Some builds use cross to do cross-compilation builds for architectures otherwise not natively supported by the runner.
This builds a full Windows installer using cargo-wix. This requires some setup beforehand with some dependencies:
This builds a full Windows installer using cargo-wix and WiX Toolset 4.
+This requires some setup beforehand with some dependencies:
After that, cache is enabled, and cargo wix takes care of the rest.
After that, cargo wix takes care of the rest.
.debCurrently, .deb files are built for x86 and ARM architectures (armv7, aarch64). This is handled by cargo-deb.
bottom","text":"A customizable cross-platform graphical process/system monitor for the terminal, supporting Linux, macOS, and Windows. Inspired by other tools like gtop, gotop, and htop.
This site serves as extended documentation for bottom alongside the README.md.
Warning
Some areas of this site are still in progress and may be missing details. Feel free to suggest/contribute changes!
"},{"location":"#installation","title":"Installation","text":"Tip
It's as good idea to first check out the Support page to see if your system is officially supported!
Tip
If you're facing some issues during/after installation, check out the Troubleshooting page for some common problems and solutions.
To install bottom, refer to the installation section of the README.md, which contains a list of all the installation methods.
The command to run bottom is btm.
You can refer to the usage pages for more details on using bottom (e.g. keybinds, some features, a general overview of what each widget does).
To configure bottom (e.g. how it behaves, how it looks, etc.) refer to the command-line options page for temporary settings, or the config file page for more permanent settings.
"},{"location":"#contribution","title":"Contribution","text":"New contributors are always welcome! See the contribution section for how to contribute to bottom, whether it be filing issues, writing documentation, creating pull requests, etc.
"},{"location":"troubleshooting/","title":"Troubleshooting/Known Issues","text":""},{"location":"troubleshooting/#the-graph-points-look-brokenstrange","title":"The graph points look broken/strange","text":"It's possible that your graphs don't look great out of the box due to the reliance on braille characters to draw them. This could cause problems if your terminal's font does not support them, or your terminal is not configured properly to draw them.
An example of missing braille fonts in PowershellSome possible solutions are included below.
"},{"location":"troubleshooting/#use-dot-markers-instead","title":"Use dot markers instead","text":"One alternative is to use the --dot_marker option to render graph charts using dots instead of the braille characters, which generally seems better supported out of the box, at the expense of looking less intricate:
btm --dot_marker"},{"location":"troubleshooting/#use-a-font-that-supports-braille-fonts","title":"Use a font that supports braille fonts","text":"Another (better) alternative is to install a font that supports braille fonts, and configure your terminal emulator to use it. For example, installing something like UBraille or Iosevka and ensuring your terminal uses it should work.
"},{"location":"troubleshooting/#linuxmacosunix","title":"Linux/macOS/Unix","text":"Solutions mostly depend on what terminal emulator you are using, so unfortunately, I can't give specific instructions. Here are some possible solutions:
gnu-free-fonts if installed, as that is known to cause problems with braille markersttf-symbola or ttf-ubraille for your terminal emulator to try and automatically fall back toU+2800 to U+28FF range.symbol_map U+2800-U+28FF Symbola.For some more possible solutions:
Note: I would advise backing up your registry beforehand if you aren't sure what you are doing!
Let's say you're installing Iosevka. The steps you can take are:
Win+R and opening regedit, or just opening it from the Start Menu.In the registry editor, go to
HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Windows NT\\CurrentVersion\\Console\\TrueTypeFont\nHere, add a new String value, and set the Name to a bunch of 0's (e.g. 000 - make sure the name isn't already used), then set the Data to the font name (e.g. Iosevka).
The last entry is the new entry for Iosevka
Then, open the Command Prompt/PowerShell, and right-click on the top bar, and open \"Properties\":
From here, go to \"Font\", and set the font to your new font (so in this example, Iosevka):
This is a known issue, some sensors may require admin privileges to get sensor data.
"},{"location":"troubleshooting/#why-dont-i-see-dual-batteries-on-windows-reported-separately-eg-thinkpads","title":"Why don't I see dual batteries on Windows reported separately? (e.g. Thinkpads)","text":"This is a known issue which seems to be with how batteries are being detected on Windows.
"},{"location":"troubleshooting/#why-cant-i-see-all-my-temperature-sensors-on-wsl","title":"Why can't I see all my temperature sensors on WSL?","text":"This is a known limitation with WSL. Due to how it works, hosts may not expose their temperature sensors and therefore, temperature sensors might be missing.
"},{"location":"troubleshooting/#why-does-wsl2-not-match-task-manager","title":"Why does WSL2 not match Task Manager?","text":"This is a known limitation with WSL2. Due to how WSL2 works, the two might not match up in terms of reported data.
"},{"location":"troubleshooting/#why-cant-i-see-all-my-processesprocess-data-on-macos","title":"Why can't I see all my processes/process data on macOS?","text":"You may have to run the program with elevated privileges to work around it - for example:
sudo btm\nWarning
Please note that you should be certain that you trust any software you grant root privileges.
There are measures taken to try to maximize the amount of information obtained without elevated privileges. For example, one can modify the instructions found on the htop wiki on how to run htop without sudo for bottom. However, please understand the potential security risks before doing so!
"},{"location":"troubleshooting/#my-configuration-file-isnt-working","title":"My configuration file isn't working","text":"If your configuration files aren't working, here are a few things to try:
"},{"location":"troubleshooting/#check-the-formatting","title":"Check the formatting","text":"It may be handy to refer to the automatically generated config files or the sample configuration files. The config files also follow the TOML format.
Also make sure your config option settings are in the right location - for example, to set your temperature type, you must set it under the [flags] table:
[flags]\ntemperature_type = \"f\"\nMeanwhile, if you want to set a custom colour or styling scheme, it would be under the [styles] table - for example:
[styles.tables.headers]\ncolour = \"LightBlue\"\nbold = true\nTo help validate your configuration files, there is JSON Schema support if your IDE/editor supports it.
"},{"location":"troubleshooting/#check-the-configuration-file-location","title":"Check the configuration file location","text":"Make sure bottom is reading the right configuration file. By default, bottom looks for config files at these locations:
OS Default Config Location macOS$HOME/Library/Application Support/bottom/bottom.toml ~/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Linux ~/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Windows C:\\Users\\<USER>\\AppData\\Roaming\\bottom\\bottom.toml If you want to use a config file in another location, use the --config or -C flags along with the path to the configuration file, like so:
btm -C path_to_config\nMake sure bottom is given the correct permissions in order to collect data. Snapcraft explains how to do so, but the TL;DR is:
sudo snap connect bottom:mount-observe\nsudo snap connect bottom:hardware-observe\nsudo snap connect bottom:system-observe\nsudo snap connect bottom:process-control\nThe underlying interface we use for NVIDIA GPU information, nvml, only works with glibc and does not work with musl at the moment (see this forum post for some more details). As such, bottom may fail to get NVIDIA GPU information when using a musl-based binary until this is resolved. This applies to Linux and Windows from my understanding.
To resolve this, use glibc-based binary builds if possible (e.g. the gnu binaries/non-musl packages in releases).
If you're still having issues, feel free to open a discussion question about it, and I (or others) can try to help.
"},{"location":"configuration/command-line-options/","title":"Command-line Options","text":"The following options can be provided to bottom in the command line to change the behaviour of the program. You can also see information on these options by running btm -h, or run btm --help to display more detailed information on each option:
--autohide_time Temporarily shows the time scale in graphs. -b, --basic Hides graphs and uses a more basic look. -C, --config_location <PATH> Sets the location of the config file. -t, --default_time_value <TIME> Default time value for graphs. --default_widget_count <N> Sets the N'th selected widget type as the default. --default_widget_type <WIDGET> Sets the default widget type. Use --help for more info. --disable_click Disables mouse clicks. --disable_keys Disables keyboard shortcuts, INCLUDING the ones that stop bottom. -m, --dot_marker Uses a dot marker for graphs. -e, --expanded Expand the default widget upon starting the app. --hide_time Hides the time scale from being shown. -r, --rate <TIME> Sets how often data is refreshed. --retention <TIME> How far back data will be stored up to. --show_table_scroll_position Shows the list scroll position tracker in the widget title for table widgets. -d, --time_delta <TIME> The amount of time changed upon zooming."},{"location":"configuration/command-line-options/#process-options","title":"Process Options","text":"Option Behaviour -S, --case_sensitive Enables case sensitivity by default when searching. -u, --current_usage Calculates process CPU usage as a percentage of current usage rather than total usage. --disable_advanced_kill Hides additional stopping options on Unix-like systems. --read_only Prevents performing any actions that affect the system (e.g. stopping processes). --get_threads Also gather process thread information. -g, --group_processes Groups processes with the same name by default. No effect if --tree is set. --hide_k_threads Hide kernel threads from being shown. --process_memory_as_value Defaults to showing process memory usage by value. --process_command Shows the full command name instead of the process name by default. -R, --regex Enables regex by default while searching. -T, --tree Makes the process widget use tree mode by default. --tree_collapse Collapse the process tree by default. -n, --unnormalized_cpu Show process CPU% usage without averaging over the number of CPU cores. -W, --whole_word Enables whole-word matching by default while searching."},{"location":"configuration/command-line-options/#temperature-options","title":"Temperature Options","text":"Option Behaviour -c, --celsius Use Celsius as the temperature unit. Default. -f, --fahrenheit Use Fahrenheit as the temperature unit. -k, --kelvin Use Kelvin as the temperature unit."},{"location":"configuration/command-line-options/#cpu-options","title":"CPU Options","text":"Option Behaviour --cpu_left_legend Puts the CPU chart legend on the left side. --default_cpu_entry Sets which CPU entry type is selected by default. -a, --hide_avg_cpu Hides the average CPU usage entry."},{"location":"configuration/command-line-options/#memory-options","title":"Memory Options","text":"Option Behaviour --memory_legend <POSITION> Where to place the legend for the memory chart widget. --enable_cache_memory Enable collecting and displaying cache and buffer memory. --free_arc Subtract ARC memory that can be freed from memory usage. --short_gpu_names Use short GPU names in the memory widget."},{"location":"configuration/command-line-options/#network-options","title":"Network Options","text":"Option Behaviour --network_legend <POSITION> Where to place the legend for the network chart widget. --network_use_bytes Displays the network widget using bytes. --network_use_binary_prefix Displays the network widget with binary prefixes. --network_use_log Displays the network widget with a log scale. --show_packets Displays packet rate and average packet size info. --use_old_network_legend Uses a separate network legend. --network_start_zeroed Show total network usage from app startup rather than on boot."},{"location":"configuration/command-line-options/#battery-options","title":"Battery Options","text":"Option Behaviour --battery Shows the battery widget in non-custom layouts."},{"location":"configuration/command-line-options/#gpu-options","title":"GPU Options","text":"Option Behaviour --disable_gpu Disable collecting and displaying NVIDIA and AMD GPU information."},{"location":"configuration/command-line-options/#style-options","title":"Style Options","text":"Option Behaviour --theme <SCHEME> Use a built-in colour theme, use '--help' for info on the colours."},{"location":"configuration/command-line-options/#other-options","title":"Other Options","text":"Option Behaviour -h, --help Prints help info (for more details use --help.) -V, --version Prints version information."},{"location":"configuration/config-file/","title":"Config File","text":"For persistent configuration, and for certain configuration options, bottom supports config files.
"},{"location":"configuration/config-file/#default-config-file","title":"Default Config File","text":"If no config file argument is given, it will automatically look for a config file at these locations:
OS Default Config Location macOS$HOME/Library/Application Support/bottom/bottom.toml $HOME/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Linux $HOME/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Windows C:\\Users\\<USER>\\AppData\\Roaming\\bottom\\bottom.toml If the config file doesn't exist at the path, bottom will automatically try to create a new config file at the location with default values.
"},{"location":"configuration/config-file/#json-schema","title":"JSON Schema","text":"The configuration file also has JSON Schema support to make it easier to manage, if your IDE/editor supports it.
"},{"location":"configuration/config-file/cpu-graph/","title":"CPU Configuration","text":"If you want to change some of the default behaviour of the CPU graph widget, you can configure things under the [cpu] section.
You can configure which CPU graph is shown by default on startup by setting cpu.default. Defaults to \"all\", which shows all entries.
[cpu]\n# One of \"all\" (default), \"average\"/\"avg\"\ndefault = \"average\"\nYou can configure whether CPU usage values are shown with a decimal place by setting cpu.show_decimal. Defaults to false.
[cpu]\nshow_decimal = true\nYou can hide the average CPU entry entirely by setting cpu.hide_avg_cpu. Defaults to false.
[cpu]\nhide_avg_cpu = true\nYou can place the CPU chart legend on the left side by setting cpu.left_legend. Defaults to false.
[cpu]\nleft_legend = true\nIn basic mode, you can give the average CPU entry a dedicated row by setting cpu.basic_average_cpu_row. Defaults to false.
[cpu]\nbasic_average_cpu_row = true\nThe disk I/O graph widget is configured under [disk_io_graph].
By default, both the read and write rate lines are shown. They can be toggled off independently.
[disk_io_graph]\n# Whether to show the read rate line. Defaults to true.\nshow_read = true\n\n# Whether to show the write rate line. Defaults to true.\nshow_write = true\nBy default, legend entries are labelled by disk name (e.g. sda, nvme0n1). To use mount points instead (e.g. /, /home):
[disk_io_graph]\nlegend = \"mount\"\nValid values are \"disk\" (default) and \"mount\".
The y-axis can be switched to a logarithmic scale:
[disk_io_graph]\nuse_log = true\nThe location of the legend can be set with legend_position. Valid values are none, top-left, top, top-right, left, right, bottom-left, bottom, and bottom-right. Defaults to top-right.
[disk_io_graph]\nlegend_position = \"top-right\"\nBy default, only mounted devices are shown. To also show unmounted devices on Linux, enable include_unmounted:
[disk_io_graph]\ninclude_unmounted = true\nYou can filter out what entries to show by configuring [disk_io_graph.name_filter] and [disk_io_graph.mount_filter] to filter by name and mount point respectively. In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
If we wanted to ignoring any entry with a name that matches /dev/sda:
[disk_io_graph.name_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"/dev/sda\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nYou can configure which columns are shown by the disk table widget by setting the columns setting:
[disk]\n# Pick which columns you want to use in any order.\ncolumns = [\"Disk\", \"Mount\", \"Used\", \"Free\", \"Total\", \"Used%\", \"R/s\", \"W/s\"]\nYou can customize the default sort column (by default, it sorts by disk name in ascending order). For example, to sort by the read rate:
[disk]\ndefault_sort = \"R/s\"\nYou can use any valid column name here (e.g. \"Disk\", \"Mount\", etc.). Note that if you put a column name that is not actually used, the default sort will just be the first column shown.
