Knowledge Base (Posts about fluxbox)

Gxkb is replacing fbxkb in my workflows, and GTK2 status in Devuan

bgstack15 — Sat, 07 Feb 2026 14:08:00 GMT

The story

I use a patched Fluxbox as my window manager. I use fbxkb as a system tray icon to show me quickly which keyboard layout is active in X11.

I was reading on Ludditus about how Debian is working even harder to remove gtk2 more. I remember when some of the gtk2 dev packages were first removed from Debian, because I had to switch to using gtk3 to build Palemoon web browser and FreeFileSync. But this recent conversation from Ludditus was the first I'd heard of this recent change, which was a week after the original announcement by a guy to the debian-devel mailing list who is not even subscribed to the mailing list. I seem to recall that most mailing lists are set up so you must subscribe in order to send to it. I wonder if this user who started the conversation was abusing permissions or taking advantage of a bad configuration. There's a whole thread you could read for more info on the thunderous applause leading to the death of liberty.

Anyway, trying to get back to the topic at hand, I discovered that my workflows still dependent on the "offensive" package include the following.

$ sudo apt-get remove libgtk2.0-0t64 libgtk2.0-0
The following packages will be REMOVED:
asunder fbxkb gtk-chtheme libgail-common libgail18t64 libgtk2.0-0t64 libgtk2.0-bin libgtkspell0 palemoon pidgin
0 upgraded, 0 newly installed, 10 to remove and 265 not upgraded.

I have yet to address pidgin, which I actively use. I have found GTK3 forks for asunder: GrimRipper and asunder-gtk3. I have yet to address Palemoon, but I think that one already was using gtk3 also, so I hope it was just an optional dependency on gtk2. I haven't used it in a while; I'll have to grab Steven Pusser's latest palemoon work.

Gxkb and Fbxkb

Gxkb is already very similar to fbxkb, but it already uses gtk3. I started testing it, and had to poke around for quite a while before I realized the solution for my needs is very simple! All I need to do is set up my {country_code}_{variant}.png in the correct directory.

mkdir -p ~/.local/share/gxkb/flags
cp -pi /mnt/public/Support/Programs/gxkb/us_dvorak.png ~/.local/share/gxkb/flags/

That's it. I used ImageMagick (or whatever has supplanted it, et al) convert to invert the colors so Dvorak shows up as an inverted-color US flag. That is all.

I had forked fbxkb to add some features, but I need zero customization of gxkb! This simplifies my maintenance efforts, not that I had to rebuild fbxkb ever once I built it the way I liked. Here are the customizations I had added to fbxkb, and the current resolution for gxkb:

Display menu on left click, and increment input method on middle click: Can live without the menu; I have only 2 variants for now. If that changes, I might have to implement a menu.
Enable rotating input methods on scrollwheel scrolling action: Already have this feature.
Add variant support, e.g., dvorak: Already have this feature.
Add tooltip: Already have this feature!

References and additional reading

Weblinks

There might be additional or repeated entries in-line in the post.

Files

file:///mnt/public/Support/Programs/gxkb/gxkb-keyboard-layout-indicator.md

New package for Devuan: stackrpms-acer-chromebook

bgstack15 — Fri, 17 May 2024 12:33:07 GMT

I picked up a really old Acer Chromebook, and wanted to make sure my customizations are repeatable. Also, I needed to prove I can control the power button.

So, go download package stackrpms-acer-chromebook now! And of course, the source code is right nearby.

The fluxbox templates react to the keystroke for the power button. Also, I mapped the brightness and volume keys to CTRL+SHIFT, so I can still use the regular F6-F10 functionality.

