than using >/dev/null. Also, seems like when dunst is running and I'm trying to exit pidgin using the ctrl+q shortcut pidgin simply freeze, but after I kill dunst it close right after so it must be related to dunst. It's very annoying to kill and start dunst every time I want to quit pidgin.
I've also noticed that when someone types something in Mumble, Mumble kind of freezes and it takes a few moments (10 or so seconds) for the notification to get through. Killing the dunst PID fixes this as well, so I'm thinking it's related somehow.
Offline
First, nice app.
thank you
It would be nice to hide the messages from the append function, other than using >/dev/null.
I've added a new option (-v) for verbose mode. Without this option, only errors are printed to stderr. With this option enabled you also get the notifications printed to stdout as it was before.
Also, seems like when dunst is running and I'm trying to exit pidgin using the ctrl+q shortcut pidgin simply freeze, but after I kill dunst it close right after so it must be related to dunst. It's very annoying to kill and start dunst every time I want to quit pidgin.
I think I found the reason. Pidgin sends a request to the notification server to close all opened notifications and waits for a reply. Since this feature isn't implemented yet, dunst didn't reply.
Please give the new changes a try and tell me, if the problem is still present. Please note, dunst still can't handle the deletion request. It just sends a reply to make pidgin happy.
I've also noticed that when someone types something in Mumble, Mumble kind of freezes and it takes a few moments (10 or so seconds) for the notification to get through. Killing the dunst PID fixes this as well, so I'm thinking it's related somehow.
I don't use mumble so I can't test it right now, but maybe it is caused by the same problem that caused pidgin to hang. So please let me know if the problem is still present after the last changes and I'll look further into it.
Offline
seems to work fine now @ pidgin.
another issue, I don't know if it's related or not to ktorrent notifications 'cause I was using ktorrent when it crash but I can't make it crash again.
some output from tty:
type unknown not a basic type
type end_struct not a basic type
also regarding commit f4508c1, there is a typo in usage().
Last edited by JokerBoy (2011-10-03 19:50:39)
Arch64/DWM || My Dropbox referral link
Offline
Awesome application, replaced my previous notification daemon perfectly.
My single issue so far is that Xmonad doesn't seem to be able to detect it as a strut, and thus avoid it ( avoidStruts: http://xmonad.org/xmonad-docs/xmonad-co … voidStruts )
Offline
I get this when I try to run dunst:
Connection Error (Failed to connect to socket /tmp/dbus-GWvD9MS5KR: Connection refused)
dbus_con == NULL
What am I missing?
EDIT: nm, works after reboot.
EDIT 2: Actually, this keeps happening. I haven't figured out the exact cause, but sometimes dunst runs fine, other times I get this error and it doesn't run at all anymore until a reboot. Any ideas?
Last edited by hellomynameisphil (2011-10-28 15:53:45)
Offline
another issue, I don't know if it's related or not to ktorrent notifications 'cause I was using ktorrent when it crash but I can't make it crash again.
some output from tty:
type unknown not a basic type type end_struct not a basic type
bump.
every time ktorrent sends a notifications dunst is crashing.
Arch64/DWM || My Dropbox referral link
Offline
I just merged a bunch of new stuff.
new stuff since I last posted here:
close all messages on right-click
Added different verbosity levels ( use -v to -vvv)
moved default values to config.h. You can define your default values here. So you don't have to override everything with commandline switches
The possibility to apply rules to notifications that match a given pattern. You can override the urgency, timeout, colors and format. An example usage for this is to give 'signed on/off' messages from pidgin different colors or set the urgency for messages from important buddies to high. This is all done within the config.h
To change default values or rules, you have to copy the config.def.h to config.h and make your changes before compiling. Dwm users should feel right home
Remember that you have to recompile dunst everytime you make changes to your config.h.
JokerBoy wrote:another issue, I don't know if it's related or not to ktorrent notifications 'cause I was using ktorrent when it crash but I can't make it crash again.
some output from tty:
type unknown not a basic type type end_struct not a basic type
bump.
every time ktorrent sends a notifications dunst is crashing.
I can't reproduce this with ktorrent, but I've added some changes that should keep dunst from exiting when it encounters these errors.
Please run dunst with '-vvv' after you've updated and paste the output.
Offline
new dbus message
extracting appname
extracting nid
extracting icon
extracting summary
extracting body
extracting hints
extracting expires
extracting hints
type unknown not a basic type
type unknown not a basic type
KTorrent randomtorrent has completed downloading.<br>Average speed: 1,05 MiB/s DL / 0 B/s UL.
{
appname: KTorrent
summary: KTorrent
body: <b>randomtorrent</b> has completed downloading.<br>Average speed: 1,05 MiB/s DL / 0 B/s UL.
icon: ktorrent
urgency: 1
timeout: 10
}
At least it doesn't crash anymore. For now.
Also, please replace <br />, <br/> and <br> with spaces..
Last edited by JokerBoy (2011-11-17 19:26:14)
Arch64/DWM || My Dropbox referral link
Offline
new dbus message extracting appname extracting nid extracting icon extracting summary extracting body extracting hints extracting expires extracting hints type unknown not a basic type type unknown not a basic type KTorrent randomtorrent has completed downloading.<br>Average speed: 1,05 MiB/s DL / 0 B/s UL. { appname: KTorrent summary: KTorrent body: <b>randomtorrent</b> has completed downloading.<br>Average speed: 1,05 MiB/s DL / 0 B/s UL. icon: ktorrent urgency: 1 timeout: 10 }
At least it doesn't crash anymore. For now.