You can also set the sort order by changing disk.sort_order with \"Ascending\" or \"Descending\". Defaults to \"Ascending\":
[disk]\nsort_order = \"Ascending\"\nBy default, only mounted devices are shown. To also show unmounted devices on Linux, enable include_unmounted:
[disk]\ninclude_unmounted = true\nUnmounted devices show I/O activity but have no used/free space (those require a live mount), so those columns appear as N/A. Pseudo-devices (loop*, ram*, zram*) are skipped, and you can hide other noisy devices with the name filter.
You can filter out what entries to show by configuring [disk.name_filter] and [disk.mount_filter] to filter by name and mount point respectively. In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, consider a disk widget showing these entries:
If we wanted to ignoring any entry with a name that matches /dev/sda:
[disk.name_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"/dev/sda\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nThis would give us:
We can also combine both the name filter and mount filter. For example:
[disk.name_filter]\nis_list_ignored = false\nlist = [\"/dev/sda\"]\nregex = true\ncase_sensitive = false\nwhole_word = false\n\n[disk.mount_filter]\nis_list_ignored = true\nlist = [\"/mnt/.*\", \"/\"]\nregex = true\ncase_sensitive = false\nwhole_word = true\nThis gives us:
"},{"location":"configuration/config-file/flags/","title":"Flags","text":"Warning
This section is in progress, and is just copied from the old documentation.
You can configure flags by putting them in [flags] table. Example:
[flags]\nhide_avg_cpu = true\nMost of the command line flags have config file equivalents to avoid having to type them out each time:
Field Type Functionalityhide_avg_cpu Boolean Deprecated - use cpu.hide_avg_cpu. Hides the average CPU usage. dot_marker Boolean Uses a dot marker for graphs. cpu_left_legend Boolean Deprecated - use cpu.left_legend. Puts the CPU chart legend to the left side. current_usage Boolean Deprecated - use processes.current_usage. Sets process CPU% to be based on current CPU%. group_processes Boolean Deprecated - use processes.default_grouped. Groups processes with the same name by default. case_sensitive Boolean Deprecated - use processes.case_sensitive. Enables case sensitivity by default. whole_word Boolean Deprecated - use processes.whole_word. Enables whole-word matching by default. regex Boolean Deprecated - use processes.regex. Enables regex by default. basic Boolean Hides graphs and uses a more basic look. use_old_network_legend Boolean DEPRECATED - uses the older network legend. battery Boolean Shows the battery widget. rate Unsigned Int (represents milliseconds) or String (represents human time) Sets a refresh rate in ms. default_time_value Unsigned Int (represents milliseconds) or String (represents human time) Default time value for graphs in ms. time_delta Unsigned Int (represents milliseconds) or String (represents human time) The amount in ms changed upon zooming. hide_time Boolean Hides the time scale. temperature_type String (one of [\"k\", \"f\", \"c\", \"kelvin\", \"fahrenheit\", \"celsius\"]) Sets the temperature unit type. default_widget_type String (one of [\"cpu\", \"proc\", \"net\", \"temp\", \"mem\", \"disk\"], same as layout options) Sets the default widget type, use --help for more info. default_widget_count Unsigned Int (represents which default_widget_type) Sets the n'th selected widget type as the default. disable_click Boolean Disables mouse clicks. enable_cache_memory Boolean Deprecated - use memory.cache_memory. Enable cache and buffer memory stats (not available on Windows). process_memory_as_value Boolean Deprecated - use processes.default_memory_value. Defaults to showing process memory usage by value. tree Boolean Deprecated - use processes.default_tree. Defaults to showing the process widget in tree mode. show_table_scroll_position Boolean Shows the scroll position tracker in table widgets. show_table_scroll_bar Boolean Shows a scroll bar on the right edge of table widgets. process_command Boolean Deprecated - use processes.process_command. Show processes as their commands by default. disable_advanced_kill Boolean Deprecated - use processes.disable_advanced_kill. Disable being able to send signals to processes on supported Unix-like systems. Only available on Linux, macOS, and FreeBSD. read_only Boolean Prevents performing any actions that affect the system (e.g. stopping processes). network_use_binary_prefix Boolean Deprecated - use network_graph.use_binary_prefix. Displays the network widget with binary prefixes. network_use_bytes Boolean Deprecated - use network_graph.use_bytes. Displays the network widget using bytes. network_use_log Boolean Deprecated - use network_graph.use_log. Displays the network widget with a log scale. disable_gpu Boolean Disable NVIDIA and AMD GPU data collection. retention String (human readable time, such as \"10m\", \"1h\", etc.) How much data is stored at once in terms of time. unnormalized_cpu Boolean Deprecated - use processes.unnormalized_cpu. Show process CPU% without normalizing over the number of cores. expanded Boolean Expand the default widget upon starting the app. memory_legend String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Deprecated - use memory_graph.legend_position or memory.legend_position; where to place the legend for the memory widget. network_legend String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Deprecated - use network_graph.legend_position or network.legend_position; where to place the legend for the network widget. average_cpu_row Boolean Deprecated - use cpu.basic_average_cpu_row. Moves the average CPU usage entry to its own row when using basic mode. tree_collapse Boolean Deprecated - use processes.tree_collapse. Collapse process tree by default. autohide_time Boolean Temporarily shows the time scale in graphs. table_gap String (one of [\"none\", \"space\", \"line\"]) Controls the gap between table headers and data rows. Defaults to \"space\". disable_keys Boolean Disables keyboard shortcuts, including the ones that stop bottom. no_write Boolean Disables writing to the config file. hide_k_threads Boolean Deprecated - use processes.hide_k_threads. Hide kernel threads from being shown. free_arc Boolean Deprecated - use memory.free_arc. Subtract ARC memory that can be freed from memory usage."},{"location":"configuration/config-file/layout/","title":"Layout","text":"Warning
This section is in progress, and is just copied from the old documentation.
bottom supports customizable layouts via the config file. Currently, layouts are controlled by using TOML objects and arrays.
For example, given the sample layout:
[[row]]\n [[row.child]]\n type=\"cpu\"\n[[row]]\n ratio=2\n [[row.child]]\n ratio=4\n type=\"mem\"\n [[row.child]]\n ratio=3\n [[row.child.child]]\n type=\"temp\"\n [[row.child.child]]\n type=\"disk\"\nThis would give a layout that has two rows, with a 1:2 ratio. The first row has only the CPU widget. The second row is split into two columns with a 4:3 ratio. The first column contains the memory widget. The second column is split into two rows with a 1:1 ratio. The first is the temperature widget, the second is the disk widget.
This is what the layout would look like when run:
Each [[row]] represents a row in the layout. A row can have any number of child values. Each [[row.child]] represents either a column or a widget. A column can have any number of child values as well. Each [[row.child.child]] represents a widget. A widget is represented by having a type field set to a string.
The following type values are supported:
\"cpu\" CPU chart and legend \"mem\", \"memory\" Memory chart \"net\", \"network\" Network chart and legend \"proc\", \"process\", \"processes\" Process table and search \"temp\", \"temperature\" Temperature table \"temp_graph\", \"temperature_graph\" Temperature graph \"disk\" Disk table \"disk_io_graph\" Disk I/O graph \"empty\" An empty space \"batt\", \"battery\" Battery statistics Each component of the layout accepts a ratio value. If this is not set, it defaults to 1.
Furthermore, you can have duplicate widgets.
For an example, look at the default config, which contains the default layout.
"},{"location":"configuration/config-file/memory-graph/","title":"Memory Graph Configuration","text":"If you want to change some of the default behaviour of the memory graph widget, you can configure things under the [memory_graph] (or [memory]) section.
You can change where the legend for the graph is placed within the widget itself (or hidden with \"none\"). The default is \"top-right\".
[memory_graph]\n# One of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]\nlegend_position = \"top-left\"\nOn Linux, you can change whether the memory used by cache/slabs is collected and shown. By default, it is false.
[memory_graph]\ncache_memory = true\nIf GPU support is enabled, you can use short_gpu_names to display simplified GPU names in the memory graph widget instead of the full GPU names. If enabled, then if there's only one GPU, it'll just list the entry as GPU; if there are multiple GPUs, it'll use entries like GPU0, GPU1, etc. By default, this is disabled.
[memory_graph]\nshort_gpu_names = true\nIf ZFS is detected (note that the zfs feature must be enabled if built manually), you can enable memory_graph.free_arc to not count ARC memory in the RAM usage calculations. Disabled by default.
[memory_graph]\nfree_arc = true\nIf you want to change some of the default behaviour of the network graph widget, you can configure some things in the config file.
Field Type Functionalityshow_packets Boolean Displays packet rate and average packet size info. legend_position String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Where to place the legend for the network widget. use_bytes Boolean Displays the network widget using bytes. Defaults to bits. use_log Boolean Displays the network widget with a log scale. Defaults to a non-log scale. use_binary_prefix Boolean Displays the network widget with a binary prefix (e.g. kibibits) rather than a decimal prefix (e.g. kilobits). Defaults to decimal prefixes. start_zeroed Boolean Zeroes out the total network usage (\"All\") counters such that it shows the total usage since the app is started, rather than the total usage since boot."},{"location":"configuration/config-file/network-graph/#filtering-entries","title":"Filtering Entries","text":"You can filter out what entries to show by configuring [network_graph.interface_filter] . In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, here we are ignoring any entry with a name that matches the regex eth0.*, or specifically virbr0.
[network_graph.interface_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"virbr0\", \"eth0slab\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nIf you want to change some of the default behaviour of the processes widget, you can configure some things in the config file.
Field Type Functionalityget_threads Boolean Gather process thread information. hide_k_threads Boolean Hide kernel threads from being shown. Linux only. tree_collapse Boolean Collapse the process tree by default when tree mode is set. process_command Boolean Shows the full command name instead of the process name by default. disable_advanced_kill Boolean Disable the advanced kill dialog and just show the basic one with no options. Linux, macOS, FreeBSD only. default_memory_value Boolean Defaults to showing process memory usage by value. default_grouped Boolean Groups processes with the same name by default. No effect if --tree is set. regex Boolean Enables regex by default while searching. case_sensitive Boolean Enables case sensitivity by default when searching. whole_word Boolean Enables whole-word matching by default while searching. default_tree Boolean Makes the process widget use tree mode by default. current_usage Boolean Calculates process CPU usage as a percentage of current usage rather than total usage. unnormalized_cpu Boolean Show process CPU% usage without averaging over the number of CPU cores."},{"location":"configuration/config-file/processes/#columns","title":"Columns","text":"You can configure which columns are shown by the process widget by setting the columns setting:
[processes]\n# Pick which columns you want to use in any order.\ncolumns = [\"cpu%\", \"mem%\", \"pid\", \"name\", \"read\", \"write\", \"tread\", \"twrite\", \"state\", \"user\", \"time\", \"gmem%\", \"gpu%\"]\nBy default, the process widget starts sorted by CPU usage in descending order. You can change the column it sorts by at startup:
[processes]\ndefault_sort = \"mem\"\nAny of the column names accepted by columns work here (e.g. \"cpu%\", \"mem\", \"pid\", \"name\", \"read\"). If the column you pick is not actually shown by the widget, the built-in default is used instead.
The same setting is also exposed as a CLI flag, which takes precedence over the config file:
btm --process_default_sort mem\nYou can also set the default order by changing processes.sort_order with \"Ascending\" or \"Descending\". If not set, then by default, it will use a sort order that is reasonable (e.g. descending for CPU/memory usage, ascending for process names).
[processes]\nsort_order = \"Ascending\"\nVarious parts of the bottom can be styled, using either built-in themes or custom theming.
"},{"location":"configuration/config-file/styling/#precedence","title":"Precedence","text":"As there are a few ways styles can be applied to bottom, the order of which styles are prioritized are, in order of highest precedence to lowest precedence:
btm --theme gruvbox)If nothing is set, it will fall back to the default theme.
"},{"location":"configuration/config-file/styling/#built-in-styles","title":"Built-in styles","text":"bottom has a few built-in themes:
These themes all also have light variants to support terminals using lighter colours.
To set the theme from the command line:
btm --theme gruvbox\nTo set the theme using the config file:
[styles]\ntheme = \"gruvbox\"\nbottom's components can also be individually styled by the user to control the colour of the text style.
"},{"location":"configuration/config-file/styling/#colours","title":"Colours","text":"Info
Note that anywhere \"colour\" is used, it can be substituted for \"color\":
# This is okay\n[styles.widgets.widget_title]\ncolour = \"black\"\nbg_colour = \"white\"\n\n[styles.cpu]\nall_entry_colour = \"green\"\n\n# This is also okay\n[styles.widgets.widget_title]\ncolor = \"black\"\nbg_color = \"white\"\n\n[styles.cpu]\nall_entry_color = \"green\"\nYou can configure the colours for components with strings that are either hex colours (e.g. \"#ffffff\"), RGB colours (e.g. \"255, 255, 255\"), or named colours. Named colours are one of the following strings:
\"Black\"\"Red\"\"Green\"\"Yellow\"\"Blue\"\"Magenta\"\"Cyan\"\"Gray\"\"DarkGray\"\"LightRed\"\"LightGreen\"\"LightYellow\"\"LightBlue\"\"LightMagenta\"\"LightCyan\"\"White\"Text can generally be styled using the following TOML table:
[field]\n# The foreground colour of text.\ncolour = \"black\"\n\n# The background colour of text.\nbg_colour = \"blue\"\n\n# Whether to make the text bold.\nbold = false\n\n# Inline table version\nfield = { colour = \"black\", bg_colour = \"blue\", bold = false }\nAll fields are optional; by default if bg_colour is not set then there will be no background colour.