To disable the default listening behavior of the power button and lid switch, I learned about elogind-inhibit, and I wrote a small program to use it. Just call this from the provided fluxbox/startup file (in /etc/stackrpms-acer-chromebook).

files/2024/listings/disable-powerbutton-daemon (Source)

#!/bin/sh
# File: /usr/bin/disable-powerbutton-daemon
# Location: stackrpms-acer-chromebook package
# Author: bgstack15
# SPDX-License-Identifier: GPL-3.0
# Startdate: 2024-04-29-2 20:22
# Title: Power Button Suppressor Daemon
# Project: disable-powerbutton
# Purpose: Runs elogind-inhibit so we can use the power button to run logout-manager
# History:
# Usage:
#   In ~/.fluxbox/startup:   disable-powerbutton-daemon &
#   Which enables this behavior to work:
#      In ~/.fluxbox/keys:   124 :Exec logout-manager-gtk
# Reference:
#    Enlightenment desktop, man elogind-inhibit
# Related:
#    /usr/libexec/stackrpms-acer-chromebook/disable-powerbutton-loop
#    /etc/stackrpms-acer-chromebook/fluxbox.keys
#    /etc/stackrpms-acer-chromebook/fluxbox.startup
#    /etc/default/disable-powerbutton
# Improve:
# Dependencies:
#    dep-devuan: logout-manager, elogind
# Documentation:
#    To run this manually:
#    elogind-inhibit --what shutdown:handle-suspend-key:handle-hibernate-key:handle-lid-switch:idle:sleep --who "stackrpms5" --why "i hate the power button" --mode=block '/bin/forever.sh'
DAEMON_NAME="disable-powerbutton"
test -f "/etc/default/${DAEMON_NAME}" && . "/etc/default/${DAEMON_NAME}"
test -f "/etc/sysconfig/${DAEMON_NAME}" && . "/etc/sysconfig/${DAEMON_NAME}"
test -z "${DP_WHAT}" && DP_WHAT="shutdown:sleep:idle:handle-power-key:handle-suspend-key:handle-hibernate-key:handle-lid-switch"
test -z "${DP_COMMAND}" && DP_COMMAND="/usr/libexec/stackrpms-acer-chromebook/disable-powerbutton-loop"
export DP_SAFETY_FILE
elogind-inhibit --what "${DP_WHAT}"--who "disable-powerbutton-daemon" --why "Let power button call logout-manager" --mode=block "${DP_COMMAND}"

That can use /etc/default/disable-powerbutton or not, because the included defaults are good enough.

It runs an infinite loop script:

files/2024/listings/disable-powerbutton-loop (Source)

#!/bin/sh
# File: /usr/libexec/stackrpms-acer-chromebook/disable-powerbutton-loop
# Location: stackrpms-acer-chromebook package
# Author: bgstack15
# Startdate: 2024-04-29-2 19:56
# Title: Sleep forever
# Project: disable-powerbutton
# Purpose: Sleep forever until a safety file exists
# History:
# Usage:
#    called by disable-powerbutton-daemon
# Reference:
# Related:
#    /usr/bin/disable-powerbutton-daemon
# Improve:
#    Probably should be a very small inotify loop
# Dependencies:
# Documentation:
test -z "${DP_SAFETY_FILE}" && DP_SAFETY_FILE=/tmp/stop-disable-powerbutton
while test ! -f "${DP_SAFETY_FILE}" ;
do
   sleep 900
done

Fluxbox handle capslock media keys

bgstack15 — Sat, 06 Jan 2024 13:40:25 GMT

VLC lets you configure global hotkeys, so you can map your XF86AudioPrev and related keys. However, when my Capslock is on, my keyboard sends different keycodes. Here's my fix for that.

I wrote a shell script to handle the logic.

files/2024/listings/vlc-command (Source)

#!/bin/sh
# Startdate: 2024-01-01-2 14:36
# Purpose: to be a hotkey target for when CapsLock is on but I still press "Play/Pause" media button on the keyboard.
# Usage:
#    In ~/.fluxbox/keys:
#    # For capslock, media previous, pause, skip
#    171 :Exec bin/vlc-command skip
#    172 :Exec bin/vlc-command play-pause
#    173 :Exec bin/vlc-command previous

case "${1}" in
   "play-pause")
      dbus-send --type=method_call --dest=org.mpris.MediaPlayer2.vlc /org/mpris/MediaPlayer2 org.mpris.MediaPlayer2.Player.PlayPause
      ;;
   "previous"|"back"|"left")
      dbus-send --type=method_call --dest=org.mpris.MediaPlayer2.vlc /org/mpris/MediaPlayer2 org.mpris.MediaPlayer2.Player.Previous
      ;;
   "next"|"skip"|"right")
      dbus-send --type=method_call --dest=org.mpris.MediaPlayer2.vlc /org/mpris/MediaPlayer2 org.mpris.MediaPlayer2.Player.Next
      ;;
esac

And then in my ~/.fluxbox/keys file I added these statements:

# For capslock, media previous, pause, skip
171 :Exec bin/vlc-command skip
172 :Exec bin/vlc-command play-pause
173 :Exec bin/vlc-command previous

I derived these of course with xev(1).

