1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
\r
2 "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
\r
3 <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
\r
5 <link rel="icon" type="image/x-icon" href="/favicon.ico">
\r
6 <meta http-equiv="Content-Type" content="application/xhtml+xml; charset=UTF-8" />
\r
7 <meta name="generator" content="AsciiDoc 8.6.9" />
\r
8 <title>i3: Hacking i3: How To</title>
\r
9 <link rel="stylesheet" href="/css/style.css" type="text/css" />
\r
10 <link rel="stylesheet" href="/css/xhtml11.css" type="text/css" />
\r
11 <script type="text/javascript">
\r
13 document.addEventListener("DOMContentLoaded", function(){asciidoc.footnotes(); asciidoc.toc(2);}, false);
\r
16 <script type="text/javascript" src="/js/asciidoc-xhtml11.js"></script>
\r
18 <body class="article">
\r
21 <a href="/"><h1 id="title">i3 - improved tiling WM</h1></a>
\r
23 <li><a style="border-bottom: 2px solid #fff" href="/docs">Docs</a></li>
\r
24 <li><a href="/screenshots">Screens</a></li>
\r
25 <li><a href="https://www.reddit.com/r/i3wm/">FAQ</a></li>
\r
26 <li><a href="/contact">Contact</a></li>
\r
27 <li><a href="https://github.com/i3/i3/issues">Bugs</a></li>
\r
29 <br style="clear: both">
\r
32 <h1>Hacking i3: How To</h1>
\r
33 <span id="author">Michael Stapelberg</span><br />
\r
34 <span id="email"><tt><<a href="mailto:michael@i3wm.org">michael@i3wm.org</a>></tt></span><br />
\r
35 <span id="revdate">February 2013</span>
\r
37 <div id="toctitle">Table of Contents</div>
38 <noscript><p><b>JavaScript must be enabled in your browser to display the table of contents.</b></p></noscript>
42 <div class="sectionbody">
\r
43 <div class="paragraph"><p>This document is intended to be the first thing you read before looking and/or
\r
44 touching i3’s source code. It should contain all important information to help
\r
45 you understand why things are like they are. If it does not mention something
\r
46 you find necessary, please do not hesitate to contact me.</p></div>
\r
50 <h2 id="_window_managers">1. Window Managers</h2>
\r
51 <div class="sectionbody">
\r
52 <div class="paragraph"><p>A window manager is not necessarily needed to run X, but it is usually used in
\r
53 combination with X to facilitate some things. The window manager’s job is to
\r
54 take care of the placement of windows, to provide the user with some mechanisms
\r
55 to change the position/size of windows and to communicate with clients to a
\r
56 certain extent (for example handle fullscreen requests of clients such as
\r
58 <div class="paragraph"><p>There are no different contexts in which X11 clients run, so a window manager
\r
59 is just another client, like all other X11 applications. However, it handles
\r
60 some events which normal clients usually don’t handle.</p></div>
\r
61 <div class="paragraph"><p>In the case of i3, the tasks (and order of them) are the following:</p></div>
\r
62 <div class="olist arabic"><ol class="arabic">
\r
65 Grab the key bindings (events will be sent upon keypress/keyrelease)
\r
70 Iterate through all existing windows (if the window manager is not started as
\r
71 the first client of X) and manage them (reparent them, create window
\r
77 When new windows are created, manage them
\r
82 Handle the client’s <tt>_WM_STATE</tt> property, but only <tt>_WM_STATE_FULLSCREEN</tt> and
\r
83 <tt>_NET_WM_STATE_DEMANDS_ATTENTION</tt>
\r
88 Handle the client’s <tt>WM_NAME</tt> property
\r
93 Handle the client’s size hints to display them proportionally
\r
98 Handle the client’s urgency hint
\r
103 Handle enter notifications (focus follows mouse)
\r
108 Handle button (as in mouse buttons) presses for focus/raise on click
\r
113 Handle expose events to re-draw own windows such as decorations
\r
118 React to the user’s commands: Change focus, Move windows, Switch workspaces,
\r
119 Change the layout mode of a container (default/stacking/tabbed), start a new
\r
120 application, restart the window manager
\r
124 <div class="paragraph"><p>In the following chapters, each of these tasks and their implementation details
\r
125 will be discussed.</p></div>
\r
126 <div class="sect2">
\r
127 <h3 id="_tiling_window_managers">1.1. Tiling window managers</h3>
\r
128 <div class="paragraph"><p>Traditionally, there are two approaches to managing windows: The most common
\r
129 one nowadays is floating, which means the user can freely move/resize the
\r
130 windows. The other approach is called tiling, which means that your window
\r
131 manager distributes windows to use as much space as possible while not
\r
132 overlapping each other.</p></div>
\r
133 <div class="paragraph"><p>The idea behind tiling is that you should not need to waste your time
\r
134 moving/resizing windows while you usually want to get some work done. After
\r
135 all, most users sooner or later tend to lay out their windows in a way which
\r
136 corresponds to tiling or stacking mode in i3. Therefore, why not let i3 do this
\r
137 for you? Certainly, it’s faster than you could ever do it.</p></div>
\r
138 <div class="paragraph"><p>The problem with most tiling window managers is that they are too inflexible.
\r
139 In my opinion, a window manager is just another tool, and similar to vim which
\r
140 can edit all kinds of text files (like source code, HTML, …) and is not limited
\r
141 to a specific file type, a window manager should not limit itself to a certain
\r
142 layout (like dwm, awesome, …) but provide mechanisms for you to easily create
\r
143 the layout you need at the moment.</p></div>
\r
145 <div class="sect2">
\r
146 <h3 id="_the_layout_tree">1.2. The layout tree</h3>
\r
147 <div class="paragraph"><p>The data structure which i3 uses to keep track of your windows is a tree. Every
\r
148 node in the tree is a container (type <tt>Con</tt>). Some containers represent actual
\r
149 windows (every container with a <tt>window != NULL</tt>), some represent split
\r
150 containers and a few have special purposes: they represent workspaces, outputs
\r
151 (like VGA1, LVDS1, …) or the X11 root window.</p></div>
\r
152 <div class="paragraph"><p>So, when you open a terminal and immediately open another one, they reside in
\r
153 the same split container, which uses the default layout. In case of an empty
\r
154 workspace, the split container we are talking about is the workspace.</p></div>
\r
155 <div class="paragraph"><p>To get an impression of how different layouts are represented, just play around
\r
156 and look at the data structures — they are exposed as a JSON hash. See
\r
157 <a href="http://i3wm.org/docs/ipc.html#_tree_reply">http://i3wm.org/docs/ipc.html#_tree_reply</a> for documentation on that and an
\r
162 <div class="sect1">
\r
163 <h2 id="_files">2. Files</h2>
\r
164 <div class="sectionbody">
\r
165 <div class="dlist"><dl>
\r
166 <dt class="hdlist1">
\r
167 include/atoms.xmacro
\r
171 A file containing all X11 atoms which i3 uses. This file will be included
\r
172 various times (for defining, requesting and receiving the atoms), each time
\r
173 with a different definition of xmacro().
\r
176 <dt class="hdlist1">
\r
181 Contains data definitions used by nearly all files. You really need to read
\r
185 <dt class="hdlist1">
\r
190 Contains forward definitions for all public functions, as well as
\r
191 doxygen-compatible comments (so if you want to get a bit more of the big
\r
192 picture, either browse all header files or use doxygen if you prefer that).