If you just want to style text by setting the foreground colour, for brevity, then you can also just set the field to be the colour itself. For example:
[styles.widgets]\nselected_text = \"#fff\"\nThese can be set under [styles.cpu]:
all_entry_colour The colour of the \"All\" CPU label all_entry_colour = \"Red\" avg_entry_colour The colour of the average CPU label and graph line avg_entry_colour = \"255, 0, 255\" cpu_core_colours Colour of each CPU threads' label and graph line. Read in order. cpu_core_colours = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#temperature-graph","title":"Temperature Graph","text":"These can be set under [styles.temp_graph]:
temp_graph_colour_styles Colour of each temperature sensor's graph line. Read in order. temp_graph_colour_styles = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#memory","title":"Memory","text":"These can be set under [styles.memory]:
ram_colour The colour of the RAM label and graph line ram_colour = \"Red\" cache_colour The colour of the cache label and graph line. Does not do anything on Windows. cache_colour = \"#ffffff\" swap_colour The colour of the swap label and graph line swap_colour = \"255, 0, 255\" arc_colour The colour of the ARC label and graph line arc_colour = \"Blue\" gpu_colours Colour of each GPU's memory label and graph line. Read in order. gpu_colours = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#network","title":"Network","text":"These can be set under [styles.network]:
rx_colour The colour of the RX (download) label and graph line rx_colour = \"Red\" tx_colour The colour of the TX (upload) label and graph line tx_colour = \"#ffffff\" rx_total_colour The colour of the total RX (download) label in basic mode rx_total_colour = \"0, 0, 0\" tx_total_colour The colour of the total TX (upload) label in basic mode tx_total_colour = \"#000\""},{"location":"configuration/config-file/styling/#battery","title":"Battery","text":"These can be set under [styles.battery]:
high_battery_colour The colour of the battery widget bar when the battery is over 50% high_battery_colour = \"Red\" medium_battery_colour The colour of the battery widget bar when the battery between 10% to 50% medium_battery_colour = \"#ffffff\" low_battery_colour The colour of the battery widget bar when the battery is under 10% low_battery_colour = \"0, 0, 0\""},{"location":"configuration/config-file/styling/#tables","title":"Tables","text":"These can be set under [styles.tables]:
headers Text styling for table headers headers = { colour = \"red\", bg_colour = \"black\", bold = true }"},{"location":"configuration/config-file/styling/#graphs","title":"Graphs","text":"These can be set under [styles.graphs]:
graph_colour The general colour of the parts of the graph graph_colour = \"white\" legend_text Text styling for graph's legend text legend_text = { colour = \"black\", bg_colour = \"blue\", bold = true }"},{"location":"configuration/config-file/styling/#general-widget-settings","title":"General widget settings","text":"These can be set under [styles.widgets]:
border_colour The colour of the widgets' borders border_colour = \"white\" selected_border_colour The colour of a widget's borders when the widget is selected selected_border_colour = \"white\" widget_title Text styling for a widget's title widget_title = { colour = \"black\", bg_colour = \"blue\", bold = true } bg_colour The background colour of the widgets bg_colour = \"black\" text Text styling for text in general text = { colour = \"black\", bg_colour = \"blue\", bold = true } selected_text Text styling for text when representing something that is selected selected_text = { colour = \"black\", bg_colour = \"blue\", bold = true } disabled_text Text styling for text when representing something that is disabled disabled_text = { colour = \"black\", bg_colour = \"blue\", bold = true } thread_text Text styling for text when representing process thread, only usable on Linux at the moment thread_text = { colour = \"green\", bg_colour = \"blue\", bold = true }"},{"location":"configuration/config-file/temperature-graph/","title":"Temperature Graph Configuration","text":"The temperature graph widget is configured under [temperature_graph].
The location of the legend can be set with legend_position. Valid values are none, top-left, top, top-right, left, right, bottom-left, bottom, and bottom-right. Defaults to top-right.
[temperature_graph]\nlegend_position = \"top-right\"\nBy default, the y-axis is bounded at 100\u00b0C (or the equivalent in the configured temperature unit) and grows automatically if a reading exceeds that. An explicit upper bound can be set with max_temp (uses the same unit as the configured temperature unit). Sensor readings above this value will be drawn off the chart.
[temperature_graph]\nmax_temp = 90.0\nYou can filter which sensors to plot by configuring [temperature_graph.sensor_filter]. This works the same way as the temperature table's filter.
For example, here we are ignoring any sensor that has \"cpu\" or \"wifi\" in its name:
[temperature_graph.sensor_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"cpu\", \"wifi\"]\n\n# Whether to use regex. Defaults to false.\nregex = false\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nThe temperature table widget is configured under [temperature].
You can customize the default sort column (by default, it sorts by temperature sensor name in ascending order). For example, to sort by temperature:
[temperature]\ndefault_sort = \"Temp\"\nYou can also set the sort order by changing temperature.sort_order with \"Ascending\" or \"Descending\". Defaults to \"Ascending\":
[temperature]\nsort_order = \"Ascending\"\nYou can filter out what entries to show by configuring [temperature.sensor_filter]. In particular you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, here we are ignoring any sensor that has \"cpu\" or \"wifi\" in it.
[temperature.sensor_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"cpu\", \"wifi\"]\n\n# Whether to use regex. Defaults to false.\nregex = false\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nREADME.md, changelog, etc.)There are a few areas where documentation changes are often needed:
README.mdCHANGELOG.mdFork the repository to make changes in.
Where you're adding documentation will probably affect what you need to do:
README.md or CHANGELOG.md For changes to README.md and CHANGELOG.md, just follow the formatting provided and use any editor.
Generally, changes to CHANGELOG.md will be handled by a maintainer, and the contents of the file should follow the Keep a Changelog format, as well as link to the relevant PR or issues.
For changes to the help menu, try to refer to the existing code within src/constants.rs on how the help menu is generated.
For changes to the extended documentation, you'll probably want at least Python 3.11 (older and newer versions should be fine), MkDocs, Material for MkDocs, mdx_truly_sane_lists, and optionally Mike installed. These can help with validating your changes locally.
You can do so through pip or your system's package managers. If you use pip, you can use venv to cleanly install the documentation dependencies:
# Change directories to the documentation.\ncd docs/\n\n # Create venv, install the dependencies, and serve the page.\n./serve.sh\nThis will serve a local version of the docs that you can open on your browser. It will update as you make changes.
Once you have your documentation changes done, submit it as a pull request. For more information regarding that, refer to Issues, Pull Requests, and Discussions.
Documentation should never be generated by AI. The point of documentation in this project is to communicate things between humans. Follow the AI policy; changes that do not follow it may be closed or hidden.
"},{"location":"contribution/issues-and-pull-requests/","title":"Issues, Pull Requests, and Discussions","text":""},{"location":"contribution/issues-and-pull-requests/#discussions","title":"Discussions","text":"Discussions are open in the repo. As for the difference between discussions and issues:
When filing a bug report, please use the bug report template and fill in as much as you can. It is incredibly difficult for a maintainer to fix a bug when it cannot be reproduced, and giving as much detail as possible generally helps to make it easier to reproduce the problem!
"},{"location":"contribution/issues-and-pull-requests/#feature-requests","title":"Feature requests","text":"Please use the feature request template and fill it out. Remember to give details about what the feature is along with why you think this suggestion will be useful.
Also, please check whether an existing issue has covered your specific feature request!
"},{"location":"contribution/issues-and-pull-requests/#pull-requests","title":"Pull requests","text":"The expected workflow for a pull request is:
main branch. Please fill out the pull request template and checklist.Some other things related to making changes or pull requests:
Note that any contribution, whether it be opening an issue, PRs, or commenting, should follow the AI policy. Contributions that do not follow this may be closed/hidden.
"},{"location":"contribution/packaging-and-distribution/","title":"Packaging and Distribution","text":"Package maintainers are always welcome and appreciated! Here's some info on how one can help with package distribution and bottom.
"},{"location":"contribution/packaging-and-distribution/#pre-built-binaries","title":"Pre-built binaries","text":"The latest stable release can be found here, where you can find pre-built binaries in either a tar.gz or zip format. Binaries here also include automatically generated shell completion files for zsh, bash, fish, and Powershell, which you may want to also install during the packaging process.
You can also find a nightly build in the releases page, built every day at 00:00 UTC off of the main branch.
In both cases, we use GitHub Actions to create our release binaries. build_releases.yml contains the GitHub Action workflow for this.
If you want to manually build bottom rather than distributing a pre-built binary, you'll need the most recent version of stable Rust, which you can get with:
rustup update stable\nYou'll then want to build with:
cargo build --release --locked\nbottom uses a build.rs script to automatically generate a manpage and shell completions for the following shells:
If you want to generate manpages and/or completion files, set the BTM_GENERATE env var to a non-empty value. For example, run something like this:
BTM_GENERATE=true cargo build --release --locked\nThis will automatically generate completion and manpage files in target/tmp/bottom/. If you wish to regenerate the files, modify/delete either these files or set BTM_GENERATE to some other non-empty value to retrigger the build script.
You may override the default directories used to generate both completion and manpage files by specifying the COMPLETION_DIR and MANPAGE_DIR environment variables respectively.
For more information, you may want to look at either the build.rs file or the binary build CI workflow.
Once you've finished your installation source, if you want to mention it in the main bottom repo, fork the repo and add the installation method and any details to the README.md file under the Installation section, as well as a corresponding table of contents entry. Once that's done, open a pull request - these will usually be approved of very quickly.
You can find more info on the contribution process here.
"},{"location":"contribution/development/build_process/","title":"Build Process","text":"Warning
This section is currently somewhat WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/build_process/#overview","title":"Overview","text":"bottom manages its own binary builds for nightly and stable release purposes. The core build workflow is handled by build_releases.yml, called by a wrapper workflow for nightly and stable releases. Builds take place via GitHub Actions.
The main things built are:
.deb package for Debian and its derivativesThis documentation gives a high-level overview of the build process for each part. For the most up-to-date and detailed reference, definitely refer back to the build_releases.yml file.
Binaries are built currently for various targets. Note that not all of these are officially supported. The following general steps are performed:
Build a release build with:
--features deploy, which enables only crates needed for release builds.--locked to lock the dependency versions.The following env variables set:
BTM_GENERATE: trueCOMPLETION_DIR: \"target/tmp/bottom/completion/\"MANPAGE_DIR: \"target/tmp/bottom/manpage/\"These generate the manpages and shell completions (see Packaging for some more information).
Some builds use cross to do cross-compilation builds for architectures otherwise not natively supported by the runner.
This builds a full Windows installer using cargo-wix. This requires some setup beforehand with some dependencies:
After that, cache is enabled, and cargo wix takes care of the rest.
.deb","text":"Currently, .deb files are built for x86 and ARM architectures (armv7, aarch64). This is handled by cargo-deb.
cargo-deb..deb file.There are additional checks via dpkg to ensure the architecture is correctly set.
Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/deploy_process/#overview","title":"Overview","text":"bottom currently has two main deploy processes to worry about:
Stable: a stable deployment, triggered manually or upon creation of a valid tag. This is a GitHub action that builds binary/installer files and uploads them to a new GitHub release.
Furthermore, this workflow does not handle the following deployments, which must be manually handled:
This is, for the most part, automatic, though it can also be used as a way of testing build workflow changes and seeing if binaries can be successfully built at all against all the targets we want to build for.
If one does not want to actually update the nightly release, and just want to test the general builds and workflow, one can run the workflow manually on a branch of choice with \"mock\" set as the parameter. Changing it to anything else will trigger a non-mock run.
"},{"location":"contribution/development/deploy_process/#stable","title":"Stable","text":"This can be manually triggered, though the general use-case is setting a tag of the form x.y.z (after checking everything is good, of course). For example:
git tag 0.6.9 && git push origin 0.6.9\nThis will automatically trigger the deployment workflow, and create a draft release with the files uploaded. One still needs to fill in the details and release it.
Furthermore, there are some deployments that are handled by maintainers of bottom that this workflow does not automatically finish. These must be manually handled.
"},{"location":"contribution/development/deploy_process/#cratesio","title":"crates.io","text":"Validate everything builds properly and works (you should have done this before releasing though). If good, then deploying on crates.io is as simple as:
cargo publish\nUpon releasing on GitHub, choco-bottom will automatically be updated with a new PR with the correct deployment files for Chocolatey. Check the PR, merge it if it is correct, then pull locally and deploy following the instructions in the README. Make sure to test installation and running at least once before deploying!
If done correctly, there should be a new build on Chocolatey, which will take some time to validate.
"},{"location":"contribution/development/deploy_process/#winget","title":"winget","text":"Sometimes, people will kindly do it, but it can also be manually triggered using winget-bottom.
Note this requires regenerating some secrets.
"},{"location":"contribution/development/dev_env/","title":"Development Environment","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/logging/","title":"Logging","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/testing/","title":"Testing","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"support/official/","title":"Official support","text":"bottom officially supports the following operating systems and corresponding architectures:
x86_64, aarch64)x86_64, i686, aarch64)x86_64, i686)These platforms are tested to work (with caveats, see below) and issues on these platforms will be fixed if possible.
Furthermore, binaries are expected to be built and tested using the most recent version of stable Rust - if you are manually building bottom from the repo/source, then please try that as well.
If you encounter any issues, see the troubleshooting page or open an issue/discussion on GitHub.
"},{"location":"support/unofficial/","title":"Unofficial support","text":"Systems and architectures that aren't officially supported may still work, but there are no guarantees on how much will work. For example, it might only compile, or it might run with bugs/broken features. Furthermore, while it will depend on the problem at the end of the day, issues on unsupported platforms are likely to go unfixed.
Unofficially supported platforms known to compile/work:
If you encounter any issues, see the troubleshooting page or open an issue/discussion on GitHub - though as mentioned, unsupported platforms may go unfixed.
"},{"location":"usage/autocomplete/","title":"Auto-Complete","text":"The release binaries in the releases page are packaged with shell auto-completion files for Bash, Zsh, fish, Powershell, Elvish, Fig, and Nushell. To install them:
btm.bash to $XDG_CONFIG_HOME/bash_completion or /etc/bash_completion.d/._btm to one of your $fpath directories.btm.fish to $HOME/.config/fish/completions/._btm.ps1 to your PowerShell profile.btm.elv.btm.ts.btm.nu.The individual auto-completion files are also included in the stable/nightly releases as completion.tar.gz if needed.
Basic mode is a special layout that removes all of the graphs and provides an interface that resembles (a very stripped-down version of) htop.
Basic mode can be enabled either through a command line flag:
btm -b\n\n# or\n\nbtm --basic\nor through the config:
[flags]\nbasic = true\nIn this mode, widgets that use tables (temperatures, processes, disks, and batteries) are only shown one at a time. One can switch between these widgets either by clicking the arrow buttons or by using the general widget selection shortcuts (for example, Ctrl+Left or H) to switch which widget is shown.
Also note that in this mode, widget expansion and custom layouts are disabled.
"},{"location":"usage/basic-mode/#key-bindings","title":"Key bindings","text":"Basic mode follows the same key bindings as normal, barring widget expansion being disabled, and that the % key while selecting the memory widget toggles between total usage and percentage.
"},{"location":"usage/general-usage/","title":"General Usage","text":"You can run bottom with:
btm\nFor help regarding the command-line options, use:
# For a simple overview of flags\nbtm -h\n\n# For more details\nbtm --help\nYou can also see keybinds and basic usage details in bottom by pressing ?, which will open a help menu.
"},{"location":"usage/general-usage/#features","title":"Features","text":""},{"location":"usage/general-usage/#expansion","title":"Expansion","text":"By default, bottom is somewhat like a dashboard - a bunch of different widgets, all showing different things, and they all cram together to fit into one terminal.
If you instead just want to see one widget - maybe you want to look at a graph in more detail, for example - you can \"expand\" the currently selected widget using the e key, which will hide all other widgets and make that widget take up all available terminal space.
You can leave this state by either pressing e again or pressing Esc.
"},{"location":"usage/general-usage/#widget-selection","title":"Widget selection","text":"To allow for widget-specific keybindings and expansion, there is the idea of widget selection in bottom, where you can focus on a specific widget to work with it. This can be done with the mouse (just click on the widget of interest) or keyboard (ex: Ctrl+Direction, see Key bindings for alternatives).
"},{"location":"usage/general-usage/#key-bindings","title":"Key bindings","text":"These are global or common keyboard shortcuts for the application, which you can see in-app through the ? shortcut. Note that key bindings are generally case-sensitive.