Errata

And the reason I sometimes end up with capslock on is I always use capslock to help wake on a monitor. I thought it was an XKCD comic that pointed out that Windows admins would use Ctrl+Alt+Delete to wake up a system, and that by default that reboots a Linux system, so when I have switched over to using all-Linux years ago, I use other methods (Capslock) to see if a system is awake/locked.

Helper script: pulsemixer-notification

bgstack15 — Fri, 06 Oct 2023 13:18:32 GMT

Story

Due to a minor hardware problem where my sound card keeps jumping from aplay -l position 0 to position 1, my wine malfunctioned with audio. I set the specific card to use in the winecfg tab for Audio. After a reboot (and the alsa card position switched again), winecfg would crash on the Audio tab and I was unable to figure how to wipe the settings back to all defaults. This wine audio problem meant everything in wine that uses audio (every game) crashes. It was frustrating.

So I bit the bullet and installed pulseaudio on my Devuan GNU+Linux desktop. It was a matter of adding start-pulseaudio-x11 & to ~/.fluxbox/startup. I learned you need to keep alsa-utils and alsamixergui installed. Pulseaudio sits on top of alsa instead of replacing it.

There is a great program named pulsemixer that is a very powerful TUI and CLI tool for controlling your inputs and outputs. But I still needed alsamixer to disable the "auto-mute" when I plug in headphones, because I accidentally discovered a few years ago that if sound comes out of both speakers and headphones, it makes my life easier. I'm probably weird for that reason (and dozens of others), but I like it and I'm going to keep it that way, and it takes alsamixer to control that. (Take that, ... atheists? pulseaudio peeps anyways.)

So, pulsemixer does provide a great, simple way to toggle mute, and change volume up and down. I was only missing notifications about changing the volume.

Configuring other apps

volumeicon

I still use volumeicon but now I have configured middle-click to "open mixer," with mixer defined as pulsemixer --togle-mute. I found that while the volumeicon can still use its built-in "Mute Volume" setting, it mutes it on the alsa level, which mutes the master volume in pulseaudio, but it cannot unmute that pulseaudio master volume. So it is now always muted until I use pulsemixer.

I also had to disable volumeicon from handling the XF86AudioRaiseVolume keypresses. For that, I modified my fluxbox config.

fluxbox

For bare pulsemixer control you can just run:

122 :Exec pulsemixer --change-volume -5
123 :Exec pulsemixer --change-volume +5
121 :Exec pulsemixer --toggle-mute

Which worked well but it doesn't have a cute little popup telling me what the current volume level is. So I had to write the following script, and configure fluxbox to use it.

121 :Exec pulsemixer-notification toggle
122 :Exec pulsemixer-notification down
123 :Exec pulsemixer-notification up

The script

Explanation

This script displays a relative volume level (0/33/66/100) and muted symbol. It will always unmute if you change volume, which is the behavior I expect because that's how alsa handled volume. A neat little feature is that it uses a notification ID, so it can replace it if you're scrolling up on the volume; it will just replace the current notification instead of slowly (1.25s per step) trickling through them.

Listing

files/2023/10/listings/pulsemixer-notification (Source)

#!/bin/sh
# File: pulsemixer-notification
# Location: /usr/bin
# Author: bgstack15
# Startdate: 2023-10-04-4 09:06
# SPDX-License-Identifier: GPL-3.0
# Title: pulsemixer wrapper script that includes notifications
# Package: bgscripts
# Purpose: wrap around pulsemixer hotkey invocations to include status notifications
# History:
# Usage:
#    In ~/.fluxbox/keys:
#       121 :Exec pulsemixer-notification toggle
#       122 :Exec pulsemixer-notification down
#       123 :Exec pulsemixer-notification up
# Reference:
# Improve:
# Dependencies:
#    devuan-dep: pulsemixer, pulseaudio, notify-send
# Documentation:

# Load configuration
test -z "${devtty}" && devtty=/dev/null
test -z "${PN_TEMPFILE}" && PN_TEMPFILE=~/.cache/audio.temp
test -z "${PN_STEP}" && PN_STEP=3
test -z "${PN_EXPIRE_MS}" && PN_EXPIRE_MS=1250
test -f "${HOME}/.config/pulsemixer-notification" && . "${HOME}/.config/pulsemixer-notification"
test -f "/etc/pulsemixer-notification.conf" && . "/etc/pulsemixer-notification.conf"

# runtime
exec 1>>"${devtty}"
action="${1}"
unset icon vol relative_level message replacestring

case "${action}" in
   toggle)
      pulsemixer --toggle-mute
      case "$( pulsemixer --get-mute )" in
         0)
            action="unmute"
            ;;
         1)
            icon="audio-volume-muted"
            message="muted"
            ;;
      esac
      ;;
   up) pulsemixer --unmute ; pulsemixer --change-volume +"${PN_STEP}" ;;
   down) pulsemixer --unmute ; pulsemixer --change-volume -"${PN_STEP}" ;;
   *)
      echo "unknown: ${action}"
      ;;
esac

get_vol_level() {
   vol="$( pulsemixer --get-volume | awk '{print $1}' )"
   is_low="$( printf '%s\n' "${vol}<=33" | bc )"
   is_med="$( printf '%s\n' "(${vol}<=66)*(${vol}>33)" | bc )"
   is_hih="$( printf '%s\n' "${vol}>66" | bc )"
   is_off="$( printf '%s\n' "${vol}==0" | bc )"
   test "${is_low}" = "1" && printf '%s' "low" && return 0
   test "${is_med}" = "1" && printf '%s' "medium" && return 0
   test "${is_hih}" = "1" && printf '%s' "high" && return 0
   test "${is_off}" = "1" && printf '%s' "muted" && return 0
}

# awk: cheat and just use left-side volume of stereo volume.
vol="$( pulsemixer --get-volume | awk '{print $1}' )"
relative_level="$( get_vol_level )"

case "${action}" in
   up|down)
      icon="audio-volume-${relative_level}"
      message="${vol}"
      ;;
   unmute)
      icon="audio-volume-${relative_level}"
      message="unmuted"
      ;;
esac

test -f "${PN_TEMPFILE}" && replacestring="--replace-id=$( cat "${PN_TEMPFILE}" 2>/dev/null )"
# leave replacestring unquoted in case it is empty
notify-send ${replacestring} --icon "${icon}" --transient "${message}" --urgency low --expire-time "${PN_EXPIRE_MS}" --print-id > "${PN_TEMPFILE}"

xdg-mime command for window managers

bgstack15 — Sun, 05 Feb 2023 14:36:52 GMT

I don't use a desktop environment for most of my desktop systems, I use the fluxbox (my blog posts) window manager. It hasn't changed since before I started using it, and probably won't ever. I think it's mostly dead, but that also means it won't ever change! I did start packaging the latest (probable) upstream (github) along with a patch I wrote for myself though.

So, because I don't use a desktop environment, and I had forgotten I could use my package of mime_types_editor or anything similar, I had a small problem when I try to use the "open directory" function beside a downloaded item in my web browser. So this command solved that for me:

xdg-mime default xfe.desktop inode/directory

And also to help myautomount tray icon menu:

xdg-mime default xfe.desktop inode/mount-point

Update 2023-11-08

For dvds, particularly with myautomount:

xdg-mime default vlc.desktop x-scheme-handler/dvd

Which adds to your mimeapps.list:

[Default Applications]
x-scheme-handler/dvd=vlc.desktop

Fluxbox show "(on $HOSTNAME)" if X11 forwarded program

bgstack15 — Sun, 13 Nov 2022 13:41:49 GMT

I solved my problem from last time! I wrote a patch for Fluxbox to add the "(on $CLIENT)" message to titlebars of X11-forwarded client windows. See the source, and the debian/ for building a dpkg on my cgit instance (master branch).