I should rework the whole dbus/libnotify code anyway. It's a bit messy atm.
I just don't find the time and more important the motivation to do so
Also, please replace <br />, <br/> and <br> with spaces..
done
Offline
@knopwob
Many thanks for these further enhancements - am experimenting with them now.
PS: Any chance of Xft support if you're getting bored ...?!
Offline
@knopwob
Many thanks for these further enhancements - am experimenting with them now.
PS: Any chance of Xft support if you're getting bored ...?!
I don't know yet. I'll have to take a look at it, to see how much work it would be. I've planned to do some work on dunst over the weekend. After that, I should be able to tell you more.
Offline
It's new feature time again.
sort notifications by urgency. This way the most important notifications are always displayed at the top of the list.
This can be disabled in the config.h and with the -ns switch.
Show the count of hidden messages if there are more notifications in the queue than the geometry allows, This also can be disabled in the config.h
Please note that you may have to update your config.h with these new options:
int sort = True; /* sort messages by urgency */
int indicate_hidden = True; /* show count of hidden messages */
Or I've you didn't make any changes to your config.h, you can simply delete it.
Offline
Just wanted to say, excellent work, I have been looking for something like this and it's perfect, thanks!
Offline
Very nice program, thanks knopwob.
Functional and the defaults look pleasing enough already, without having started to modify the config.def.h (other than switching the font, since i don't have the terminus-medium it seems (i do have the terminus pkg from pacman installed but dunst still complained. But i don't use terminus at all anyway. I switched it to Linux Biolinum ).
One thing i noticed (in the less than a dozen minutes i used it so far): Any mouse action other than left click seems to dismiss the whole the whole set of notifications, not just right click, was that intended? (I noticed because on a notebook touchpad it can be quite easy to accidentally scroll'wheel' instead of click for example)
Anyway, unless it is going to nomz my computer, this will be my new notification daemon
Ogion
(my-dotfiles)
"People willing to trade their freedom for temporary security deserve neither and will lose both." - Benjamin Franklin
"Enlightenment is man's leaving his self-caused immaturity." - Immanuel Kant
Offline
Very nice program, thanks knopwob.
Functional and the defaults look pleasing enough already, without having started to modify the config.def.h (other than switching the font, since i don't have the terminus-medium it seems (i do have the terminus pkg from pacman installed but dunst still complained. But i don't use terminus at all anyway. I switched it to Linux Biolinum ).One thing i noticed (in the less than a dozen minutes i used it so far): Any mouse action other than left click seems to dismiss the whole the whole set of notifications, not just right click, was that intended? (I noticed because on a notebook touchpad it can be quite easy to accidentally scroll'wheel' instead of click for example)
No that's not intentional. Left-click should delete the topmost message and right-click should delete all messages. Every other mouse button should be ignored. I'll look into that when I'm at home.
Anyway, unless it is going to nomz my computer, this will be my new notification daemon
Ogion
I need a larger userbase, otherwise it's not enough fun to let dunst destroy computers
Offline
Offline
It's new feature time again.
Bravo! Encore!
... love the new features
Offline
Hm, i just recompiled (deleted the pkg and src dirs, to make sure it really got new sources), and it still closes all for any mouse action other than leftclick. (Not that i mind much)
Also noticed, it only shows three at once, and any new notifications after that get queued and displayed when old notifications are closed. I suspect indicate_hidden is meant for that? Except i can't see any (wether in the notifications or in the output of -vvv) indicator of how many messages are hidden (if this is what is meant by hidden).
I need a larger userbase, otherwise it's not enough fun to let dunst destroy computers tongue
Ich hab keine blassen Dunst wovon du redet…*naiv*
Ogion
(my-dotfiles)
"People willing to trade their freedom for temporary security deserve neither and will lose both." - Benjamin Franklin
"Enlightenment is man's leaving his self-caused immaturity." - Immanuel Kant
Offline
Hm, i just recompiled (deleted the pkg and src dirs, to make sure it really got new sources), and it still closes all for any mouse action other than leftclick. (Not that i mind much)
Also noticed, it only shows three at once, and any new notifications after that get queued and displayed when old notifications are closed. I suspect indicate_hidden is meant for that? Except i can't see any (wether in the notifications or in the output of -vvv) indicator of how many messages are hidden (if this is what is meant by hidden).
That doesn't sound right. You can double-check that you have the latest sources by going into src/dunst and src/dunst-build
and run
git log
Both should have
commit 22759b87c5b1c473296f8927f85f0f6b3b6ca2a1
Author: Sascha Kruse <knopwob@googlemail.com>
Date: Fri Nov 18 16:33:31 2011 +0100ignore middle-click
at the top. If that's the case and still doesn't show the new changes in behaviour, could you please post the output of makepkg?
knopwob wrote:I need a larger userbase, otherwise it's not enough fun to let dunst destroy computers tongue
Ich hab keine blassen Dunst wovon du redet…*naiv*
OT: what do you use to edit the messages in the forum? I'm just curious because it seems to replace with 'tongue'
Offline