\r
195 <dt class="hdlist1">
\r
196 src/config_parser.c
\r
200 Contains a custom configuration parser. See src/command_parser.c for rationale
\r
201 on why we use a custom parser.
\r
204 <dt class="hdlist1">
\r
209 Contains all functions which handle mouse button clicks (right mouse button
\r
210 clicks initiate resizing and thus are relatively complex).
\r
213 <dt class="hdlist1">
\r
214 src/command_parser.c
\r
218 Contains a hand-written parser to parse commands (commands are what
\r
219 you bind on keys and what you can send to i3 using the IPC interface, like
\r
220 <em>move left</em> or <em>workspace 4</em>).
\r
223 <dt class="hdlist1">
\r
228 Contains all functions which deal with containers directly (creating
\r
229 containers, searching containers, getting specific properties from containers,
\r
233 <dt class="hdlist1">
\r
238 Contains all functions handling the configuration file (calling the parser
\r
239 src/config_parser.c) with the correct path, switching key bindings mode).
\r
242 <dt class="hdlist1">
\r
247 Functions to get/set certain EWMH properties easily.
\r
250 <dt class="hdlist1">
\r
255 Contains functions for floating mode (mostly resizing/dragging).
\r
258 <dt class="hdlist1">
\r
263 Contains all handlers for all kinds of X events (new window title, new hints,
\r
264 unmapping, key presses, button presses, …).
\r
267 <dt class="hdlist1">
\r
272 Contains code for the IPC interface.
\r
275 <dt class="hdlist1">
\r
280 Contains code for loading layouts from JSON files.
\r
283 <dt class="hdlist1">
\r
288 Contains the logging functions.
\r
291 <dt class="hdlist1">
\r
296 Initializes the window manager.
\r
299 <dt class="hdlist1">
\r
304 Looks at existing or new windows and decides whether to manage them. If so, it
\r
305 reparents the window and inserts it into our data structures.
\r
308 <dt class="hdlist1">
\r
313 A "match" is a data structure which acts like a mask or expression to match
\r
314 certain windows or not. For example, when using commands, you can specify a
\r
315 command like this: [title="<strong>Firefox</strong>"] kill. The title member of the match
\r
316 data structure will then be filled and i3 will check each window using
\r
317 match_matches_window() to find the windows affected by this command.
\r
320 <dt class="hdlist1">
\r
325 Contains code to move a container in a specific direction.
\r
328 <dt class="hdlist1">
\r
333 Functions to handle CT_OUTPUT cons.
\r
336 <dt class="hdlist1">
\r
341 The RandR API is used to get (and re-query) the configured outputs (monitors,
\r
345 <dt class="hdlist1">
\r
350 Renders the tree data structure by assigning coordinates to every node. These
\r
351 values will later be pushed to X11 in <tt>src/x.c</tt>.
\r
354 <dt class="hdlist1">
\r
359 Contains the functions to resize containers.
\r
362 <dt class="hdlist1">
\r
363 src/restore_layout.c
\r
367 Everything for restored containers that is not pure state parsing (which can be
\r
368 found in load_layout.c).
\r
371 <dt class="hdlist1">
\r
376 Handles <tt>SIGSEGV</tt>, <tt>SIGABRT</tt> and <tt>SIGFPE</tt> by showing a dialog that i3 crashed.
\r
377 You can chose to let it dump core, to restart it in-place or to restart it
\r
378 in-place but forget about the layout.
\r
381 <dt class="hdlist1">
\r
386 Contains functions which open or close containers in the tree, change focus or
\r
387 cleanup ("flatten") the tree. See also <tt>src/move.c</tt> for another similar
\r
388 function, which was moved into its own file because it is so long.
\r
391 <dt class="hdlist1">
\r
396 Contains useful functions which are not really dependent on anything.
\r
399 <dt class="hdlist1">
\r
404 Handlers to update X11 window properties like <tt>WM_CLASS</tt>, <tt>_NET_WM_NAME</tt>,
\r
405 <tt>CLIENT_LEADER</tt>, etc.
\r
408 <dt class="hdlist1">
\r
413 Contains all functions related to workspaces (displaying, hiding, renaming…)
\r
416 <dt class="hdlist1">
\r
421 Transfers our in-memory tree (see <tt>src/render.c</tt>) to X11.
\r
424 <dt class="hdlist1">
\r
429 Contains wrappers to use xcb more easily.
\r
432 <dt class="hdlist1">
\r
437 XCursor functions (for cursor themes).
\r
440 <dt class="hdlist1">
\r
445 Legacy support for Xinerama. See <tt>src/randr.c</tt> for the preferred API.
\r
451 <div class="sect1">
\r
452 <h2 id="_data_structures">3. Data structures</h2>
\r
453 <div class="sectionbody">
\r
454 <div class="paragraph"><p>See include/data.h for documented data structures. The most important ones are
\r
455 explained right here.</p></div>
\r
456 <div class="paragraph"><p>So, the hierarchy is:</p></div>
\r
457 <div class="olist arabic"><ol class="arabic">
\r
460 <strong>X11 root window</strong>, the root container
\r
465 <strong>Output container</strong> (LVDS1 in this example)
\r
470 <strong>Content container</strong> (there are also containers for dock windows)
\r
475 <strong>Workspaces</strong> (Workspace 1 in this example, with horizontal orientation)
\r
480 <strong>Split container</strong> (vertically split)
\r
485 <strong>X11 window containers</strong>
\r
489 <div class="paragraph"><p>The data type is <tt>Con</tt>, in all cases.</p></div>
\r
490 <div class="sect2">
\r
491 <h3 id="_x11_root_window">3.1. X11 root window</h3>
\r
492 <div class="paragraph"><p>The X11 root window is a single window per X11 display (a display is identified
\r
493 by <tt>:0</tt> or <tt>:1</tt> etc.). The root window is what you draw your background image
\r
494 on. It spans all the available outputs, e.g. <tt>VGA1</tt> is a specific part of the
\r
495 root window and <tt>LVDS1</tt> is a specific part of the root window.</p></div>
\r
497 <div class="sect2">
\r
498 <h3 id="_output_container">3.2. Output container</h3>
\r
499 <div class="paragraph"><p>Every active output obtained through RandR is represented by one output
\r
500 container. Outputs are considered active when a mode is configured (meaning
\r
501 something is actually displayed on the output) and the output is not a clone.</p></div>
\r
502 <div class="paragraph"><p>For example, if your notebook has a screen resolution of 1280x800 px and you
\r
503 connect a video projector with a resolution of 1024x768 px, set it up in clone
\r
504 mode (<tt>xrandr --output VGA1 --mode 1024x768 --same-as LVDS1</tt>), i3 will
\r
505 reduce the resolution to the lowest common resolution and disable one of the
\r
506 cloned outputs afterwards.</p></div>
\r
507 <div class="paragraph"><p>However, if you configure it using <tt>xrandr --output VGA1 --mode 1024x768
\r
508 --right-of LVDS1</tt>, i3 will set both outputs active. For each output, a new
\r
509 workspace will be assigned. New workspaces are created on the output you are
\r
510 currently on.</p></div>
\r
512 <div class="sect2">
\r
513 <h3 id="_content_container">3.3. Content container</h3>
\r
514 <div class="paragraph"><p>Each output has multiple children. Two of them are dock containers which hold
\r
515 dock clients. The other one is the content container, which holds the actual
\r
516 content (workspaces) of this output.</p></div>
\r
518 <div class="sect2">
\r
519 <h3 id="_workspace">3.4. Workspace</h3>
\r
520 <div class="paragraph"><p>A workspace is identified by its name. Basically, you could think of
\r
521 workspaces as different desks in your office, if you like the desktop