Just in case that ever goes away (and somehow this blog doesn't), here's the source:

Author: bgstack15
Date: 2022-11-09
Version: fluxbox 1.4.0
Source: original
Summary: Add "(on $CLIENT)" to titlebar for forwarded windows
Message: 
Inspired by xfwm4's ability to display the remote host running an X11 window. This fails on xfe which somehow lacks the WM_CLIENT_MACHINE property.
--- a/src/Xutil.cc
+++ b/src/Xutil.cc
@@ -43,6 +43,51 @@ using std::endl;

 namespace Xutil {

+FbTk::FbString getWMClientMachine(Window window) {
+
+    if (window == None)
+        return FbTk::FbString("");
+
+    Display *display = FbTk::App::instance()->display();
+
+    XTextProperty text_prop;
+    text_prop.value = 0;
+    char **list = 0;
+    int num = 0;
+    _FB_USES_NLS;
+    FbTk::FbString name;
+
+    if (XGetWMClientMachine(display, window, &text_prop)) {
+        if (text_prop.value && text_prop.nitems > 0) {
+            if (text_prop.encoding != XA_STRING) {
+
+                text_prop.nitems = strlen((char *) text_prop.value);
+                XmbTextPropertyToTextList(display, &text_prop, &list, &num);
+
+                if (num > 0 && list != 0)
+                    name = FbTk::FbStringUtil::LocaleStrToFb(static_cast<char *>(*list));
+                else
+                    name = text_prop.value ? FbTk::FbStringUtil::XStrToFb((char *)text_prop.value) : "";
+
+                if (list)
+                    XFreeStringList(list);
+
+            } else
+                name = text_prop.value ? FbTk::FbStringUtil::XStrToFb((char *)text_prop.value) : "";
+
+            XFree(text_prop.value);
+
+        } else { // default name
+            name = _FB_XTEXT(Window, Unnamed, "Unnamed", "Default name for a window without a WM_NAME");
+        }
+    } else {
+        // default name
+        name = _FB_XTEXT(Window, Unnamed, "Unnamed", "Default name for a window without a WM_NAME");
+    }
+
+    return name;
+}
+
 FbTk::FbString getWMName(Window window) {

     if (window == None)
--- a/src/Xutil.hh
+++ b/src/Xutil.hh
@@ -28,6 +28,7 @@

 namespace Xutil {

+FbTk::FbString getWMClientMachine(Window window);
 FbTk::FbString getWMName(Window window);

 FbTk::FbString getWMClassName(Window win);
--- a/src/WinClient.cc
+++ b/src/WinClient.cc
@@ -49,6 +49,7 @@
 #else
   #include <string.h>
 #endif
+#include<unistd.h>

 using std::string;
 using std::list;
@@ -217,6 +218,10 @@ bool WinClient::getAttrib(XWindowAttribu
     return XGetWindowAttributes(display(), window(), &attr);
 }

+bool WinClient::getWMClientMachine(XTextProperty &textprop) const {
+    return XGetWMClientMachine(display(), window(), &textprop);
+}
+
 bool WinClient::getWMName(XTextProperty &textprop) const {
     return XGetWMName(display(), window(), &textprop);
 }
@@ -319,7 +324,14 @@ void WinClient::updateTitle() {
     if (m_title_override)
         return;

-    m_title.setLogical(FbTk::FbString(Xutil::getWMName(window()), 0, 512));
+    FbTk::FbString fullname = FbTk::FbString(Xutil::getWMName(window()), 0, 512);
+    FbTk::FbString clientmachine = FbTk::FbString(Xutil::getWMClientMachine(window()), 0, 512);
+    char *host = new char[512]; gethostname(host, 512);
+    FbTk::FbString hostname = FbTk::FbString(host);
+    if (clientmachine != "Unnamed" && clientmachine != "" && clientmachine != hostname) {
+        fullname += " (on " + clientmachine + ")";
+    }
+    m_title.setLogical(fullname);
     m_title_update_timer.start();
 }