Binding Action q, Q, Ctrl+c Quit Esc Close dialog windows, search, widgets, or exit expanded mode Ctrl+r Reset display and any collected data f Freeze/unfreeze updating with new data ? Open help menu e Toggle expanding the currently selected widget Ctrl+Up Shift+Up K W Select the widget above Ctrl+Down Shift+Down J S Select the widget below Ctrl+Left Shift+Left H A Select the widget on the left Ctrl+Right Shift+Right L D Select the widget on the right Up , k Move up within a widget Down , j Move down within a widget Left h Alt+h Move left within a widget Right l Alt+l Move right within a widget g+g , Home Jump to the first entry G , End Jump to the last entry Page Up , Page Down Scroll up/down a table by a page Ctrl+u Scroll up a table by half a page Ctrl+d Scroll down a table by half a page"},{"location":"usage/general-usage/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects the widget"},{"location":"usage/widgets/battery/","title":"Battery Widget","text":"Warning
The battery features are unavailable if the binary is compiled with the battery feature disabled or if there are no batteries on the system!
The battery widget provides information about batteries on the system.
The battery widget can be enabled through either the --battery flag, the battery = true option in a config file, or specifying the widget in a custom layout.
The following data is displayed for batteries:
The battery widget also supports devices with multiple batteries, and you can switch between them using the keyboard or the mouse.
"},{"location":"usage/widgets/battery/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action Left h Alt+h Moves to the battery entry to the left of the current one Right l Alt+l Moves to the battery entry to the right of the current one"},{"location":"usage/widgets/battery/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects a battery entry"},{"location":"usage/widgets/cpu-graph/","title":"CPU Widget","text":"The CPU widget displays a visual representation of CPU usage over a time range.
"},{"location":"usage/widgets/cpu-graph/#features","title":"Features","text":"The CPU widget is composed of two parts: the graph and the legend:
Users can scroll through the legend using either the keyboard or mouse to select which entry to display on the graph. The \"All\" option shows every entry at the same time, though this may get a bit hard to follow if you have a large number of cores/threads.
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
"},{"location":"usage/widgets/cpu-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
"},{"location":"usage/widgets/cpu-graph/#graph","title":"Graph","text":"Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/cpu-graph/#legend","title":"Legend","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the legend G , End Jump to the last entry in the legend"},{"location":"usage/widgets/cpu-graph/#mouse-bindings","title":"Mouse bindings","text":""},{"location":"usage/widgets/cpu-graph/#graph_1","title":"Graph","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/cpu-graph/#legend_1","title":"Legend","text":"Binding Action Scroll Scroll through options to display in the graph Left Button Selects a CPU thread/average to show in the graph"},{"location":"usage/widgets/disk-io-graph/","title":"Disk I/O Graph Widget","text":"The disk I/O graph widget shows read and write rates over time.
"},{"location":"usage/widgets/disk-io-graph/#features","title":"Features","text":"The legend displays the current reads and writes per second in bytes for each mount/disk item. The y-axis automatically scales based on shown read/write values. By default this is in a linear scale, but it can be switched to a logarithmic scale through the config file.
The displayed time range can be adjusted through either the keyboard or mouse.
"},{"location":"usage/widgets/disk-io-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/disk-io-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/disk/","title":"Disk Widget","text":"The disk widget provides a table of useful disk and partition information, like I/O per second and total usage.
"},{"location":"usage/widgets/disk/#features","title":"Features","text":"The disk widget provides the following information:
Note that key bindings are generally case-sensitive.
Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table d Sort by disk, press again to reverse sorting order m Sort by mount, press again to reverse sorting order u Sort by amount used, press again to reverse sorting order n Sort by amount free, press again to reverse sorting order t Sort by total space available, press again to reverse sorting order p Sort by percentage used, press again to reverse sorting order r Sort by read rate, press again to reverse sorting order w Sort by write rate, press again to reverse sorting order"},{"location":"usage/widgets/disk/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects an entry in the table"},{"location":"usage/widgets/memory-graph/","title":"Memory Widget","text":"The memory widget provides a visual representation of RAM and swap usage over time.
"},{"location":"usage/widgets/memory-graph/#features","title":"Features","text":"The legend displays the current usage in terms of percentage and actual usage in binary units (KiB, MiB, GiB, etc.). If the total RAM or swap available is 0, then it is automatically hidden from the legend and graph.
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
This widget can also be configured to display Nvidia and AMD GPU memory usage (--disable_gpu on Linux/Windows to disable) or cache memory usage (--enable_cache_memory).
Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/memory-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/memory-graph/#how-are-memory-values-determined","title":"How are memory values determined?","text":""},{"location":"usage/widgets/memory-graph/#linux","title":"Linux","text":"Memory usage is calculated using the following formula based on values from /proc/meminfo (based on htop's implementation):
MemTotal - MemFree - Buffers - (Cached + SReclaimable - Shmem)\nYou can find more info on /proc/meminfo and its fields here.
In Windows, we calculate swap by querying Get-Counter \"\\Paging File(*)\\% Usage\". This is also what some libraries like psutil use. However, note there are also a few other valid methods of representing \"swap\" in Windows (e.g. using GetPerformanceInfo), which all slightly don't match.
The network widget provides a visual representation of network input and output per second, as well as noting the total amount received and transmitted.
"},{"location":"usage/widgets/network-graph/#features","title":"Features","text":"The legend displays the current reads and writes per second in bits, as well as the total amount read/written.
The y-axis automatically scales based on shown read/write values, and by default, is a linear scale based on base-10 units (e.x. kilobit, gigabit, etc.). Through configuration, the read/write per second unit can be changed to bytes, while the y-axis can be changed to a log scale and/or use base-2 units (e.x. kibibit, gibibit, etc.).
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
"},{"location":"usage/widgets/network-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/network-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/process/","title":"Process Widget","text":"The process widget displays a table containing information regarding a running process, along with sorting, searching, and process control features.
"},{"location":"usage/widgets/process/#features","title":"Features","text":"The process widget has three main components:
By default, the main process table displays the following information for each process:
Process uptime
Info
On Windows, the I/O counters will report all reads/writes, not just disk. See here for more details.
With the feature flag (--disable_gpu on Linux/Windows to disable) and gpu process columns enabled in the configuration:
See the processes configuration page on how to customize which columns are shown.
"},{"location":"usage/widgets/process/#sorting","title":"Sorting","text":"The table can be sorted by clicking on the table headers, which will either sort the table by that column, or if already sorting by that column, reverse the sorting order.
Alternatively, one can sort using the sort menu sub-widget, which is brought up using s or F6, and can be controlled by arrow keys or the mouse.
"},{"location":"usage/widgets/process/#grouping","title":"Grouping","text":"Pressing Tab in the table will group entries with the same name together. The PID column will be replaced with the number of entries in each group, and usage is added together when displayed.
Info
Note that the process state and user columns are disabled in this mode.
Info
Note that if tree mode is also active, processes cannot be grouped together due to the behaviour of the two modes somewhat clashing. This also reflects with default modes like group_processes.
Pressing d+d or F9 will allow you to terminate the currently selected process/process group. On Unix-like operating systems, you are also able to control which specific signals to send (e.g. SIGKILL, SIGTERM).
If you're on Windows, or if the disable_advanced_kill flag is set in the options or command-line (only available on Linux, macOS, and FreeBSD), then a simpler termination screen with just yes or no options will be shown.
Pressing t or F5 in the table toggles tree mode in the process widget, displaying processes in regard to their parent-child process relationships.
A process in tree mode can also be \"collapsed\", hiding its children and any descendants, using the either the -, +, or Left keys, or clicking on an entry. It can be expanded by using the -, +, or Right keys, or by clicking on the entry again. The Space key can also be used to toggle between the collapsed and expanded states.
Info
Note that if tree mode is active, processes cannot be grouped together due to the behaviour of the two modes somewhat clashing. This also reflects with default modes like group_processes.
You can show the full command instead of just the process name by pressing P.
"},{"location":"usage/widgets/process/#search","title":"Search","text":"Pressing / or Ctrl+f will open up the search sub-widget. By default, just typing in something will search by the process name.
This search can be further enhanced by matching by case, matching the entire word, or by regex.
We are able to also search for multiple things/conditions.
And if our search uses a keyword, we need to use quotation marks around the term to properly search it.
Lastly, we can refine our search even further based on the other columns, like PID, CPU usage, etc., as well as grouping together conditions.
You can also paste search queries (e.g. Shift+Ins, Ctrl+Shift+v).
"},{"location":"usage/widgets/process/#keywords","title":"Keywords","text":"Note all keywords are case-insensitive. To search for a process/command that collides with a keyword, surround the term with quotes (e.x. \"cpu\").
btm Matches by process or command name; supports regex pid pid=1044 Matches by PID; supports regex cpu cpu% cpu > 0.5 Matches the CPU column; supports comparison operators memb memb > 1000 b Matches the memory column in terms of bytes; supports comparison operators mem mem% mem < 0.5 Matches the memory column in terms of percent; supports comparison operators read r/s rps read = 1 mb Matches the read/s column in terms of bytes; supports comparison operators write w/s wps write >= 1 kb Matches the write/s column in terms of bytes; supports comparison operators tread t.read tread <= 1024 gb Matches the total read column in terms of bytes; supports comparison operators twrite t.write twrite > 1024 tb Matches the total write column in terms of bytes; supports comparison operators user user=root Matches by user; supports regex state state=running Matches by state; supports regex () (<COND 1> AND <COND 2>) OR <COND 3> Group together a condition gmem gmem > 1000 b Matches the gpu memory column in terms of bytes; supports comparison operators gmem% gmem% < 0.5 Matches the gpu memory column in terms of percent; supports comparison operators gpu% gpu% > 0 Matches the gpu usage column in terms of percent; supports comparison operators"},{"location":"usage/widgets/process/#comparison-operators","title":"Comparison operators","text":"Keywords Description = Checks if the values are equal != Checks if the values are not equal > Checks if the left value is strictly greater than the right < Checks if the left value is strictly less than the right >= Checks if the left value is greater than or equal to the right <= Checks if the left value is less than or equal to the right"},{"location":"usage/widgets/process/#logical-operators","title":"Logical operators","text":"Note all operators are case-insensitive, and the and operator takes precedence over the or operator.
and && <Space> <COND 1> and <COND 2> <COND 1> && <COND 2> <COND 1> <COND 2> Requires both conditions to be true to match or || <COND 1> or <COND 2> <COND 1> || <COND 2> Requires at least one condition to be true to match ! !<COND> !(<COND 1> or <COND 2>) Inverts the following condition or group ! is reserved as an operator, so it cannot appear bare as a value. To match a literal ! in a name or string field, quote it \u2014 e.g. \"foo!\" or user = \"!\". A bare ! with nothing parseable after it (such as user = ! or ! on its own) is rejected.
All units are case-insensitive.
Keywords DescriptionB Bytes KB Kilobytes MB Megabytes GB Gigabytes TB Terabytes KiB Kibibytes MiB Mebibytes GiB Gibibytes TiB Tebibytes"},{"location":"usage/widgets/process/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
"},{"location":"usage/widgets/process/#process-table","title":"Process table","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table d+d , F9 Send a kill signal to the selected process c Sort by CPU usage, press again to reverse sorting order m Sort by memory usage, press again to reverse sorting order p Sort by PID name, press again to reverse sorting order n Sort by process name, press again to reverse sorting order Tab Toggle grouping processes with the same name P Toggle between showing the full command or just the process name Ctrl+f , / Toggle showing the search sub-widget s , F6, Del (Fn+Del on macOS) Toggle showing the sort sub-widget I Invert the current sort % Toggle between values and percentages for memory usage t , F5 Toggle tree mode M Sort by gpu memory usage, press again to reverse sorting order C Sort by gpu usage, press again to reverse sorting order z Toggle the hiding of kernel threads"},{"location":"usage/widgets/process/#sort-sub-widget","title":"Sort sub-widget","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table Esc Close the sort sub-widget Enter Sorts the corresponding process table"},{"location":"usage/widgets/process/#search-sub-widget","title":"Search sub-widget","text":"Binding Action Left h Alt+h Moves the cursor left Right l Alt+l Moves the cursor right Esc Close the search widget (retains the filter) Ctrl+a Skip to the start of the search query Ctrl+e Skip to the end of the search query Ctrl+u Clear the current search query Ctrl+w Delete a word behind the cursor Ctrl+h Delete the character behind the cursor Backspace Delete the character behind the cursor Del (Fn+Del on macOS) Delete the character at the cursor Alt+c , F1 Toggle matching case Alt+w , F2 Toggle matching the entire word Alt+r , F3 Toggle using regex"},{"location":"usage/widgets/process/#mouse-bindings","title":"Mouse bindings","text":""},{"location":"usage/widgets/process/#process-table_1","title":"Process table","text":"Binding Action Scroll Selects a CPU thread/average to show in the graph Left Button Table header: Sorts/reverse sorts the table by the column Table entry: Selects an entry in the table, if in tree mode, collapses/expands the entry's children"},{"location":"usage/widgets/process/#sort-sub-widget_1","title":"Sort sub-widget","text":"Binding Action Left Button Selects an entry in the table"},{"location":"usage/widgets/temperature-graph/","title":"Temperature Graph Widget","text":"The temperature graph widget provides temperature readings over time.
"},{"location":"usage/widgets/temperature-graph/#features","title":"Features","text":"Each detected sensor is drawn as its own line. The y-axis is in the configured temperature unit (Celsius by default; see the --celsius/--fahrenheit/--kelvin arguments or flags.temperature_type config option).
By default the y-axis is bounded at 100\u00b0C (or the equivalent in the configured unit) and grows automatically. An upper bound can also be set explicitly via the config file.
The displayed time range can be adjusted through either the keyboard or mouse.
"},{"location":"usage/widgets/temperature-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/temperature-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/temperature-table/","title":"Temperature Widget","text":"The temperature widget provides a table of temperature sensors and their current temperature.
"},{"location":"usage/widgets/temperature-table/#features","title":"Features","text":"The temperature widget provides the sensor name as well as its current temperature.
This widget can also be configured to display Nvidia and AMD GPU temperatures (--disable_gpu on Linux/Windows to disable).
Note that key bindings are generally case-sensitive.
Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table t Sort by temperature, press again to reverse sorting order s Sort by sensor name, press again to reverse sorting order"},{"location":"usage/widgets/temperature-table/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects an entry in the table"}]} \ No newline at end of file +{"config":{"lang":["en"],"separator":"[\\s\\-]+","pipeline":["stopWordFilter"],"fields":{"title":{"boost":1000.0},"text":{"boost":1.0},"tags":{"boost":1000000.0}}},"docs":[{"location":"","title":"bottom","text":"A customizable cross-platform graphical process/system monitor for the terminal, supporting Linux, macOS, and Windows. Inspired by other tools like gtop, gotop, and htop.
This site serves as extended documentation for bottom alongside the README.md.
Warning
Some areas of this site are still in progress and may be missing details. Feel free to suggest/contribute changes!
"},{"location":"#installation","title":"Installation","text":"Tip
It's as good idea to first check out the Support page to see if your system is officially supported!