\r
522 metaphor. They just contain different sets of windows and are completely
\r
523 separate of each other. Other window managers also call this “Virtual
\r
524 desktops”.</p></div>
\r
526 <div class="sect2">
\r
527 <h3 id="_split_container">3.5. Split container</h3>
\r
528 <div class="paragraph"><p>A split container is a container which holds an arbitrary amount of split
\r
529 containers or X11 window containers. It has an orientation (horizontal or
\r
530 vertical) and a layout.</p></div>
\r
531 <div class="paragraph"><p>Split containers (and X11 window containers, which are a subtype of split
\r
532 containers) can have different border styles.</p></div>
\r
534 <div class="sect2">
\r
535 <h3 id="_x11_window_container">3.6. X11 window container</h3>
\r
536 <div class="paragraph"><p>An X11 window container holds exactly one X11 window. These are the leaf nodes
\r
537 of the layout tree, they cannot have any children.</p></div>
\r
541 <div class="sect1">
\r
542 <h2 id="_list_queue_macros">4. List/queue macros</h2>
\r
543 <div class="sectionbody">
\r
544 <div class="paragraph"><p>i3 makes heavy use of the list macros defined in BSD operating systems. To
\r
545 ensure that the operating system on which i3 is compiled has all the expected
\r
546 features, i3 comes with <tt>include/queue.h</tt>. On BSD systems, you can use man
\r
547 <tt>queue(3)</tt>. On Linux, you have to use google (or read the source).</p></div>
\r
548 <div class="paragraph"><p>The lists used are <tt>SLIST</tt> (single linked lists), <tt>CIRCLEQ</tt> (circular
\r
549 queues) and <tt>TAILQ</tt> (tail queues). Usually, only forward traversal is necessary,
\r
550 so an <tt>SLIST</tt> works fine. If inserting elements at arbitrary positions or at
\r
551 the end of a list is necessary, a <tt>TAILQ</tt> is used instead. However, for the
\r
552 windows inside a container, a <tt>CIRCLEQ</tt> is necessary to go from the currently
\r
553 selected window to the window above/below.</p></div>
\r
556 <div class="sect1">
\r
557 <h2 id="_naming_conventions">5. Naming conventions</h2>
\r
558 <div class="sectionbody">
\r
559 <div class="paragraph"><p>There is a row of standard variables used in many events. The following names
\r
560 should be chosen for those:</p></div>
\r
561 <div class="ulist"><ul>
\r
564 “conn” is the xcb_connection_t
\r
569 “event” is the event of the particular type
\r
574 “con” names a container
\r
579 “current” is a loop variable when using <tt>TAILQ_FOREACH</tt> etc.
\r
585 <div class="sect1">
\r
586 <h2 id="_startup_src_mainx_c_main">6. Startup (src/mainx.c, main())</h2>
\r
587 <div class="sectionbody">
\r
588 <div class="ulist"><ul>
\r
591 Establish the xcb connection
\r
596 Check for XKB extension on the separate X connection, load Xcursor
\r
601 Check for RandR screens (with a fall-back to Xinerama)
\r
606 Grab the keycodes for which bindings exist
\r
611 Manage all existing windows
\r
616 Enter the event loop
\r
622 <div class="sect1">
\r
623 <h2 id="_keybindings">7. Keybindings</h2>
\r
624 <div class="sectionbody">
\r
625 <div class="sect2">
\r
626 <h3 id="_grabbing_the_bindings">7.1. Grabbing the bindings</h3>
\r
627 <div class="paragraph"><p>Grabbing the bindings is quite straight-forward. You pass X your combination of
\r
628 modifiers and the keycode you want to grab and whether you want to grab them
\r
629 actively or passively. Most bindings (everything except for bindings using
\r
630 Mode_switch) are grabbed passively, that is, just the window manager gets the
\r
631 event and cannot replay it.</p></div>
\r
632 <div class="paragraph"><p>We need to grab bindings that use Mode_switch actively because of a bug in X.
\r
633 When the window manager receives the keypress/keyrelease event for an actively
\r
634 grabbed keycode, it has to decide what to do with this event: It can either
\r
635 replay it so that other applications get it or it can prevent other
\r
636 applications from receiving it.</p></div>
\r
637 <div class="paragraph"><p>So, why do we need to grab keycodes actively? Because X does not set the
\r
638 state-property of keypress/keyrelease events properly. The Mode_switch bit is
\r
639 not set and we need to get it using XkbGetState. This means we cannot pass X
\r
640 our combination of modifiers containing Mode_switch when grabbing the key and
\r
641 therefore need to grab the keycode itself without any modifiers. This means,
\r
642 if you bind Mode_switch + keycode 38 ("a"), i3 will grab keycode 38 ("a") and
\r
643 check on each press of "a" if the Mode_switch bit is set using XKB. If yes, it
\r
644 will handle the event, if not, it will replay the event.</p></div>
\r
646 <div class="sect2">
\r
647 <h3 id="_handling_a_keypress">7.2. Handling a keypress</h3>
\r
648 <div class="paragraph"><p>As mentioned in "Grabbing the bindings", upon a keypress event, i3 first gets
\r
649 the correct state.</p></div>
\r
650 <div class="paragraph"><p>Then, it looks through all bindings and gets the one which matches the received
\r
652 <div class="paragraph"><p>The bound command is parsed by the cmdparse lexer/parser, see <tt>parse_cmd</tt> in
\r
653 <tt>src/cmdparse.y</tt>.</p></div>
\r
657 <div class="sect1">
\r
658 <h2 id="_manage_windows_src_main_c_manage_window_and_reparent_window">8. Manage windows (src/main.c, manage_window() and reparent_window())</h2>
\r
659 <div class="sectionbody">
\r
660 <div class="paragraph"><p><tt>manage_window()</tt> does some checks to decide whether the window should be
\r
661 managed at all:</p></div>
\r
662 <div class="ulist"><ul>
\r
665 Windows have to be mapped, that is, visible on screen
\r
670 The override_redirect must not be set. Windows with override_redirect shall
\r
671 not be managed by a window manager
\r
675 <div class="paragraph"><p>Afterwards, i3 gets the initial geometry and reparents the window (see
\r
676 <tt>reparent_window()</tt>) if it wasn’t already managed.</p></div>
\r
677 <div class="paragraph"><p>Reparenting means that for each window which is reparented, a new window,
\r
678 slightly larger than the original one, is created. The original window is then
\r
679 reparented to the bigger one (called "frame").</p></div>
\r
680 <div class="paragraph"><p>After reparenting, the window type (<tt>_NET_WM_WINDOW_TYPE</tt>) is checked to see
\r
681 whether this window is a dock (<tt>_NET_WM_WINDOW_TYPE_DOCK</tt>), like dzen2 for
\r
682 example. Docks are handled differently, they don’t have decorations and are not
\r
683 assigned to a specific container. Instead, they are positioned at the bottom
\r
684 or top of the screen (in the appropriate dock area containers). To get the
\r
685 height which needs to be reserved for the window, the <tt>_NET_WM_STRUT_PARTIAL</tt>
\r
686 property is used.</p></div>
\r
687 <div class="paragraph"><p>Furthermore, the list of assignments (to other workspaces, which may be on
\r
688 other screens) is checked. If the window matches one of the user’s criteria,
\r
689 it may either be put in floating mode or moved to a different workspace. If the
\r
690 target workspace is not visible, the window will not be mapped.</p></div>
\r
693 <div class="sect1">
\r
694 <h2 id="_what_happens_when_an_application_is_started">9. What happens when an application is started?</h2>
\r
695 <div class="sectionbody">