@@ -328,7 +340,14 @@ void WinClient::emitTitleSig() {
 }

 void WinClient::setTitle(const FbTk::FbString &title) {
-    m_title.setLogical(title);
+    FbTk::FbString fullname = title;
+    FbTk::FbString clientmachine = FbTk::FbString(Xutil::getWMClientMachine(window()), 0, 512);
+    char *host = new char[512]; gethostname(host, 512);
+    FbTk::FbString hostname = FbTk::FbString(host);
+    if (clientmachine != "Unnamed" && clientmachine != "" && clientmachine != hostname) {
+        fullname += " (on " + clientmachine + ")";
+    }
+    m_title.setLogical(fullname);
     m_title_override = true;
     m_title_update_timer.start();
 }
--- a/src/WinClient.hh
+++ b/src/WinClient.hh
@@ -91,6 +91,7 @@ public:
     //

     bool getAttrib(XWindowAttributes &attr) const;
+    bool getWMClientMachine(XTextProperty &textprop) const;
     bool getWMName(XTextProperty &textprop) const;
     bool getWMIconName(XTextProperty &textprop) const;
     std::string getWMRole() const;

And here's the readme:

Readme for fluxbox

Overview

This is for my customized dpkg of Fluxbox which exists quite happily in vanilla form in Debian Sid.

fluxbox upstream

I modified the debian/ from snapshot.debian.org/package/fluxbox/

This set of overlay patches probably will not be maintained, so user beware! That being said, upstream is frozen so you won't get behind here either.

Reason for being in stackrpms

I wrote exactly one patch: debian/patches/add-clientmachine-if-forwarded.patch Also by using the custom package here I can incorporate all the useful patches added in what appears to be a neo-upstream on github which has neat patches like toolbar: allow labeled general action buttons and document iconified title deco which are not present in the Debian Fluxbox 1.3.5 release.

I chose Fluxbox as my daily driver desktop environment/window manager because it is stable. And it is stable, with a few minor flaws!

So this one patch I wrote adds to the titlebar of a forwarded window "(on $CLIENT)" which was inspired by Xfwm4.

Improvements

Add a configurable option for "Show (on $CLIENT)" so users can decide if they like it.
Some programs do not properly maintain WM_CLIENT_MACHINE X property, like xfe(1) so that could be misleading.
Maybe store local hostname on the Display, like Xfwm4 so we do not need to keep calling gethostname()
Find way to rearrange the iconbar window entries (issue #94 in original repo) which was also requested in one of the github forks (luebking/fluxbox#1 with comments:

If you however mean to drag around the icon, that's probably not too hard to do in src/Icon{Button,barTool}*

Alternatives

Just use the native package fluxbox for the given distro.

Github indicates a large number of forks exist of the source. Almost all of them do not provide substantive improvements.

Dependencies

Exact same as native package fluxbox.

References

Every link in this document. Official upstream link?: http://git.fluxbox.org/fluxbox.git/

Differences from upstream

Only one d/p/series and d/p/add-clientmachine-if-forwarded.patch

Call for help: Fluxbox display "(on $HOSTNAME)" for X11 forwarding

bgstack15 — Wed, 09 Nov 2022 14:06:48 GMT

I recently discovered that my X11-forwarded client windows do not display the appended string "(on $CLIENT)" like I recall seeing in the past.

Upon some research, including a Unix Stack Exchange question, I learned that it depends on the window manager! It's not a built-in of Xorg.

Xfwm4 file client.c#215 adds this bit inside function clientCreateTitleName().

Fluxbox file WinClient.cc#322 in function WinClient::updateTitle() would be where you add the logic for adding "(on $CLIENT)" string.

If anybody likes C++ development and wants to write me a custom patch and earn my dying thanks, I would greatly appreciate it! Maybe someday I'll learn how to do this part. If a patch for this ever gets written, I would have to build fluxbox myself with the patch, which I am willing to do.

References

https://salsa.debian.org/xfce-team/desktop/xfwm4 https://gitlab.xfce.org/xfce/xfwm4

http://fluxbox.org/download/

Linux: Fluxbox on Nvidia driver: Fonts are huge

bgstack15 — Tue, 06 Sep 2022 12:34:19 GMT

I installed Devuan GNU+Linux on a desktop computer with a Nvidia graphics card. Once I used the nvidia-driver dpkg and rebooted, everything worked correctly. Fluxbox had a problem where its size 8 fonts were gigantic! Investigating revealed that the Nvidia settings utility shows a DPI of 304x305. An Internet search for nvidia fluxbox large font showed a very useful solution from the FreeBSD forum

Inside the generated xorg.conf (which I placed in /etc/X11/xorg.conf.d/hostname.conf), add to the "Monitor" section:

Option         "UseEditDpi" "false"
Option         "DPI" "96 x 96"

Also, I learned that lspci -v will show you what kernel module is in use for a given PCI device. Here is my graphics card with the proprietary Nvidia driver.