Tip
If you're facing some issues during/after installation, check out the Troubleshooting page for some common problems and solutions.
To install bottom, refer to the installation section of the README.md, which contains a list of all the installation methods.
The command to run bottom is btm.
You can refer to the usage pages for more details on using bottom (e.g. keybinds, some features, a general overview of what each widget does).
To configure bottom (e.g. how it behaves, how it looks, etc.) refer to the command-line options page for temporary settings, or the config file page for more permanent settings.
"},{"location":"#contribution","title":"Contribution","text":"New contributors are always welcome! See the contribution section for how to contribute to bottom, whether it be filing issues, writing documentation, creating pull requests, etc.
"},{"location":"troubleshooting/","title":"Troubleshooting/Known Issues","text":""},{"location":"troubleshooting/#the-graph-points-look-brokenstrange","title":"The graph points look broken/strange","text":"It's possible that your graphs don't look great out of the box due to the reliance on braille characters to draw them. This could cause problems if your terminal's font does not support them, or your terminal is not configured properly to draw them.
An example of missing braille fonts in PowershellSome possible solutions are included below.
"},{"location":"troubleshooting/#use-dot-markers-instead","title":"Use dot markers instead","text":"One alternative is to use the --dot_marker option to render graph charts using dots instead of the braille characters, which generally seems better supported out of the box, at the expense of looking less intricate:
btm --dot_marker"},{"location":"troubleshooting/#use-a-font-that-supports-braille-fonts","title":"Use a font that supports braille fonts","text":"Another (better) alternative is to install a font that supports braille fonts, and configure your terminal emulator to use it. For example, installing something like UBraille or Iosevka and ensuring your terminal uses it should work.
"},{"location":"troubleshooting/#linuxmacosunix","title":"Linux/macOS/Unix","text":"Solutions mostly depend on what terminal emulator you are using, so unfortunately, I can't give specific instructions. Here are some possible solutions:
gnu-free-fonts if installed, as that is known to cause problems with braille markersttf-symbola or ttf-ubraille for your terminal emulator to try and automatically fall back toU+2800 to U+28FF range.symbol_map U+2800-U+28FF Symbola.For some more possible solutions:
Note: I would advise backing up your registry beforehand if you aren't sure what you are doing!
Let's say you're installing Iosevka. The steps you can take are:
Win+R and opening regedit, or just opening it from the Start Menu.In the registry editor, go to
HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Windows NT\\CurrentVersion\\Console\\TrueTypeFont\nHere, add a new String value, and set the Name to a bunch of 0's (e.g. 000 - make sure the name isn't already used), then set the Data to the font name (e.g. Iosevka).
The last entry is the new entry for Iosevka
Then, open the Command Prompt/PowerShell, and right-click on the top bar, and open \"Properties\":
From here, go to \"Font\", and set the font to your new font (so in this example, Iosevka):
This is a known issue, some sensors may require admin privileges to get sensor data.
"},{"location":"troubleshooting/#why-dont-i-see-dual-batteries-on-windows-reported-separately-eg-thinkpads","title":"Why don't I see dual batteries on Windows reported separately? (e.g. Thinkpads)","text":"This is a known issue which seems to be with how batteries are being detected on Windows.
"},{"location":"troubleshooting/#why-cant-i-see-all-my-temperature-sensors-on-wsl","title":"Why can't I see all my temperature sensors on WSL?","text":"This is a known limitation with WSL. Due to how it works, hosts may not expose their temperature sensors and therefore, temperature sensors might be missing.
"},{"location":"troubleshooting/#why-does-wsl2-not-match-task-manager","title":"Why does WSL2 not match Task Manager?","text":"This is a known limitation with WSL2. Due to how WSL2 works, the two might not match up in terms of reported data.
"},{"location":"troubleshooting/#why-cant-i-see-all-my-processesprocess-data-on-macos","title":"Why can't I see all my processes/process data on macOS?","text":"You may have to run the program with elevated privileges to work around it - for example:
sudo btm\nWarning
Please note that you should be certain that you trust any software you grant root privileges.
There are measures taken to try to maximize the amount of information obtained without elevated privileges. For example, one can modify the instructions found on the htop wiki on how to run htop without sudo for bottom. However, please understand the potential security risks before doing so!
"},{"location":"troubleshooting/#my-configuration-file-isnt-working","title":"My configuration file isn't working","text":"If your configuration files aren't working, here are a few things to try:
"},{"location":"troubleshooting/#check-the-formatting","title":"Check the formatting","text":"It may be handy to refer to the automatically generated config files or the sample configuration files. The config files also follow the TOML format.
Also make sure your config option settings are in the right location - for example, to set your temperature type, you must set it under the [flags] table:
[flags]\ntemperature_type = \"f\"\nMeanwhile, if you want to set a custom colour or styling scheme, it would be under the [styles] table - for example:
[styles.tables.headers]\ncolour = \"LightBlue\"\nbold = true\nTo help validate your configuration files, there is JSON Schema support if your IDE/editor supports it.
"},{"location":"troubleshooting/#check-the-configuration-file-location","title":"Check the configuration file location","text":"Make sure bottom is reading the right configuration file. By default, bottom looks for config files at these locations:
OS Default Config Location macOS$HOME/Library/Application Support/bottom/bottom.toml ~/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Linux ~/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Windows C:\\Users\\<USER>\\AppData\\Roaming\\bottom\\bottom.toml If you want to use a config file in another location, use the --config or -C flags along with the path to the configuration file, like so:
btm -C path_to_config\nMake sure bottom is given the correct permissions in order to collect data. Snapcraft explains how to do so, but the TL;DR is:
sudo snap connect bottom:mount-observe\nsudo snap connect bottom:hardware-observe\nsudo snap connect bottom:system-observe\nsudo snap connect bottom:process-control\nThe underlying interface we use for NVIDIA GPU information, nvml, only works with glibc and does not work with musl at the moment (see this forum post for some more details). As such, bottom may fail to get NVIDIA GPU information when using a musl-based binary until this is resolved. This applies to Linux and Windows from my understanding.
To resolve this, use glibc-based binary builds if possible (e.g. the gnu binaries/non-musl packages in releases).
If you're still having issues, feel free to open a discussion question about it, and I (or others) can try to help.
"},{"location":"configuration/command-line-options/","title":"Command-line Options","text":"The following options can be provided to bottom in the command line to change the behaviour of the program. You can also see information on these options by running btm -h, or run btm --help to display more detailed information on each option:
--autohide_time Temporarily shows the time scale in graphs. -b, --basic Hides graphs and uses a more basic look. -C, --config_location <PATH> Sets the location of the config file. -t, --default_time_value <TIME> Default time value for graphs. --default_widget_count <N> Sets the N'th selected widget type as the default. --default_widget_type <WIDGET> Sets the default widget type. Use --help for more info. --disable_click Disables mouse clicks. --disable_keys Disables keyboard shortcuts, INCLUDING the ones that stop bottom. -m, --dot_marker Uses a dot marker for graphs. -e, --expanded Expand the default widget upon starting the app. --hide_time Hides the time scale from being shown. -r, --rate <TIME> Sets how often data is refreshed. --retention <TIME> How far back data will be stored up to. --show_table_scroll_position Shows the list scroll position tracker in the widget title for table widgets. -d, --time_delta <TIME> The amount of time changed upon zooming."},{"location":"configuration/command-line-options/#process-options","title":"Process Options","text":"Option Behaviour -S, --case_sensitive Enables case sensitivity by default when searching. -u, --current_usage Calculates process CPU usage as a percentage of current usage rather than total usage. --disable_advanced_kill Hides additional stopping options on Unix-like systems. --read_only Prevents performing any actions that affect the system (e.g. stopping processes). --get_threads Also gather process thread information. -g, --group_processes Groups processes with the same name by default. No effect if --tree is set. --hide_k_threads Hide kernel threads from being shown. --process_memory_as_value Defaults to showing process memory usage by value. --process_command Shows the full command name instead of the process name by default. -R, --regex Enables regex by default while searching. -T, --tree Makes the process widget use tree mode by default. --tree_collapse Collapse the process tree by default. -n, --unnormalized_cpu Show process CPU% usage without averaging over the number of CPU cores. -W, --whole_word Enables whole-word matching by default while searching."},{"location":"configuration/command-line-options/#temperature-options","title":"Temperature Options","text":"Option Behaviour -c, --celsius Use Celsius as the temperature unit. Default. -f, --fahrenheit Use Fahrenheit as the temperature unit. -k, --kelvin Use Kelvin as the temperature unit."},{"location":"configuration/command-line-options/#cpu-options","title":"CPU Options","text":"Option Behaviour --cpu_left_legend Puts the CPU chart legend on the left side. --default_cpu_entry Sets which CPU entry type is selected by default. -a, --hide_avg_cpu Hides the average CPU usage entry."},{"location":"configuration/command-line-options/#memory-options","title":"Memory Options","text":"Option Behaviour --memory_legend <POSITION> Where to place the legend for the memory chart widget. --enable_cache_memory Enable collecting and displaying cache and buffer memory. --free_arc Subtract ARC memory that can be freed from memory usage. --short_gpu_names Use short GPU names in the memory widget."},{"location":"configuration/command-line-options/#network-options","title":"Network Options","text":"Option Behaviour --network_legend <POSITION> Where to place the legend for the network chart widget. --network_use_bytes Displays the network widget using bytes. --network_use_binary_prefix Displays the network widget with binary prefixes. --network_use_log Displays the network widget with a log scale. --show_packets Displays packet rate and average packet size info. --use_old_network_legend Uses a separate network legend. --network_start_zeroed Show total network usage from app startup rather than on boot."},{"location":"configuration/command-line-options/#battery-options","title":"Battery Options","text":"Option Behaviour --battery Shows the battery widget in non-custom layouts."},{"location":"configuration/command-line-options/#gpu-options","title":"GPU Options","text":"Option Behaviour --disable_gpu Disable collecting and displaying NVIDIA and AMD GPU information."},{"location":"configuration/command-line-options/#style-options","title":"Style Options","text":"Option Behaviour --theme <SCHEME> Use a built-in colour theme, use '--help' for info on the colours."},{"location":"configuration/command-line-options/#other-options","title":"Other Options","text":"Option Behaviour -h, --help Prints help info (for more details use --help.) -V, --version Prints version information."},{"location":"configuration/config-file/","title":"Config File","text":"For persistent configuration, and for certain configuration options, bottom supports config files.
"},{"location":"configuration/config-file/#default-config-file","title":"Default Config File","text":"If no config file argument is given, it will automatically look for a config file at these locations:
OS Default Config Location macOS$HOME/Library/Application Support/bottom/bottom.toml $HOME/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Linux $HOME/.config/bottom/bottom.toml $XDG_CONFIG_HOME/bottom/bottom.toml Windows C:\\Users\\<USER>\\AppData\\Roaming\\bottom\\bottom.toml If the config file doesn't exist at the path, bottom will automatically try to create a new config file at the location with default values.
"},{"location":"configuration/config-file/#json-schema","title":"JSON Schema","text":"The configuration file also has JSON Schema support to make it easier to manage, if your IDE/editor supports it.
"},{"location":"configuration/config-file/cpu-graph/","title":"CPU Configuration","text":"If you want to change some of the default behaviour of the CPU graph widget, you can configure things under the [cpu] section.
You can configure which CPU graph is shown by default on startup by setting cpu.default. Defaults to \"all\", which shows all entries.
[cpu]\n# One of \"all\" (default), \"average\"/\"avg\"\ndefault = \"average\"\nYou can configure whether CPU usage values are shown with a decimal place by setting cpu.show_decimal. Defaults to false.
[cpu]\nshow_decimal = true\nYou can hide the average CPU entry entirely by setting cpu.hide_avg_cpu. Defaults to false.
[cpu]\nhide_avg_cpu = true\nYou can place the CPU chart legend on the left side by setting cpu.left_legend. Defaults to false.
[cpu]\nleft_legend = true\nIn basic mode, you can give the average CPU entry a dedicated row by setting cpu.basic_average_cpu_row. Defaults to false.
[cpu]\nbasic_average_cpu_row = true\nThe disk I/O graph widget is configured under [disk_io_graph].
By default, both the read and write rate lines are shown. They can be toggled off independently.
[disk_io_graph]\n# Whether to show the read rate line. Defaults to true.\nshow_read = true\n\n# Whether to show the write rate line. Defaults to true.\nshow_write = true\nBy default, legend entries are labelled by disk name (e.g. sda, nvme0n1). To use mount points instead (e.g. /, /home):
[disk_io_graph]\nlegend = \"mount\"\nValid values are \"disk\" (default) and \"mount\".
The y-axis can be switched to a logarithmic scale:
[disk_io_graph]\nuse_log = true\nThe location of the legend can be set with legend_position. Valid values are none, top-left, top, top-right, left, right, bottom-left, bottom, and bottom-right. Defaults to top-right.
[disk_io_graph]\nlegend_position = \"top-right\"\nBy default, only mounted devices are shown. To also show unmounted devices on Linux, enable include_unmounted:
[disk_io_graph]\ninclude_unmounted = true\nYou can filter out what entries to show by configuring [disk_io_graph.name_filter] and [disk_io_graph.mount_filter] to filter by name and mount point respectively. In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
If we wanted to ignoring any entry with a name that matches /dev/sda:
[disk_io_graph.name_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"/dev/sda\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nYou can configure which columns are shown by the disk table widget by setting the columns setting:
[disk]\n# Pick which columns you want to use in any order.\ncolumns = [\"Disk\", \"Mount\", \"Used\", \"Free\", \"Total\", \"Used%\", \"R/s\", \"W/s\"]\nYou can customize the default sort column (by default, it sorts by disk name in ascending order). For example, to sort by the read rate:
[disk]\ndefault_sort = \"R/s\"\nYou can use any valid column name here (e.g. \"Disk\", \"Mount\", etc.). Note that if you put a column name that is not actually used, the default sort will just be the first column shown.
You can also set the sort order by changing disk.sort_order with \"Ascending\" or \"Descending\". Defaults to \"Ascending\":
[disk]\nsort_order = \"Ascending\"\nBy default, only mounted devices are shown. To also show unmounted devices on Linux, enable include_unmounted:
[disk]\ninclude_unmounted = true\nUnmounted devices show I/O activity but have no used/free space (those require a live mount), so those columns appear as N/A. Pseudo-devices (loop*, ram*, zram*) are skipped, and you can hide other noisy devices with the name filter.
You can filter out what entries to show by configuring [disk.name_filter] and [disk.mount_filter] to filter by name and mount point respectively. In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, consider a disk widget showing these entries:
If we wanted to ignoring any entry with a name that matches /dev/sda:
[disk.name_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"/dev/sda\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nThis would give us:
We can also combine both the name filter and mount filter. For example:
[disk.name_filter]\nis_list_ignored = false\nlist = [\"/dev/sda\"]\nregex = true\ncase_sensitive = false\nwhole_word = false\n\n[disk.mount_filter]\nis_list_ignored = true\nlist = [\"/mnt/.*\", \"/\"]\nregex = true\ncase_sensitive = false\nwhole_word = true\nThis gives us:
"},{"location":"configuration/config-file/flags/","title":"Flags","text":"Warning
This section is in progress, and is just copied from the old documentation.