\r
696 <div class="paragraph"><p>i3 does not care about applications. All it notices is when new windows are
\r
697 mapped (see <tt>src/handlers.c</tt>, <tt>handle_map_request()</tt>). The window is then
\r
698 reparented (see section "Manage windows").</p></div>
\r
699 <div class="paragraph"><p>After reparenting the window, <tt>render_tree()</tt> is called which renders the
\r
700 internal layout table. The new window has been placed in the currently focused
\r
701 container and therefore the new window and the old windows (if any) need to be
\r
702 moved/resized so that the currently active layout (default/stacking/tabbed mode)
\r
703 is rendered correctly. To move/resize windows, a window is “configured” in
\r
704 X11-speak.</p></div>
\r
705 <div class="paragraph"><p>Some applications, such as MPlayer obviously assume the window manager is
\r
706 stupid and try to configure their windows by themselves. This generates an
\r
707 event called configurerequest. i3 handles these events and tells the window the
\r
708 size it had before the configurerequest (with the exception of not yet mapped
\r
709 windows, which get configured like they want to, and floating windows, which
\r
710 can reconfigure themselves).</p></div>
\r
713 <div class="sect1">
\r
714 <h2 id="_net_wm_state">10. _NET_WM_STATE</h2>
\r
715 <div class="sectionbody">
\r
716 <div class="paragraph"><p>Only the _NET_WM_STATE_FULLSCREEN and _NET_WM_STATE_DEMANDS_ATTENTION atoms
\r
717 are handled.</p></div>
\r
718 <div class="paragraph"><p>The former calls “toggle_fullscreen()” for the specific client which just
\r
719 configures the client to use the whole screen on which it currently is.
\r
720 Also, it is set as fullscreen_client for the i3Screen.</p></div>
\r
721 <div class="paragraph"><p>The latter is used to set, read and display urgency hints.</p></div>
\r
724 <div class="sect1">
\r
725 <h2 id="_wm_name">11. WM_NAME</h2>
\r
726 <div class="sectionbody">
\r
727 <div class="paragraph"><p>When the WM_NAME property of a window changes, its decoration (containing the
\r
728 title) is re-rendered. Note that WM_NAME is in COMPOUND_TEXT encoding which is
\r
729 totally uncommon and cumbersome. Therefore, the _NET_WM_NAME atom will be used
\r
730 if present.</p></div>
\r
733 <div class="sect1">
\r
734 <h2 id="_net_wm_name">12. _NET_WM_NAME</h2>
\r
735 <div class="sectionbody">
\r
736 <div class="paragraph"><p>Like WM_NAME, this atom contains the title of a window. However, _NET_WM_NAME
\r
737 is encoded in UTF-8. i3 will recode it to UCS-2 in order to be able to pass it
\r
738 to X. Using an appropriate font (ISO-10646), you can see most special
\r
739 characters (every special character contained in your font).</p></div>
\r
742 <div class="sect1">
\r
743 <h2 id="_size_hints">13. Size hints</h2>
\r
744 <div class="sectionbody">
\r
745 <div class="paragraph"><p>Size hints specify the minimum/maximum size for a given window as well as its
\r
746 aspect ratio. This is important for clients like mplayer, who only set the
\r
747 aspect ratio and resize their window to be as small as possible (but only with
\r
748 some video outputs, for example in Xv, while when using x11, mplayer does the
\r
749 necessary centering for itself).</p></div>
\r
750 <div class="paragraph"><p>So, when an aspect ratio was specified, i3 adjusts the height of the window
\r
751 until the size maintains the correct aspect ratio. For the code to do this, see
\r
752 src/layout.c, function resize_client().</p></div>
\r
755 <div class="sect1">
\r
756 <h2 id="_rendering_src_layout_c_render_layout_and_render_container">14. Rendering (src/layout.c, render_layout() and render_container())</h2>
\r
757 <div class="sectionbody">
\r
758 <div class="paragraph"><p>Rendering in i3 version 4 is the step which assigns the correct sizes for
\r
759 borders, decoration windows, child windows and the stacking order of all
\r
760 windows. In a separate step (<tt>x_push_changes()</tt>), these changes are pushed to
\r
762 <div class="paragraph"><p>Keep in mind that all these properties (<tt>rect</tt>, <tt>window_rect</tt> and <tt>deco_rect</tt>)
\r
763 are temporary, meaning they will be overwritten by calling <tt>render_con</tt>.
\r
764 Persistent position/size information is kept in <tt>geometry</tt>.</p></div>
\r
765 <div class="paragraph"><p>The entry point for every rendering operation (except for the case of moving
\r
766 floating windows around) currently is <tt>tree_render()</tt> which will re-render
\r
767 everything that’s necessary (for every output, only the currently displayed
\r
768 workspace is rendered). This behavior is expected to change in the future,
\r
769 since for a lot of updates, re-rendering everything is not actually necessary.
\r
770 Focus was on getting it working correct, not getting it work very fast.</p></div>
\r
771 <div class="paragraph"><p>What <tt>tree_render()</tt> actually does is calling <tt>render_con()</tt> on the root
\r
772 container and then pushing the changes to X11. The following sections talk
\r
773 about the different rendering steps, in the order of "top of the tree" (root
\r
774 container) to the bottom.</p></div>
\r
775 <div class="sect2">
\r
776 <h3 id="_rendering_the_root_container">14.1. Rendering the root container</h3>
\r
777 <div class="paragraph"><p>The i3 root container (<tt>con->type == CT_ROOT</tt>) represents the X11 root window.
\r
778 It contains one child container for every output (like LVDS1, VGA1, …), which
\r
779 is available on your computer.</p></div>
\r
780 <div class="paragraph"><p>Rendering the root will first render all tiling windows and then all floating
\r
781 windows. This is necessary because a floating window can be positioned in such
\r
782 a way that it is visible on two different outputs. Therefore, by first
\r
783 rendering all the tiling windows (of all outputs), we make sure that floating
\r
784 windows can never be obscured by tiling windows.</p></div>
\r
785 <div class="paragraph"><p>Essentially, though, this code path will just call <tt>render_con()</tt> for every
\r
786 output and <tt>x_raise_con(); render_con()</tt> for every floating window.</p></div>
\r
787 <div class="paragraph"><p>In the special case of having a "global fullscreen" window (fullscreen mode
\r
788 spanning all outputs), a shortcut is taken and <tt>x_raise_con(); render_con()</tt> is
\r
789 only called for the global fullscreen window.</p></div>
\r
791 <div class="sect2">
\r
792 <h3 id="_rendering_an_output">14.2. Rendering an output</h3>
\r
793 <div class="paragraph"><p>Output containers (<tt>con->layout == L_OUTPUT</tt>) represent a hardware output like
\r
794 LVDS1, VGA1, etc. An output container has three children (at the moment): One
\r
795 content container (having workspaces as children) and the top/bottom dock area
\r
796 containers.</p></div>
\r
797 <div class="paragraph"><p>The rendering happens in the function <tt>render_l_output()</tt> in the following
\r
799 <div class="olist arabic"><ol class="arabic">
\r
802 Find the content container (<tt>con->type == CT_CON</tt>)
\r
807 Get the currently visible workspace (<tt>con_get_fullscreen_con(content,
\r
813 If there is a fullscreened window on that workspace, directly render it and
\r
814 return, thus ignoring the dock areas.