01:00.0 VGA compatible controller: NVIDIA Corporation GP107 [GeForce GTX 1050 Ti] (rev a1) (prog-if 00 [VGA controller]) Subsystem: Gigabyte Technology Co., Ltd GP107 [GeForce GTX 1050 Ti] Flags: bus master, fast devsel, latency 0, IRQ 30, IOMMU group 1 Memory at f6000000 (32-bit, non-prefetchable) [size=16M] Memory at e0000000 (64-bit, prefetchable) [size=256M] Memory at f0000000 (64-bit, prefetchable) [size=32M] I/O ports at e000 [size=128] Expansion ROM at 000c0000 [virtual] [disabled] [size=128K] Capabilities: Kernel driver in use: nvidia Kernel modules: nvidia

References

Internet searches

nvidia fluxbox large font

Weblinks

https://forums.freebsd.org/threads/fluxbox-fonts-are-huge.22849/

update 2024-04-13

While you might be tempted to place a file, /etc/X11/Xresources/dpi with contents:

! bgstack15 2023-07-16-1 17:29
Xft.dpi: 96

Which tends to help most of the time, it doesn't fix the "hardware" dpi that this post helps with. See also https://old.reddit.com/r/archlinux/comments/gsq0ix/nvidia_drivers_ruins_my_resolution_and_dpi/ which indicates to place this in a file, e.g., /etc/X11/xorg.conf.d/hostname.conf:

Section "OutputClass"
    Identifier "nvidia dpi settings"
    MatchDriver "nvidia-drm"
    Option "UseEdidDpi" "False"
    Option "DPI" "96 x 96"
EndSection

Package for devuan: xdgmenumaker

bgstack15 — Tue, 03 Mar 2020 13:50:21 GMT

After a healthy discussion on the Devuan forum about how to get xdg-compliant menus for non-xdg-compliant window managers, I went and packaged up xdgmenumaker for Devuan! You can go install this utility as an easy dpkg from my OBS, which includes a few of my own wrapper tools. One is a nice menu entry for rebuilding the fluxbox menus. The nature of xdg menus is to use the system-wide /usr/share/applications directory and its contents, and also include your local ~/.local/share/applications (or other $XDG_DATA_DIRS/applications/ locations as defined), so using the system-wide xdg fluxbox menu would not be complete. Therefore, I highly recommend adding to your ~/.fluxbox/startup script:

/usr/bin/xdgmenumaker -f fluxbox -i > ~/.fluxbox/xdg-menu &

And adjust your ~/.fluxbox/menu file to include at least these entries:

[include] (.fluxbox/xdg-menu)
#[include] (/etc/xdgmenumaker/fluxbox)

I recommend the above commented line, because in case for some reason your per-user stuff is broken, The dpkg I build provides a menu entry for reloading it, so you don't have to go to the command line and run the above command.

Rearrange windows on iconbar in Fluxbox

bgstack15 — Thu, 05 Sep 2019 12:46:21 GMT

I don't know C++, so I don't know how to add a patch to Fluxbox to make it easy to drag and rearrange the windows on the iconbar (aka taskbar). But I've dabbled with some higher-level X tools, and here's my dirty solution to rearrange the windows.

#!/bin/sh
# designed to be used with a keyboard shortcut
# in ~/.fluxbox/keys that could be:
# Control Shift Mod4 X :Exec /home/username/bin/move-to-end-of-taskbar.sh
xdotool getactivewindow windowunmap windowmap

This is original research. Any arbitrary window that is active, can be hidden and then redrawn, to make it show up last in the list, so it is on the end. Do it enough, and the remaining windows will be bumped to the first in the list.

Alternatives

This functionality has been requested in fluxbox itself, but has not been implemented: Fluxbox / Feature Requests / #94 Rearrange toolbar buttons order