You can configure flags by putting them in [flags] table. Example:
[flags]\nhide_avg_cpu = true\nMost of the command line flags have config file equivalents to avoid having to type them out each time:
Field Type Functionalityhide_avg_cpu Boolean Deprecated - use cpu.hide_avg_cpu. Hides the average CPU usage. dot_marker Boolean Uses a dot marker for graphs. cpu_left_legend Boolean Deprecated - use cpu.left_legend. Puts the CPU chart legend to the left side. current_usage Boolean Deprecated - use processes.current_usage. Sets process CPU% to be based on current CPU%. group_processes Boolean Deprecated - use processes.default_grouped. Groups processes with the same name by default. case_sensitive Boolean Deprecated - use processes.case_sensitive. Enables case sensitivity by default. whole_word Boolean Deprecated - use processes.whole_word. Enables whole-word matching by default. regex Boolean Deprecated - use processes.regex. Enables regex by default. basic Boolean Hides graphs and uses a more basic look. use_old_network_legend Boolean DEPRECATED - uses the older network legend. battery Boolean Shows the battery widget. rate Unsigned Int (represents milliseconds) or String (represents human time) Sets a refresh rate in ms. default_time_value Unsigned Int (represents milliseconds) or String (represents human time) Default time value for graphs in ms. time_delta Unsigned Int (represents milliseconds) or String (represents human time) The amount in ms changed upon zooming. hide_time Boolean Hides the time scale. temperature_type String (one of [\"k\", \"f\", \"c\", \"kelvin\", \"fahrenheit\", \"celsius\"]) Sets the temperature unit type. default_widget_type String (one of [\"cpu\", \"proc\", \"net\", \"temp\", \"mem\", \"disk\"], same as layout options) Sets the default widget type, use --help for more info. default_widget_count Unsigned Int (represents which default_widget_type) Sets the n'th selected widget type as the default. disable_click Boolean Disables mouse clicks. enable_cache_memory Boolean Deprecated - use memory.cache_memory. Enable cache and buffer memory stats (not available on Windows). process_memory_as_value Boolean Deprecated - use processes.default_memory_value. Defaults to showing process memory usage by value. tree Boolean Deprecated - use processes.default_tree. Defaults to showing the process widget in tree mode. show_table_scroll_position Boolean Shows the scroll position tracker in table widgets. show_table_scroll_bar Boolean Shows a scroll bar on the right edge of table widgets. process_command Boolean Deprecated - use processes.process_command. Show processes as their commands by default. disable_advanced_kill Boolean Deprecated - use processes.disable_advanced_kill. Disable being able to send signals to processes on supported Unix-like systems. Only available on Linux, macOS, and FreeBSD. read_only Boolean Prevents performing any actions that affect the system (e.g. stopping processes). network_use_binary_prefix Boolean Deprecated - use network_graph.use_binary_prefix. Displays the network widget with binary prefixes. network_use_bytes Boolean Deprecated - use network_graph.use_bytes. Displays the network widget using bytes. network_use_log Boolean Deprecated - use network_graph.use_log. Displays the network widget with a log scale. disable_gpu Boolean Disable NVIDIA and AMD GPU data collection. retention String (human readable time, such as \"10m\", \"1h\", etc.) How much data is stored at once in terms of time. unnormalized_cpu Boolean Deprecated - use processes.unnormalized_cpu. Show process CPU% without normalizing over the number of cores. expanded Boolean Expand the default widget upon starting the app. memory_legend String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Deprecated - use memory_graph.legend_position or memory.legend_position; where to place the legend for the memory widget. network_legend String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Deprecated - use network_graph.legend_position or network.legend_position; where to place the legend for the network widget. average_cpu_row Boolean Deprecated - use cpu.basic_average_cpu_row. Moves the average CPU usage entry to its own row when using basic mode. tree_collapse Boolean Deprecated - use processes.tree_collapse. Collapse process tree by default. autohide_time Boolean Temporarily shows the time scale in graphs. table_gap String (one of [\"none\", \"space\", \"line\"]) Controls the gap between table headers and data rows. Defaults to \"space\". disable_keys Boolean Disables keyboard shortcuts, including the ones that stop bottom. no_write Boolean Disables writing to the config file. hide_k_threads Boolean Deprecated - use processes.hide_k_threads. Hide kernel threads from being shown. free_arc Boolean Deprecated - use memory.free_arc. Subtract ARC memory that can be freed from memory usage."},{"location":"configuration/config-file/layout/","title":"Layout","text":"Warning
This section is in progress, and is just copied from the old documentation.
bottom supports customizable layouts via the config file. Currently, layouts are controlled by using TOML objects and arrays.
For example, given the sample layout:
[[row]]\n [[row.child]]\n type=\"cpu\"\n[[row]]\n ratio=2\n [[row.child]]\n ratio=4\n type=\"mem\"\n [[row.child]]\n ratio=3\n [[row.child.child]]\n type=\"temp\"\n [[row.child.child]]\n type=\"disk\"\nThis would give a layout that has two rows, with a 1:2 ratio. The first row has only the CPU widget. The second row is split into two columns with a 4:3 ratio. The first column contains the memory widget. The second column is split into two rows with a 1:1 ratio. The first is the temperature widget, the second is the disk widget.
This is what the layout would look like when run:
Each [[row]] represents a row in the layout. A row can have any number of child values. Each [[row.child]] represents either a column or a widget. A column can have any number of child values as well. Each [[row.child.child]] represents a widget. A widget is represented by having a type field set to a string.
The following type values are supported:
\"cpu\" CPU chart and legend \"mem\", \"memory\" Memory chart \"net\", \"network\" Network chart and legend \"proc\", \"process\", \"processes\" Process table and search \"temp\", \"temperature\" Temperature table \"temp_graph\", \"temperature_graph\" Temperature graph \"disk\" Disk table \"disk_io_graph\" Disk I/O graph \"empty\" An empty space \"batt\", \"battery\" Battery statistics Each component of the layout accepts a ratio value. If this is not set, it defaults to 1.
Furthermore, you can have duplicate widgets.
For an example, look at the default config, which contains the default layout.
"},{"location":"configuration/config-file/memory-graph/","title":"Memory Graph Configuration","text":"If you want to change some of the default behaviour of the memory graph widget, you can configure things under the [memory_graph] (or [memory]) section.
You can change where the legend for the graph is placed within the widget itself (or hidden with \"none\"). The default is \"top-right\".
[memory_graph]\n# One of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]\nlegend_position = \"top-left\"\nOn Linux, you can change whether the memory used by cache/slabs is collected and shown. By default, it is false.
[memory_graph]\ncache_memory = true\nIf GPU support is enabled, you can use short_gpu_names to display simplified GPU names in the memory graph widget instead of the full GPU names. If enabled, then if there's only one GPU, it'll just list the entry as GPU; if there are multiple GPUs, it'll use entries like GPU0, GPU1, etc. By default, this is disabled.
[memory_graph]\nshort_gpu_names = true\nIf ZFS is detected (note that the zfs feature must be enabled if built manually), you can enable memory_graph.free_arc to not count ARC memory in the RAM usage calculations. Disabled by default.
[memory_graph]\nfree_arc = true\nIf you want to change some of the default behaviour of the network graph widget, you can configure some things in the config file.
Field Type Functionalityshow_packets Boolean Displays packet rate and average packet size info. legend_position String (one of [\"none\", \"top-left\", \"top\", \"top-right\", \"left\", \"right\", \"bottom-left\", \"bottom\", \"bottom-right\"]) Where to place the legend for the network widget. use_bytes Boolean Displays the network widget using bytes. Defaults to bits. use_log Boolean Displays the network widget with a log scale. Defaults to a non-log scale. use_binary_prefix Boolean Displays the network widget with a binary prefix (e.g. kibibits) rather than a decimal prefix (e.g. kilobits). Defaults to decimal prefixes. start_zeroed Boolean Zeroes out the total network usage (\"All\") counters such that it shows the total usage since the app is started, rather than the total usage since boot."},{"location":"configuration/config-file/network-graph/#filtering-entries","title":"Filtering Entries","text":"You can filter out what entries to show by configuring [network_graph.interface_filter] . In particular, you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, here we are ignoring any entry with a name that matches the regex eth0.*, or specifically virbr0.
[network_graph.interface_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"virbr0\", \"eth0slab\"]\n\n# Whether to use regex. Defaults to false.\nregex = true\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nIf you want to change some of the default behaviour of the processes widget, you can configure some things in the config file.
Field Type Functionalityget_threads Boolean Gather process thread information. hide_k_threads Boolean Hide kernel threads from being shown. Linux only. tree_collapse Boolean Collapse the process tree by default when tree mode is set. process_command Boolean Shows the full command name instead of the process name by default. disable_advanced_kill Boolean Disable the advanced kill dialog and just show the basic one with no options. Linux, macOS, FreeBSD only. default_memory_value Boolean Defaults to showing process memory usage by value. default_grouped Boolean Groups processes with the same name by default. No effect if --tree is set. regex Boolean Enables regex by default while searching. case_sensitive Boolean Enables case sensitivity by default when searching. whole_word Boolean Enables whole-word matching by default while searching. default_tree Boolean Makes the process widget use tree mode by default. current_usage Boolean Calculates process CPU usage as a percentage of current usage rather than total usage. unnormalized_cpu Boolean Show process CPU% usage without averaging over the number of CPU cores."},{"location":"configuration/config-file/processes/#columns","title":"Columns","text":"You can configure which columns are shown by the process widget by setting the columns setting:
[processes]\n# Pick which columns you want to use in any order.\ncolumns = [\"cpu%\", \"mem%\", \"pid\", \"name\", \"read\", \"write\", \"tread\", \"twrite\", \"state\", \"user\", \"time\", \"gmem%\", \"gpu%\"]\nBy default, the process widget starts sorted by CPU usage in descending order. You can change the column it sorts by at startup:
[processes]\ndefault_sort = \"mem\"\nAny of the column names accepted by columns work here (e.g. \"cpu%\", \"mem\", \"pid\", \"name\", \"read\"). If the column you pick is not actually shown by the widget, the built-in default is used instead.
The same setting is also exposed as a CLI flag, which takes precedence over the config file:
btm --process_default_sort mem\nYou can also set the default order by changing processes.sort_order with \"Ascending\" or \"Descending\". If not set, then by default, it will use a sort order that is reasonable (e.g. descending for CPU/memory usage, ascending for process names).
[processes]\nsort_order = \"Ascending\"\nVarious parts of the bottom can be styled, using either built-in themes or custom theming.
"},{"location":"configuration/config-file/styling/#precedence","title":"Precedence","text":"As there are a few ways styles can be applied to bottom, the order of which styles are prioritized are, in order of highest precedence to lowest precedence:
btm --theme gruvbox)If nothing is set, it will fall back to the default theme.
"},{"location":"configuration/config-file/styling/#built-in-styles","title":"Built-in styles","text":"bottom has a few built-in themes:
These themes all also have light variants to support terminals using lighter colours.
To set the theme from the command line:
btm --theme gruvbox\nTo set the theme using the config file:
[styles]\ntheme = \"gruvbox\"\nbottom's components can also be individually styled by the user to control the colour of the text style.
"},{"location":"configuration/config-file/styling/#colours","title":"Colours","text":"Info
Note that anywhere \"colour\" is used, it can be substituted for \"color\":
# This is okay\n[styles.widgets.widget_title]\ncolour = \"black\"\nbg_colour = \"white\"\n\n[styles.cpu]\nall_entry_colour = \"green\"\n\n# This is also okay\n[styles.widgets.widget_title]\ncolor = \"black\"\nbg_color = \"white\"\n\n[styles.cpu]\nall_entry_color = \"green\"\nYou can configure the colours for components with strings that are either hex colours (e.g. \"#ffffff\"), RGB colours (e.g. \"255, 255, 255\"), or named colours. Named colours are one of the following strings:
\"Black\"\"Red\"\"Green\"\"Yellow\"\"Blue\"\"Magenta\"\"Cyan\"\"Gray\"\"DarkGray\"\"LightRed\"\"LightGreen\"\"LightYellow\"\"LightBlue\"\"LightMagenta\"\"LightCyan\"\"White\"Text can generally be styled using the following TOML table:
[field]\n# The foreground colour of text.\ncolour = \"black\"\n\n# The background colour of text.\nbg_colour = \"blue\"\n\n# Whether to make the text bold.\nbold = false\n\n# Inline table version\nfield = { colour = \"black\", bg_colour = \"blue\", bold = false }\nAll fields are optional; by default if bg_colour is not set then there will be no background colour.
If you just want to style text by setting the foreground colour, for brevity, then you can also just set the field to be the colour itself. For example:
[styles.widgets]\nselected_text = \"#fff\"\nThese can be set under [styles.cpu]:
all_entry_colour The colour of the \"All\" CPU label all_entry_colour = \"Red\" avg_entry_colour The colour of the average CPU label and graph line avg_entry_colour = \"255, 0, 255\" cpu_core_colours Colour of each CPU threads' label and graph line. Read in order. cpu_core_colours = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#temperature-graph","title":"Temperature Graph","text":"These can be set under [styles.temp_graph]:
temp_graph_colour_styles Colour of each temperature sensor's graph line. Read in order. temp_graph_colour_styles = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#memory","title":"Memory","text":"These can be set under [styles.memory]:
ram_colour The colour of the RAM label and graph line ram_colour = \"Red\" cache_colour The colour of the cache label and graph line. Does not do anything on Windows. cache_colour = \"#ffffff\" swap_colour The colour of the swap label and graph line swap_colour = \"255, 0, 255\" arc_colour The colour of the ARC label and graph line arc_colour = \"Blue\" gpu_colours Colour of each GPU's memory label and graph line. Read in order. gpu_colours = [\"Red\", \"Blue\", \"Green\"]"},{"location":"configuration/config-file/styling/#network","title":"Network","text":"These can be set under [styles.network]:
rx_colour The colour of the RX (download) label and graph line rx_colour = \"Red\" tx_colour The colour of the TX (upload) label and graph line tx_colour = \"#ffffff\" rx_total_colour The colour of the total RX (download) label in basic mode rx_total_colour = \"0, 0, 0\" tx_total_colour The colour of the total TX (upload) label in basic mode tx_total_colour = \"#000\""},{"location":"configuration/config-file/styling/#battery","title":"Battery","text":"These can be set under [styles.battery]:
high_battery_colour The colour of the battery widget bar when the battery is over 50% high_battery_colour = \"Red\" medium_battery_colour The colour of the battery widget bar when the battery between 10% to 50% medium_battery_colour = \"#ffffff\" low_battery_colour The colour of the battery widget bar when the battery is under 10% low_battery_colour = \"0, 0, 0\""},{"location":"configuration/config-file/styling/#tables","title":"Tables","text":"These can be set under [styles.tables]:
headers Text styling for table headers headers = { colour = \"red\", bg_colour = \"black\", bold = true }"},{"location":"configuration/config-file/styling/#graphs","title":"Graphs","text":"These can be set under [styles.graphs]:
graph_colour The general colour of the parts of the graph graph_colour = \"white\" legend_text Text styling for graph's legend text legend_text = { colour = \"black\", bg_colour = \"blue\", bold = true }"},{"location":"configuration/config-file/styling/#general-widget-settings","title":"General widget settings","text":"These can be set under [styles.widgets]:
border_colour The colour of the widgets' borders border_colour = \"white\" selected_border_colour The colour of a widget's borders when the widget is selected selected_border_colour = \"white\" widget_title Text styling for a widget's title widget_title = { colour = \"black\", bg_colour = \"blue\", bold = true } bg_colour The background colour of the widgets bg_colour = \"black\" text Text styling for text in general text = { colour = \"black\", bg_colour = \"blue\", bold = true } selected_text Text styling for text when representing something that is selected selected_text = { colour = \"black\", bg_colour = \"blue\", bold = true } disabled_text Text styling for text when representing something that is disabled disabled_text = { colour = \"black\", bg_colour = \"blue\", bold = true } thread_text Text styling for text when representing process thread, only usable on Linux at the moment thread_text = { colour = \"green\", bg_colour = \"blue\", bold = true }"},{"location":"configuration/config-file/temperature-graph/","title":"Temperature Graph Configuration","text":"The temperature graph widget is configured under [temperature_graph].