\r
819 Sum up the space used by all the dock windows (they have a variable height
\r
825 Set the workspace rects (x/y/width/height) based on the position of the
\r
826 output (stored in <tt>con->rect</tt>) and the usable space
\r
827 (<tt>con->rect.{width,height}</tt> without the space used for dock windows).
\r
832 Recursively raise and render the output’s child containers (meaning dock
\r
833 area containers and the content container).
\r
838 <div class="sect2">
\r
839 <h3 id="_rendering_a_workspace_or_split_container">14.3. Rendering a workspace or split container</h3>
\r
840 <div class="paragraph"><p>From here on, there really is no difference anymore. All containers are of
\r
841 <tt>con->type == CT_CON</tt> (whether workspace or split container) and some of them
\r
842 have a <tt>con->window</tt>, meaning they represent an actual window instead of a
\r
843 split container.</p></div>
\r
844 <div class="sect3">
\r
845 <h4 id="_default_layout">14.3.1. Default layout</h4>
\r
846 <div class="paragraph"><p>In default layout, containers are placed horizontally or vertically next to
\r
847 each other (depending on the <tt>con->orientation</tt>). If a child is a leaf node (as
\r
848 opposed to a split container) and has border style "normal", appropriate space
\r
849 will be reserved for its window decoration.</p></div>
\r
851 <div class="sect3">
\r
852 <h4 id="_stacked_layout">14.3.2. Stacked layout</h4>
\r
853 <div class="paragraph"><p>In stacked layout, only the focused window is actually shown (this is achieved
\r
854 by calling <tt>x_raise_con()</tt> in reverse focus order at the end of <tt>render_con()</tt>).</p></div>
\r
855 <div class="paragraph"><p>The available space for the focused window is the size of the container minus
\r
856 the height of the window decoration for all windows inside this stacked
\r
857 container.</p></div>
\r
858 <div class="paragraph"><p>If border style is "1pixel" or "none", no window decoration height will be
\r
859 reserved (or displayed later on), unless there is more than one window inside
\r
860 the stacked container.</p></div>
\r
862 <div class="sect3">
\r
863 <h4 id="_tabbed_layout">14.3.3. Tabbed layout</h4>
\r
864 <div class="paragraph"><p>Tabbed layout works precisely like stacked layout, but the window decoration
\r
865 position/size is different: They are placed next to each other on a single line
\r
866 (fixed height).</p></div>
\r
868 <div class="sect3">
\r
869 <h4 id="_dock_area_layout">14.3.4. Dock area layout</h4>
\r
870 <div class="paragraph"><p>This is a special case. Users cannot choose the dock area layout, but it will be
\r
871 set for the dock area containers. In the dockarea layout (at the moment!),
\r
872 windows will be placed above each other.</p></div>
\r
875 <div class="sect2">
\r
876 <h3 id="_rendering_a_window">14.4. Rendering a window</h3>
\r
877 <div class="paragraph"><p>A window’s size and position will be determined in the following way:</p></div>
\r
878 <div class="olist arabic"><ol class="arabic">
\r
881 Subtract the border if border style is not "none" (but "normal" or "1pixel").
\r
886 Subtract the X11 border, if the window has an X11 border > 0.
\r
891 Obey the aspect ratio of the window (think MPlayer).
\r
896 Obey the height- and width-increments of the window (think terminal emulator
\r
897 which can only be resized in one-line or one-character steps).
\r
904 <div class="sect1">
\r
905 <h2 id="_pushing_updates_to_x11_drawing">15. Pushing updates to X11 / Drawing</h2>
\r
906 <div class="sectionbody">
\r
907 <div class="paragraph"><p>A big problem with i3 before version 4 was that we just sent requests to X11
\r
908 anywhere in the source code. This was bad because nobody could understand the
\r
909 entirety of our interaction with X11, it lead to subtle bugs and a lot of edge
\r
910 cases which we had to consider all over again.</p></div>
\r
911 <div class="paragraph"><p>Therefore, since version 4, we have a single file, <tt>src/x.c</tt>, which is
\r
912 responsible for repeatedly transferring parts of our tree datastructure to X11.</p></div>
\r
913 <div class="paragraph"><p><tt>src/x.c</tt> consists of multiple parts:</p></div>
\r
914 <div class="olist arabic"><ol class="arabic">
\r
917 The state pushing: <tt>x_push_changes()</tt>, which calls <tt>x_push_node()</tt>.
\r
922 State modification functions: <tt>x_con_init</tt>, <tt>x_reinit</tt>,
\r
923 <tt>x_reparent_child</tt>, <tt>x_move_win</tt>, <tt>x_con_kill</tt>, <tt>x_raise_con</tt>, <tt>x_set_name</tt>
\r
924 and <tt>x_set_warp_to</tt>.
\r
929 Expose event handling (drawing decorations): <tt>x_deco_recurse()</tt> and
\r
930 <tt>x_draw_decoration()</tt>.
\r
934 <div class="sect2">
\r
935 <h3 id="_pushing_state_to_x11">15.1. Pushing state to X11</h3>
\r
936 <div class="paragraph"><p>In general, the function <tt>x_push_changes</tt> should be called to push state
\r
937 changes. Only when the scope of the state change is clearly defined (for
\r
938 example only the title of a window) and its impact is known beforehand, one can
\r
939 optimize this and call <tt>x_push_node</tt> on the appropriate con directly.</p></div>
\r
940 <div class="paragraph"><p><tt>x_push_changes</tt> works in the following steps:</p></div>
\r
941 <div class="olist arabic"><ol class="arabic">
\r
944 Clear the eventmask for all mapped windows. This leads to not getting
\r
945 useless ConfigureNotify or EnterNotify events which are caused by our
\r
946 requests. In general, we only want to handle user input.
\r
951 Stack windows above each other, in reverse stack order (starting with the
\r
952 most obscured/bottom window). This is relevant for floating windows which
\r
953 can overlap each other, but also for tiling windows in stacked or tabbed
\r
954 containers. We also update the <tt>_NET_CLIENT_LIST_STACKING</tt> hint which is
\r
955 necessary for tab drag and drop in Chromium.
\r
960 <tt>x_push_node</tt> will be called for the root container, recursively calling
\r
961 itself for the container’s children. This function actually pushes the
\r
962 state, see the next paragraph.
\r
967 If the pointer needs to be warped to a different position (for example when
\r
968 changing focus to a different output), it will be warped now.
\r
973 The eventmask is restored for all mapped windows.
\r
978 Window decorations will be rendered by calling <tt>x_deco_recurse</tt> on the root
\r
979 container, which then recursively calls itself for the children.
\r
984 If the input focus needs to be changed (because the user focused a different
\r
985 window), it will be updated now.