The location of the legend can be set with legend_position. Valid values are none, top-left, top, top-right, left, right, bottom-left, bottom, and bottom-right. Defaults to top-right.
[temperature_graph]\nlegend_position = \"top-right\"\nBy default, the y-axis is bounded at 100\u00b0C (or the equivalent in the configured temperature unit) and grows automatically if a reading exceeds that. An explicit upper bound can be set with max_temp (uses the same unit as the configured temperature unit). Sensor readings above this value will be drawn off the chart.
[temperature_graph]\nmax_temp = 90.0\nYou can filter which sensors to plot by configuring [temperature_graph.sensor_filter]. This works the same way as the temperature table's filter.
For example, here we are ignoring any sensor that has \"cpu\" or \"wifi\" in its name:
[temperature_graph.sensor_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"cpu\", \"wifi\"]\n\n# Whether to use regex. Defaults to false.\nregex = false\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nThe temperature table widget is configured under [temperature].
You can customize the default sort column (by default, it sorts by temperature sensor name in ascending order). For example, to sort by temperature:
[temperature]\ndefault_sort = \"Temp\"\nYou can also set the sort order by changing temperature.sort_order with \"Ascending\" or \"Descending\". Defaults to \"Ascending\":
[temperature]\nsort_order = \"Ascending\"\nYou can filter out what entries to show by configuring [temperature.sensor_filter]. In particular you can set a list of things to filter with by setting list, and configure how that list is processed with the other options.
For example, here we are ignoring any sensor that has \"cpu\" or \"wifi\" in it.
[temperature.sensor_filter]\n# Whether to ignore any matches. Defaults to true.\nis_list_ignored = true\n\n# A list of filters to try and match.\nlist = [\"cpu\", \"wifi\"]\n\n# Whether to use regex. Defaults to false.\nregex = false\n\n# Whether to be case-sensitive. Defaults to false.\ncase_sensitive = false\n\n# Whether to require matching the whole word. Defaults to false.\nwhole_word = false\nREADME.md, changelog, etc.)There are a few areas where documentation changes are often needed:
README.mdCHANGELOG.mdFork the repository to make changes in.
Where you're adding documentation will probably affect what you need to do:
README.md or CHANGELOG.md For changes to README.md and CHANGELOG.md, just follow the formatting provided and use any editor.
Generally, changes to CHANGELOG.md will be handled by a maintainer, and the contents of the file should follow the Keep a Changelog format, as well as link to the relevant PR or issues.
For changes to the help menu, try to refer to the existing code within src/constants.rs on how the help menu is generated.
For changes to the extended documentation, you'll probably want at least Python 3.11 (older and newer versions should be fine), MkDocs, Material for MkDocs, mdx_truly_sane_lists, and optionally Mike installed. These can help with validating your changes locally.
You can do so through pip or your system's package managers. If you use pip, you can use venv to cleanly install the documentation dependencies:
# Change directories to the documentation.\ncd docs/\n\n # Create venv, install the dependencies, and serve the page.\n./serve.sh\nThis will serve a local version of the docs that you can open on your browser. It will update as you make changes.
Once you have your documentation changes done, submit it as a pull request. For more information regarding that, refer to Issues, Pull Requests, and Discussions.
Documentation should never be generated by AI. The point of documentation in this project is to communicate things between humans. Follow the AI policy; changes that do not follow it may be closed or hidden.
"},{"location":"contribution/issues-and-pull-requests/","title":"Issues, Pull Requests, and Discussions","text":""},{"location":"contribution/issues-and-pull-requests/#discussions","title":"Discussions","text":"Discussions are open in the repo. As for the difference between discussions and issues:
When filing a bug report, please use the bug report template and fill in as much as you can. It is incredibly difficult for a maintainer to fix a bug when it cannot be reproduced, and giving as much detail as possible generally helps to make it easier to reproduce the problem!
"},{"location":"contribution/issues-and-pull-requests/#feature-requests","title":"Feature requests","text":"Please use the feature request template and fill it out. Remember to give details about what the feature is along with why you think this suggestion will be useful.
Also, please check whether an existing issue has covered your specific feature request!
"},{"location":"contribution/issues-and-pull-requests/#pull-requests","title":"Pull requests","text":"The expected workflow for a pull request is:
main branch. Please fill out the pull request template and checklist.Some other things related to making changes or pull requests:
Note that any contribution, whether it be opening an issue, PRs, or commenting, should follow the AI policy. Contributions that do not follow this may be closed/hidden.
"},{"location":"contribution/packaging-and-distribution/","title":"Packaging and Distribution","text":"Package maintainers are always welcome and appreciated! Here's some info on how one can help with package distribution and bottom.
"},{"location":"contribution/packaging-and-distribution/#pre-built-binaries","title":"Pre-built binaries","text":"The latest stable release can be found here, where you can find pre-built binaries in either a tar.gz or zip format. Binaries here also include automatically generated shell completion files for zsh, bash, fish, and Powershell, which you may want to also install during the packaging process.
You can also find a nightly build in the releases page, built every day at 00:00 UTC off of the main branch.
In both cases, we use GitHub Actions to create our release binaries. build_releases.yml contains the GitHub Action workflow for this.
If you want to manually build bottom rather than distributing a pre-built binary, you'll need the most recent version of stable Rust, which you can get with:
rustup update stable\nYou'll then want to build with:
cargo build --release --locked\nbottom uses a build.rs script to automatically generate a manpage and shell completions for the following shells:
If you want to generate manpages and/or completion files, set the BTM_GENERATE env var to a non-empty value. For example, run something like this:
BTM_GENERATE=true cargo build --release --locked\nThis will automatically generate completion and manpage files in target/tmp/bottom/. If you wish to regenerate the files, modify/delete either these files or set BTM_GENERATE to some other non-empty value to retrigger the build script.
You may override the default directories used to generate both completion and manpage files by specifying the COMPLETION_DIR and MANPAGE_DIR environment variables respectively.
For more information, you may want to look at either the build.rs file or the binary build CI workflow.
Once you've finished your installation source, if you want to mention it in the main bottom repo, fork the repo and add the installation method and any details to the README.md file under the Installation section, as well as a corresponding table of contents entry. Once that's done, open a pull request - these will usually be approved of very quickly.
You can find more info on the contribution process here.
"},{"location":"contribution/development/build_process/","title":"Build Process","text":"Warning
This section is currently somewhat WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/build_process/#overview","title":"Overview","text":"bottom manages its own binary builds for nightly and stable release purposes. The core build workflow is handled by build_releases.yml, called by a wrapper workflow for nightly and stable releases. Builds take place via GitHub Actions.
The main things built are:
.deb package for Debian and its derivativesThis documentation gives a high-level overview of the build process for each part. For the most up-to-date and detailed reference, definitely refer back to the build_releases.yml file.
Binaries are built currently for various targets. Note that not all of these are officially supported. The following general steps are performed:
Build a release build with:
--features deploy, which enables only crates needed for release builds.--locked to lock the dependency versions.The following env variables set:
BTM_GENERATE: trueCOMPLETION_DIR: \"target/tmp/bottom/completion/\"MANPAGE_DIR: \"target/tmp/bottom/manpage/\"These generate the manpages and shell completions (see Packaging for some more information).
Some builds use cross to do cross-compilation builds for architectures otherwise not natively supported by the runner.
This builds a full Windows installer using cargo-wix and WiX Toolset 4. This requires some setup beforehand with some dependencies:
After that, cargo wix takes care of the rest.
.deb","text":"Currently, .deb files are built for x86 and ARM architectures (armv7, aarch64). This is handled by cargo-deb.
cargo-deb..deb file.There are additional checks via dpkg to ensure the architecture is correctly set.
Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/deploy_process/#overview","title":"Overview","text":"bottom currently has two main deploy processes to worry about:
Stable: a stable deployment, triggered manually or upon creation of a valid tag. This is a GitHub action that builds binary/installer files and uploads them to a new GitHub release.
Furthermore, this workflow does not handle the following deployments, which must be manually handled:
This is, for the most part, automatic, though it can also be used as a way of testing build workflow changes and seeing if binaries can be successfully built at all against all the targets we want to build for.
If one does not want to actually update the nightly release, and just want to test the general builds and workflow, one can run the workflow manually on a branch of choice with \"mock\" set as the parameter. Changing it to anything else will trigger a non-mock run.
"},{"location":"contribution/development/deploy_process/#stable","title":"Stable","text":"This can be manually triggered, though the general use-case is setting a tag of the form x.y.z (after checking everything is good, of course). For example:
git tag 0.6.9 && git push origin 0.6.9\nThis will automatically trigger the deployment workflow, and create a draft release with the files uploaded. One still needs to fill in the details and release it.
Furthermore, there are some deployments that are handled by maintainers of bottom that this workflow does not automatically finish. These must be manually handled.
"},{"location":"contribution/development/deploy_process/#cratesio","title":"crates.io","text":"Validate everything builds properly and works (you should have done this before releasing though). If good, then deploying on crates.io is as simple as:
cargo publish\nUpon releasing on GitHub, choco-bottom will automatically be updated with a new PR with the correct deployment files for Chocolatey. Check the PR, merge it if it is correct, then pull locally and deploy following the instructions in the README. Make sure to test installation and running at least once before deploying!
If done correctly, there should be a new build on Chocolatey, which will take some time to validate.
"},{"location":"contribution/development/deploy_process/#winget","title":"winget","text":"Sometimes, people will kindly do it, but it can also be manually triggered using winget-bottom.
Note this requires regenerating some secrets.
"},{"location":"contribution/development/dev_env/","title":"Development Environment","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/logging/","title":"Logging","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"contribution/development/testing/","title":"Testing","text":"Warning
This section is currently WIP.
Warning
This section is intended for people who wish to work on/build/distribute bottom, not general users.
"},{"location":"support/official/","title":"Official support","text":"bottom officially supports the following operating systems and corresponding architectures:
x86_64, aarch64)x86_64, i686, aarch64)x86_64, i686)These platforms are tested to work (with caveats, see below) and issues on these platforms will be fixed if possible.
Furthermore, binaries are expected to be built and tested using the most recent version of stable Rust - if you are manually building bottom from the repo/source, then please try that as well.
If you encounter any issues, see the troubleshooting page or open an issue/discussion on GitHub.
"},{"location":"support/unofficial/","title":"Unofficial support","text":"Systems and architectures that aren't officially supported may still work, but there are no guarantees on how much will work. For example, it might only compile, or it might run with bugs/broken features. Furthermore, while it will depend on the problem at the end of the day, issues on unsupported platforms are likely to go unfixed.
Unofficially supported platforms known to compile/work:
If you encounter any issues, see the troubleshooting page or open an issue/discussion on GitHub - though as mentioned, unsupported platforms may go unfixed.
"},{"location":"usage/autocomplete/","title":"Auto-Complete","text":"The release binaries in the releases page are packaged with shell auto-completion files for Bash, Zsh, fish, Powershell, Elvish, Fig, and Nushell. To install them:
btm.bash to $XDG_CONFIG_HOME/bash_completion or /etc/bash_completion.d/._btm to one of your $fpath directories.btm.fish to $HOME/.config/fish/completions/._btm.ps1 to your PowerShell profile.btm.elv.btm.ts.btm.nu.The individual auto-completion files are also included in the stable/nightly releases as completion.tar.gz if needed.
Basic mode is a special layout that removes all of the graphs and provides an interface that resembles (a very stripped-down version of) htop.
Basic mode can be enabled either through a command line flag:
btm -b\n\n# or\n\nbtm --basic\nor through the config:
[flags]\nbasic = true\nIn this mode, widgets that use tables (temperatures, processes, disks, and batteries) are only shown one at a time. One can switch between these widgets either by clicking the arrow buttons or by using the general widget selection shortcuts (for example, Ctrl+Left or H) to switch which widget is shown.
Also note that in this mode, widget expansion and custom layouts are disabled.
"},{"location":"usage/basic-mode/#key-bindings","title":"Key bindings","text":"Basic mode follows the same key bindings as normal, barring widget expansion being disabled, and that the % key while selecting the memory widget toggles between total usage and percentage.
"},{"location":"usage/general-usage/","title":"General Usage","text":"You can run bottom with:
btm\nFor help regarding the command-line options, use:
# For a simple overview of flags\nbtm -h\n\n# For more details\nbtm --help\nYou can also see keybinds and basic usage details in bottom by pressing ?, which will open a help menu.
"},{"location":"usage/general-usage/#features","title":"Features","text":""},{"location":"usage/general-usage/#expansion","title":"Expansion","text":"By default, bottom is somewhat like a dashboard - a bunch of different widgets, all showing different things, and they all cram together to fit into one terminal.
If you instead just want to see one widget - maybe you want to look at a graph in more detail, for example - you can \"expand\" the currently selected widget using the e key, which will hide all other widgets and make that widget take up all available terminal space.
You can leave this state by either pressing e again or pressing Esc.
"},{"location":"usage/general-usage/#widget-selection","title":"Widget selection","text":"To allow for widget-specific keybindings and expansion, there is the idea of widget selection in bottom, where you can focus on a specific widget to work with it. This can be done with the mouse (just click on the widget of interest) or keyboard (ex: Ctrl+Direction, see Key bindings for alternatives).
"},{"location":"usage/general-usage/#key-bindings","title":"Key bindings","text":"These are global or common keyboard shortcuts for the application, which you can see in-app through the ? shortcut. Note that key bindings are generally case-sensitive.