\r
990 <tt>x_push_node_unmaps</tt> will be called for the root container. This function
\r
991 only pushes UnmapWindow requests. Separating the state pushing is necessary
\r
992 to handle fullscreen windows (and workspace switches) in a smooth fashion:
\r
993 The newly visible windows should be visible before the old windows are
\r
998 <div class="paragraph"><p><tt>x_push_node</tt> works in the following steps:</p></div>
\r
999 <div class="olist arabic"><ol class="arabic">
\r
1002 Update the window’s <tt>WM_NAME</tt>, if changed (the <tt>WM_NAME</tt> is set on i3
\r
1003 containers mainly for debugging purposes).
\r
1008 Reparents a child window into the i3 container if the container was created
\r
1009 for a specific managed window.
\r
1014 If the size/position of the i3 container changed (due to opening a new
\r
1015 window or switching layouts for example), the window will be reconfigured.
\r
1016 Also, the pixmap which is used to draw the window decoration/border on is
\r
1017 reconfigured (pixmaps are size-dependent).
\r
1022 Size/position for the child window is adjusted.
\r
1027 The i3 container is mapped if it should be visible and was not yet mapped.
\r
1028 When mapping, <tt>WM_STATE</tt> is set to <tt>WM_STATE_NORMAL</tt>. Also, the eventmask of
\r
1029 the child window is updated and the i3 container’s contents are copied from
\r
1035 <tt>x_push_node</tt> is called recursively for all children of the current
\r
1040 <div class="paragraph"><p><tt>x_push_node_unmaps</tt> handles the remaining case of an i3 container being
\r
1041 unmapped if it should not be visible anymore. <tt>WM_STATE</tt> will be set to
\r
1042 <tt>WM_STATE_WITHDRAWN</tt>.</p></div>
\r
1044 <div class="sect2">
\r
1045 <h3 id="_drawing_window_decorations_borders_backgrounds">15.2. Drawing window decorations/borders/backgrounds</h3>
\r
1046 <div class="paragraph"><p><tt>x_draw_decoration</tt> draws window decorations. It is run for every leaf
\r
1047 container (representing an actual X11 window) and for every non-leaf container
\r
1048 which is in a stacked/tabbed container (because stacked/tabbed containers
\r
1049 display a window decoration for split containers, which consists of a representation
\r
1050 of the child container’s names.</p></div>
\r
1051 <div class="paragraph"><p>Then, parameters are collected to be able to determine whether this decoration
\r
1052 drawing is actually necessary or was already done. This saves a substantial
\r
1053 number of redraws (depending on your workload, but far over 50%).</p></div>
\r
1054 <div class="paragraph"><p>Assuming that we need to draw this decoration, we start by filling the empty
\r
1055 space around the child window (think of MPlayer with a specific aspect ratio)
\r
1056 in the user-configured client background color.</p></div>
\r
1057 <div class="paragraph"><p>Afterwards, we draw the appropriate border (in case of border styles "normal"
\r
1058 and "1pixel") and the top bar (in case of border style "normal").</p></div>
\r
1059 <div class="paragraph"><p>The last step is drawing the window title on the top bar.</p></div>
\r
1063 <div class="sect1">
\r
1064 <h2 id="_user_commands_parser_specs_commands_spec">16. User commands (parser-specs/commands.spec)</h2>
\r
1065 <div class="sectionbody">
\r
1066 <div class="paragraph"><p>In the configuration file and when using i3 interactively (with <tt>i3-msg</tt>, for
\r
1067 example), you use commands to make i3 do things, like focus a different window,
\r
1068 set a window to fullscreen, and so on. An example command is <tt>floating enable</tt>,
\r
1069 which enables floating mode for the currently focused window. See the
\r
1070 appropriate section in the <a href="userguide.html">User’s Guide</a> for a reference of
\r
1071 all commands.</p></div>
\r
1072 <div class="paragraph"><p>In earlier versions of i3, interpreting these commands was done using lex and
\r
1073 yacc, but experience has shown that lex and yacc are not well suited for our
\r
1074 command language. Therefore, starting from version 4.2, we use a custom parser
\r
1075 for user commands and the configuration file.
\r
1076 The input specification for this parser can be found in the file
\r
1077 <tt>parser-specs/*.spec</tt>. Should you happen to use Vim as an editor, use
\r
1078 :source parser-specs/highlighting.vim to get syntax highlighting for this file
\r
1079 (highlighting files for other editors are welcome).</p></div>
\r
1080 <div class="listingblock">
\r
1081 <div class="title">Excerpt from commands.spec</div>
\r
1082 <div class="content">
\r
1083 <pre><tt>state INITIAL:
\r
1084 '[' -> call cmd_criteria_init(); CRITERIA
\r
1087 'workspace' -> WORKSPACE
\r
1088 'exit' -> call cmd_exit()
\r
1089 'restart' -> call cmd_restart()
\r
1090 'reload' -> call cmd_reload()</tt></pre>
\r
1092 <div class="paragraph"><p>The input specification is written in an extremely simple format. The
\r
1093 specification is then converted into C code by the Perl script
\r
1094 generate-commands-parser.pl (the output file names begin with GENERATED and the
\r
1095 files are stored in the <tt>include</tt> directory). The parser implementation
\r
1096 <tt>src/commands_parser.c</tt> includes the generated C code at compile-time.</p></div>
\r
1097 <div class="paragraph"><p>The above excerpt from commands.spec illustrates nearly all features of our
\r
1098 specification format: You describe different states and what can happen within
\r
1099 each state. State names are all-caps; the state in the above excerpt is called
\r
1100 INITIAL. A list of tokens and their actions (separated by an ASCII arrow)
\r
1101 follows. In the excerpt, all tokens are literals, that is, simple text strings
\r
1102 which will be compared with the input. An action is either the name of a state
\r
1103 in which the parser will transition into, or the keyword <em>call</em>, followed by
\r
1104 the name of a function (and optionally a state).</p></div>
\r
1105 <div class="sect2">
\r
1106 <h3 id="_example_the_workspace_state">16.1. Example: The WORKSPACE state</h3>
\r
1107 <div class="paragraph"><p>Let’s have a look at the WORKSPACE state, which is a good example of all
\r
1108 features. This is its definition:</p></div>
\r
1109 <div class="listingblock">
\r
1110 <div class="title">WORKSPACE state (commands.spec)</div>
\r
1111 <div class="content">
\r
1112 <pre><tt># workspace next|prev|next_on_output|prev_on_output
\r
1113 # workspace back_and_forth
\r
1114 # workspace <name>
\r
1115 # workspace number <number>
\r
1117 direction = 'next_on_output', 'prev_on_output', 'next', 'prev'
\r
1118 -> call cmd_workspace($direction)
\r
1120 -> call cmd_workspace_back_and_forth()
\r
1122 -> WORKSPACE_NUMBER
\r
1123 workspace = string
\r
1124 -> call cmd_workspace_name($workspace)</tt></pre>
\r
1126 <div class="paragraph"><p>As you can see from the commands, there are multiple different valid variants
\r
1127 of the workspace command:</p></div>
\r
1128 <div class="dlist"><dl>
\r
1129 <dt class="hdlist1">
\r
1130 workspace <direction>
\r
1134 The word <em>workspace</em> can be followed by any of the tokens <em>next</em>,
\r
1135 <em>prev</em>, <em>next_on_output</em> or <em>prev_on_output</em>. This command will
\r
1136 switch to the next or previous workspace (optionally on the same
\r
1138 There is one function called <tt>cmd_workspace</tt>, which is defined
\r
1139 in <tt>src/commands.c</tt>. It will handle this kind of command. To know which
\r
1140 direction was specified, the direction token is stored on the stack
\r
1141 with the name "direction", which is what the "direction = " means in
\r
1142 the beginning.<br />
\r
1146 <div class="admonitionblock">
\r
1149 <div class="title">Note</div>
\r
1151 <td class="content">Note that you can specify multiple literals in the same line. This has
\r
1152 exactly the same effect as if you specified <tt>direction =
\r
1153 'next_on_output' -> call cmd_workspace($direction)</tt> and so forth.<br /></td>
\r
1156 <div class="admonitionblock">
\r
1159 <div class="title">Note</div>
\r
1161 <td class="content">Also note that the order of literals is important here: If <em>next</em> were
\r
1162 ordered before <em>next_on_output</em>, then <em>next_on_output</em> would never
\r
1166 <div class="dlist"><dl>
\r
1167 <dt class="hdlist1">
\r
1168 workspace back_and_forth
\r
1172 This is a very simple case: When the literal <em>back_and_forth</em> is found
\r
1173 in the input, the function <tt>cmd_workspace_back_and_forth</tt> will be
\r
1174 called without parameters and the parser will return to the INITIAL
\r
1175 state (since no other state was specified).