Binding Action q, Q, Ctrl+c Quit Esc Close dialog windows, search, widgets, or exit expanded mode Ctrl+r Reset display and any collected data f Freeze/unfreeze updating with new data ? Open help menu e Toggle expanding the currently selected widget Ctrl+Up Shift+Up K W Select the widget above Ctrl+Down Shift+Down J S Select the widget below Ctrl+Left Shift+Left H A Select the widget on the left Ctrl+Right Shift+Right L D Select the widget on the right Up , k Move up within a widget Down , j Move down within a widget Left h Alt+h Move left within a widget Right l Alt+l Move right within a widget g+g , Home Jump to the first entry G , End Jump to the last entry Page Up , Page Down Scroll up/down a table by a page Ctrl+u Scroll up a table by half a page Ctrl+d Scroll down a table by half a page"},{"location":"usage/general-usage/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects the widget"},{"location":"usage/widgets/battery/","title":"Battery Widget","text":"Warning
The battery features are unavailable if the binary is compiled with the battery feature disabled or if there are no batteries on the system!
The battery widget provides information about batteries on the system.
The battery widget can be enabled through either the --battery flag, the battery = true option in a config file, or specifying the widget in a custom layout.
The following data is displayed for batteries:
The battery widget also supports devices with multiple batteries, and you can switch between them using the keyboard or the mouse.
"},{"location":"usage/widgets/battery/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action Left h Alt+h Moves to the battery entry to the left of the current one Right l Alt+l Moves to the battery entry to the right of the current one"},{"location":"usage/widgets/battery/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects a battery entry"},{"location":"usage/widgets/cpu-graph/","title":"CPU Widget","text":"The CPU widget displays a visual representation of CPU usage over a time range.
"},{"location":"usage/widgets/cpu-graph/#features","title":"Features","text":"The CPU widget is composed of two parts: the graph and the legend:
Users can scroll through the legend using either the keyboard or mouse to select which entry to display on the graph. The \"All\" option shows every entry at the same time, though this may get a bit hard to follow if you have a large number of cores/threads.
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
"},{"location":"usage/widgets/cpu-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
"},{"location":"usage/widgets/cpu-graph/#graph","title":"Graph","text":"Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/cpu-graph/#legend","title":"Legend","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the legend G , End Jump to the last entry in the legend"},{"location":"usage/widgets/cpu-graph/#mouse-bindings","title":"Mouse bindings","text":""},{"location":"usage/widgets/cpu-graph/#graph_1","title":"Graph","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/cpu-graph/#legend_1","title":"Legend","text":"Binding Action Scroll Scroll through options to display in the graph Left Button Selects a CPU thread/average to show in the graph"},{"location":"usage/widgets/disk-io-graph/","title":"Disk I/O Graph Widget","text":"The disk I/O graph widget shows read and write rates over time.
"},{"location":"usage/widgets/disk-io-graph/#features","title":"Features","text":"The legend displays the current reads and writes per second in bytes for each mount/disk item. The y-axis automatically scales based on shown read/write values. By default this is in a linear scale, but it can be switched to a logarithmic scale through the config file.
The displayed time range can be adjusted through either the keyboard or mouse.
"},{"location":"usage/widgets/disk-io-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/disk-io-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/disk/","title":"Disk Widget","text":"The disk widget provides a table of useful disk and partition information, like I/O per second and total usage.
"},{"location":"usage/widgets/disk/#features","title":"Features","text":"The disk widget provides the following information:
Note that key bindings are generally case-sensitive.
Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table d Sort by disk, press again to reverse sorting order m Sort by mount, press again to reverse sorting order u Sort by amount used, press again to reverse sorting order n Sort by amount free, press again to reverse sorting order t Sort by total space available, press again to reverse sorting order p Sort by percentage used, press again to reverse sorting order r Sort by read rate, press again to reverse sorting order w Sort by write rate, press again to reverse sorting order"},{"location":"usage/widgets/disk/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects an entry in the table"},{"location":"usage/widgets/memory-graph/","title":"Memory Widget","text":"The memory widget provides a visual representation of RAM and swap usage over time.
"},{"location":"usage/widgets/memory-graph/#features","title":"Features","text":"The legend displays the current usage in terms of percentage and actual usage in binary units (KiB, MiB, GiB, etc.). If the total RAM or swap available is 0, then it is automatically hidden from the legend and graph.
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
This widget can also be configured to display Nvidia and AMD GPU memory usage (--disable_gpu on Linux/Windows to disable) or cache memory usage (--enable_cache_memory).
Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/memory-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/memory-graph/#how-are-memory-values-determined","title":"How are memory values determined?","text":""},{"location":"usage/widgets/memory-graph/#linux","title":"Linux","text":"Memory usage is calculated using the following formula based on values from /proc/meminfo (based on htop's implementation):
MemTotal - MemFree - Buffers - (Cached + SReclaimable - Shmem)\nYou can find more info on /proc/meminfo and its fields here.
In Windows, we calculate swap by querying Get-Counter \"\\Paging File(*)\\% Usage\". This is also what some libraries like psutil use. However, note there are also a few other valid methods of representing \"swap\" in Windows (e.g. using GetPerformanceInfo), which all slightly don't match.
The network widget provides a visual representation of network input and output per second, as well as noting the total amount received and transmitted.
"},{"location":"usage/widgets/network-graph/#features","title":"Features","text":"The legend displays the current reads and writes per second in bits, as well as the total amount read/written.
The y-axis automatically scales based on shown read/write values, and by default, is a linear scale based on base-10 units (e.x. kilobit, gigabit, etc.). Through configuration, the read/write per second unit can be changed to bytes, while the y-axis can be changed to a log scale and/or use base-2 units (e.x. kibibit, gibibit, etc.).
One can also adjust the displayed time range through either the keyboard or mouse, with a range of 30s to 600s.
"},{"location":"usage/widgets/network-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/network-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/process/","title":"Process Widget","text":"The process widget displays a table containing information regarding a running process, along with sorting, searching, and process control features.
"},{"location":"usage/widgets/process/#features","title":"Features","text":"The process widget has three main components:
By default, the main process table displays the following information for each process:
Process uptime
Info
On Windows, the I/O counters will report all reads/writes, not just disk. See here for more details.
With the feature flag (--disable_gpu on Linux/Windows to disable) and gpu process columns enabled in the configuration:
See the processes configuration page on how to customize which columns are shown.
"},{"location":"usage/widgets/process/#sorting","title":"Sorting","text":"The table can be sorted by clicking on the table headers, which will either sort the table by that column, or if already sorting by that column, reverse the sorting order.
Alternatively, one can sort using the sort menu sub-widget, which is brought up using s or F6, and can be controlled by arrow keys or the mouse.
"},{"location":"usage/widgets/process/#grouping","title":"Grouping","text":"Pressing Tab in the table will group entries with the same name together. The PID column will be replaced with the number of entries in each group, and usage is added together when displayed.
Info
Note that the process state and user columns are disabled in this mode.
Info
Note that if tree mode is also active, processes cannot be grouped together due to the behaviour of the two modes somewhat clashing. This also reflects with default modes like group_processes.
Pressing d+d or F9 will allow you to terminate the currently selected process/process group. On Unix-like operating systems, you are also able to control which specific signals to send (e.g. SIGKILL, SIGTERM).
If you're on Windows, or if the disable_advanced_kill flag is set in the options or command-line (only available on Linux, macOS, and FreeBSD), then a simpler termination screen with just yes or no options will be shown.
Pressing t or F5 in the table toggles tree mode in the process widget, displaying processes in regard to their parent-child process relationships.
A process in tree mode can also be \"collapsed\", hiding its children and any descendants, using the either the -, +, or Left keys, or clicking on an entry. It can be expanded by using the -, +, or Right keys, or by clicking on the entry again. The Space key can also be used to toggle between the collapsed and expanded states.
Info
Note that if tree mode is active, processes cannot be grouped together due to the behaviour of the two modes somewhat clashing. This also reflects with default modes like group_processes.
You can show the full command instead of just the process name by pressing P.
"},{"location":"usage/widgets/process/#search","title":"Search","text":"Pressing / or Ctrl+f will open up the search sub-widget. By default, just typing in something will search by the process name.
This search can be further enhanced by matching by case, matching the entire word, or by regex.
We are able to also search for multiple things/conditions.
And if our search uses a keyword, we need to use quotation marks around the term to properly search it.
Lastly, we can refine our search even further based on the other columns, like PID, CPU usage, etc., as well as grouping together conditions.
You can also paste search queries (e.g. Shift+Ins, Ctrl+Shift+v).
"},{"location":"usage/widgets/process/#keywords","title":"Keywords","text":"Note all keywords are case-insensitive. To search for a process/command that collides with a keyword, surround the term with quotes (e.x. \"cpu\").
btm Matches by process or command name; supports regex pid pid=1044 Matches by PID; supports regex cpu cpu% cpu > 0.5 Matches the CPU column; supports comparison operators memb memb > 1000 b Matches the memory column in terms of bytes; supports comparison operators mem mem% mem < 0.5 Matches the memory column in terms of percent; supports comparison operators read r/s rps read = 1 mb Matches the read/s column in terms of bytes; supports comparison operators write w/s wps write >= 1 kb Matches the write/s column in terms of bytes; supports comparison operators tread t.read tread <= 1024 gb Matches the total read column in terms of bytes; supports comparison operators twrite t.write twrite > 1024 tb Matches the total write column in terms of bytes; supports comparison operators user user=root Matches by user; supports regex state state=running Matches by state; supports regex () (<COND 1> AND <COND 2>) OR <COND 3> Group together a condition gmem gmem > 1000 b Matches the gpu memory column in terms of bytes; supports comparison operators gmem% gmem% < 0.5 Matches the gpu memory column in terms of percent; supports comparison operators gpu% gpu% > 0 Matches the gpu usage column in terms of percent; supports comparison operators"},{"location":"usage/widgets/process/#comparison-operators","title":"Comparison operators","text":"Keywords Description = Checks if the values are equal != Checks if the values are not equal > Checks if the left value is strictly greater than the right < Checks if the left value is strictly less than the right >= Checks if the left value is greater than or equal to the right <= Checks if the left value is less than or equal to the right"},{"location":"usage/widgets/process/#logical-operators","title":"Logical operators","text":"Note all operators are case-insensitive, and the and operator takes precedence over the or operator.
and && <Space> <COND 1> and <COND 2> <COND 1> && <COND 2> <COND 1> <COND 2> Requires both conditions to be true to match or || <COND 1> or <COND 2> <COND 1> || <COND 2> Requires at least one condition to be true to match ! !<COND> !(<COND 1> or <COND 2>) Inverts the following condition or group ! is reserved as an operator, so it cannot appear bare as a value. To match a literal ! in a name or string field, quote it \u2014 e.g. \"foo!\" or user = \"!\". A bare ! with nothing parseable after it (such as user = ! or ! on its own) is rejected.
All units are case-insensitive.
Keywords DescriptionB Bytes KB Kilobytes MB Megabytes GB Gigabytes TB Terabytes KiB Kibibytes MiB Mebibytes GiB Gibibytes TiB Tebibytes"},{"location":"usage/widgets/process/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
"},{"location":"usage/widgets/process/#process-table","title":"Process table","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table d+d , F9 Send a kill signal to the selected process c Sort by CPU usage, press again to reverse sorting order m Sort by memory usage, press again to reverse sorting order p Sort by PID name, press again to reverse sorting order n Sort by process name, press again to reverse sorting order Tab Toggle grouping processes with the same name P Toggle between showing the full command or just the process name Ctrl+f , / Toggle showing the search sub-widget s , F6, Del (Fn+Del on macOS) Toggle showing the sort sub-widget I Invert the current sort % Toggle between values and percentages for memory usage t , F5 Toggle tree mode M Sort by gpu memory usage, press again to reverse sorting order C Sort by gpu usage, press again to reverse sorting order z Toggle the hiding of kernel threads"},{"location":"usage/widgets/process/#sort-sub-widget","title":"Sort sub-widget","text":"Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table Esc Close the sort sub-widget Enter Sorts the corresponding process table"},{"location":"usage/widgets/process/#search-sub-widget","title":"Search sub-widget","text":"Binding Action Left h Alt+h Moves the cursor left Right l Alt+l Moves the cursor right Esc Close the search widget (retains the filter) Ctrl+a Skip to the start of the search query Ctrl+e Skip to the end of the search query Ctrl+u Clear the current search query Ctrl+w Delete a word behind the cursor Ctrl+h Delete the character behind the cursor Backspace Delete the character behind the cursor Del (Fn+Del on macOS) Delete the character at the cursor Alt+c , F1 Toggle matching case Alt+w , F2 Toggle matching the entire word Alt+r , F3 Toggle using regex"},{"location":"usage/widgets/process/#mouse-bindings","title":"Mouse bindings","text":""},{"location":"usage/widgets/process/#process-table_1","title":"Process table","text":"Binding Action Scroll Selects a CPU thread/average to show in the graph Left Button Table header: Sorts/reverse sorts the table by the column Table entry: Selects an entry in the table, if in tree mode, collapses/expands the entry's children"},{"location":"usage/widgets/process/#sort-sub-widget_1","title":"Sort sub-widget","text":"Binding Action Left Button Selects an entry in the table"},{"location":"usage/widgets/temperature-graph/","title":"Temperature Graph Widget","text":"The temperature graph widget provides temperature readings over time.
"},{"location":"usage/widgets/temperature-graph/#features","title":"Features","text":"Each detected sensor is drawn as its own line. The y-axis is in the configured temperature unit (Celsius by default; see the --celsius/--fahrenheit/--kelvin arguments or flags.temperature_type config option).
By default the y-axis is bounded at 100\u00b0C (or the equivalent in the configured unit) and grows automatically. An upper bound can also be set explicitly via the config file.
The displayed time range can be adjusted through either the keyboard or mouse.
"},{"location":"usage/widgets/temperature-graph/#key-bindings","title":"Key bindings","text":"Note that key bindings are generally case-sensitive.
Binding Action + Zoom in on chart (decrease time range) - Zoom out on chart (increase time range) = Reset zoom"},{"location":"usage/widgets/temperature-graph/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Scroll Scrolling up or down zooms in or out of the graph respectively"},{"location":"usage/widgets/temperature-table/","title":"Temperature Widget","text":"The temperature widget provides a table of temperature sensors and their current temperature.
"},{"location":"usage/widgets/temperature-table/#features","title":"Features","text":"The temperature widget provides the sensor name as well as its current temperature.
This widget can also be configured to display Nvidia and AMD GPU temperatures (--disable_gpu on Linux/Windows to disable).
Note that key bindings are generally case-sensitive.
Binding Action Up , k Move up within a widget Down , j Move down within a widget g+g , Home Jump to the first entry in the table G , End Jump to the last entry in the table t Sort by temperature, press again to reverse sorting order s Sort by sensor name, press again to reverse sorting order"},{"location":"usage/widgets/temperature-table/#mouse-bindings","title":"Mouse bindings","text":"Binding Action Left Button Selects an entry in the table"}]} \ No newline at end of file