\r
1178 <dt class="hdlist1">
\r
1179 workspace <name>
\r
1183 In this case, the workspace command is followed by an arbitrary string,
\r
1184 possibly in quotes, for example "workspace 3" or "workspace bleh".<br />
\r
1185 This is the first time that the token is actually not a literal (not in
\r
1186 single quotes), but just called string. Other possible tokens are word
\r
1187 (the same as string, but stops matching at a whitespace) and end
\r
1188 (matches the end of the input).
\r
1191 <dt class="hdlist1">
\r
1192 workspace number <number>
\r
1196 The workspace command has to be followed by the keyword <tt>number</tt>. It
\r
1197 then transitions into the state <tt>WORKSPACE_NUMBER</tt>, where the actual
\r
1198 parameter will be read.
\r
1203 <div class="sect2">
\r
1204 <h3 id="_introducing_a_new_command">16.2. Introducing a new command</h3>
\r
1205 <div class="paragraph"><p>The following steps have to be taken in order to properly introduce a new
\r
1206 command (or possibly extend an existing command):</p></div>
\r
1207 <div class="olist arabic"><ol class="arabic">
\r
1210 Define a function beginning with <tt>cmd_</tt> in the file <tt>src/commands.c</tt>. Copy
\r
1211 the prototype of an existing function.
\r
1216 After adding a comment on what the function does, copy the comment and
\r
1217 function definition to <tt>include/commands.h</tt>. Make the comment in the header
\r
1218 file use double asterisks to make doxygen pick it up.
\r
1223 Write a test case (or extend an existing test case) for your feature, see
\r
1224 <a href="testsuite.html">i3 testsuite</a>. For now, it is sufficient to simply call
\r
1225 your command in all the various possible ways.
\r
1230 Extend the parser specification in <tt>parser-specs/commands.spec</tt>. Run the
\r
1231 testsuite and see if your new function gets called with the appropriate
\r
1232 arguments for the appropriate input.
\r
1237 Actually implement the feature.
\r
1242 Document the feature in the <a href="userguide.html">User’s Guide</a>.
\r
1249 <div class="sect1">
\r
1250 <h2 id="_moving_containers">17. Moving containers</h2>
\r
1251 <div class="sectionbody">
\r
1252 <div class="paragraph"><p>The movement code is pretty delicate. You need to consider all cases before
\r
1253 making any changes or before being able to fully understand how it works.</p></div>
\r
1254 <div class="sect2">
\r
1255 <h3 id="_case_1_moving_inside_the_same_container">17.1. Case 1: Moving inside the same container</h3>
\r
1256 <div class="paragraph"><p>The reference layout for this case is a single workspace in horizontal
\r
1257 orientation with two containers on it. Focus is on the left container (1).</p></div>
\r
1258 <div class="tableblock">
\r
1259 <table rules="all"
\r
1262 cellspacing="0" cellpadding="4">
\r
1263 <col width="50%" />
\r
1264 <col width="50%" />
\r
1267 <td align="center" valign="top"><p class="table">1</p></td>
\r
1268 <td align="center" valign="top"><p class="table">2</p></td>
\r
1273 <div class="paragraph"><p>When moving the left window to the right (command <tt>move right</tt>), tree_move will
\r
1274 look for a container with horizontal orientation and finds the parent of the
\r
1275 left container, that is, the workspace. Afterwards, it runs the code branch
\r
1276 commented with "the easy case": it calls TAILQ_NEXT to get the container right
\r
1277 of the current one and swaps both containers.</p></div>
\r
1279 <div class="sect2">
\r
1280 <h3 id="_case_2_move_a_container_into_a_split_container">17.2. Case 2: Move a container into a split container</h3>
\r
1281 <div class="paragraph"><p>The reference layout for this case is a horizontal workspace with two
\r
1282 containers. The right container is a v-split with two containers. Focus is on
\r
1283 the left container (1).</p></div>
\r
1284 <div class="tableblock">
\r
1285 <table rules="all"
\r
1288 cellspacing="0" cellpadding="4">
\r
1289 <col width="50%" />
\r
1290 <col width="50%" />
\r
1293 <td rowspan="2" align="center" valign="middle"><p class="table">1</p></td>
\r
1294 <td align="center" valign="top"><p class="table">2</p></td>
\r
1297 <td align="center" valign="top"><p class="table">3</p></td>
\r
1302 <div class="paragraph"><p>When moving to the right (command <tt>move right</tt>), i3 will work like in case 1
\r
1303 ("the easy case"). However, as the right container is not a leaf container, but
\r
1304 a v-split, the left container (1) will be inserted at the right position (below
\r
1305 2, assuming that 2 is focused inside the v-split) by calling <tt>insert_con_into</tt>.</p></div>
\r
1306 <div class="paragraph"><p><tt>insert_con_into</tt> detaches the container from its parent and inserts it
\r
1307 before/after the given target container. Afterwards, the on_remove_child
\r
1308 callback is called on the old parent container which will then be closed, if
\r
1310 <div class="paragraph"><p>Afterwards, <tt>con_focus</tt> will be called to fix the focus stack and the tree will
\r
1311 be flattened.</p></div>
\r
1313 <div class="sect2">
\r
1314 <h3 id="_case_3_moving_to_non_existent_top_bottom">17.3. Case 3: Moving to non-existent top/bottom</h3>
\r
1315 <div class="paragraph"><p>Like in case 1, the reference layout for this case is a single workspace in
\r
1316 horizontal orientation with two containers on it. Focus is on the left
\r
1317 container:</p></div>
\r
1318 <div class="tableblock">
\r
1319 <table rules="all"
\r
1322 cellspacing="0" cellpadding="4">
\r
1323 <col width="50%" />
\r
1324 <col width="50%" />
\r
1327 <td align="center" valign="top"><p class="table">1</p></td>
\r
1328 <td align="center" valign="top"><p class="table">2</p></td>
\r
1333 <div class="paragraph"><p>This time however, the command is <tt>move up</tt> or <tt>move down</tt>. tree_move will look
\r
1334 for a container with vertical orientation. As it will not find any,
\r
1335 <tt>same_orientation</tt> is NULL and therefore i3 will perform a forced orientation
\r
1336 change on the workspace by creating a new h-split container, moving the
\r
1337 workspace contents into it and then changing the workspace orientation to
\r
1338 vertical. Now it will again search for parent containers with vertical
\r
1339 orientation and it will find the workspace.</p></div>
\r
1340 <div class="paragraph"><p>This time, the easy case code path will not be run as we are not moving inside
\r
1341 the same container. Instead, <tt>insert_con_into</tt> will be called with the focused
\r
1342 container and the container above/below the current one (on the level of
\r
1343 <tt>same_orientation</tt>).</p></div>
\r
1344 <div class="paragraph"><p>Now, <tt>con_focus</tt> will be called to fix the focus stack and the tree will be
\r
1345 flattened.</p></div>
\r
1347 <div class="sect2">
\r
1348 <h3 id="_case_4_moving_to_existent_top_bottom">17.4. Case 4: Moving to existent top/bottom</h3>
\r
1349 <div class="paragraph"><p>The reference layout for this case is a vertical workspace with two containers.
\r
1350 The bottom one is a h-split containing two containers (1 and 2). Focus is on
\r
1351 the bottom left container (1).</p></div>
\r
1352 <div class="tableblock">
\r
1353 <table rules="all"
\r
1356 cellspacing="0" cellpadding="4">
\r
1357 <col width="50%" />
\r
1358 <col width="50%" />
\r
1361 <td colspan="2" align="center" valign="top"><p class="table">3</p></td>
\r
1364 <td align="center" valign="top"><p class="table">1</p></td>
\r
1365 <td align="center" valign="top"><p class="table">2</p></td>
\r
1370 <div class="paragraph"><p>This case is very much like case 3, only this time the forced workspace
\r
1371 orientation change does not need to be performed because the workspace already
\r
1372 is in vertical orientation.</p></div>
\r
1374 <div class="sect2">
\r
1375 <h3 id="_case_5_moving_in_one_child_h_split">17.5. Case 5: Moving in one-child h-split</h3>
\r
1376 <div class="paragraph"><p>The reference layout for this case is a horizontal workspace with two
\r
1377 containers having a v-split on the left side with a one-child h-split on the
\r
1378 bottom. Focus is on the bottom left container (2(h)):</p></div>
\r
1379 <div class="tableblock">
\r
1380 <table rules="all"
\r
1383 cellspacing="0" cellpadding="4">
\r
1384 <col width="50%" />
\r
1385 <col width="50%" />
\r
1388 <td align="center" valign="top"><p class="table">1</p></td>
\r
1389 <td rowspan="2" align="center" valign="middle"><p class="table">3</p></td>
\r
1392 <td align="center" valign="top"><p class="table">2(h)</p></td>
\r
1397 <div class="paragraph"><p>In this case, <tt>same_orientation</tt> will be set to the h-split container around
\r
1398 the focused container. However, when trying the easy case, the next/previous
\r
1399 container <tt>swap</tt> will be NULL. Therefore, i3 will search again for a
\r
1400 <tt>same_orientation</tt> container, this time starting from the parent of the h-split
\r
1401 container.</p></div>
\r
1402 <div class="paragraph"><p>After determining a new <tt>same_orientation</tt> container (if it is NULL, the
\r
1403 orientation will be force-changed), this case is equivalent to case 2 or case
\r
1406 <div class="sect2">
\r
1407 <h3 id="_case_6_floating_containers">17.6. Case 6: Floating containers</h3>
\r
1408 <div class="paragraph"><p>The reference layout for this case is a horizontal workspace with two
\r
1409 containers plus one floating h-split container. Focus is on the floating
\r
1410 container.</p></div>
\r
1411 <div class="paragraph"><p>TODO: nice illustration. table not possible?</p></div>
\r
1412 <div class="paragraph"><p>When moving up/down, the container needs to leave the floating container and it
\r
1413 needs to be placed on the workspace (at workspace level). This is accomplished
\r
1414 by calling the function <tt>attach_to_workspace</tt>.</p></div>
\r
1418 <div class="sect1">
\r
1419 <h2 id="_click_handling">18. Click handling</h2>
\r
1420 <div class="sectionbody">
\r
1421 <div class="paragraph"><p>Without much ado, here is the list of cases which need to be considered:</p></div>
\r
1422 <div class="ulist"><ul>
\r
1425 click to focus (tiling + floating) and raise (floating)
\r
1430 click to focus/raise when in stacked/tabbed mode
\r
1435 floating_modifier + left mouse button to drag a floating con
\r
1440 floating_modifier + right mouse button to resize a floating con
\r
1445 click on decoration in a floating con to either initiate a resize (if there
\r
1446 is more than one child in the floating con) or to drag the
\r
1447 floating con (if it’s the one at the top).
\r
1452 click on border in a floating con to resize the floating con
\r
1457 floating_modifier + right mouse button to resize a tiling con
\r
1462 click on border/decoration to resize a tiling con
\r
1468 <div class="sect1">
\r
1469 <h2 id="_gotchas">19. Gotchas</h2>
\r
1470 <div class="sectionbody">
\r
1471 <div class="ulist"><ul>
\r
1474 Forgetting to call <tt>xcb_flush(conn);</tt> after sending a request. This usually
\r
1475 leads to code which looks like it works fine but which does not work under
\r
1476 certain conditions.
\r
1481 Forgetting to call <tt>floating_fix_coordinates(con, old_rect, new_rect)</tt> after
\r
1482 moving workspaces across outputs. Coordinates for floating containers are
\r
1483 not relative to workspace boundaries, so you must correct their coordinates
\r
1484 or those containers will show up in the wrong workspace or not at all.
\r
1490 <div class="sect1">
\r
1491 <h2 id="_using_git_sending_patches">20. Using git / sending patches</h2>
\r
1492 <div class="sectionbody">
\r
1493 <div class="sect2">
\r
1494 <h3 id="_introduction">20.1. Introduction</h3>
\r
1495 <div class="paragraph"><p>For a short introduction into using git, see
\r
1496 <a href="http://web.archive.org/web/20121024222556/http://www.spheredev.org/wiki/Git_for_the_lazy">http://web.archive.org/web/20121024222556/http://www.spheredev.org/wiki/Git_for_the_lazy</a>
\r
1497 or, for more documentation, see <a href="http://git-scm.com/documentation">http://git-scm.com/documentation</a></p></div>
\r
1498 <div class="paragraph"><p>Please talk to us before working on new features to see whether they will be
\r
1499 accepted. A good way for this is to open an issue and asking for opinions on it.
\r
1500 Even for accepted features, this can be a good way to refine an idea upfront. However,
\r
1501 we don’t want to see certain features in i3, e.g., switching window focus in an
\r
1502 Alt+Tab like way.</p></div>
\r
1503 <div class="paragraph"><p>When working on bugfixes, please make sure you mention that you are working on
\r
1504 it in the corresponding bug report at <a href="https://github.com/i3/i3/issues">https://github.com/i3/i3/issues</a>. In case
\r
1505 there is no bug report yet, please create one.</p></div>
\r
1506 <div class="paragraph"><p>After you are done, please submit your work for review as a pull request at
\r
1507 <a href="https://github.com/i3/i3">https://github.com/i3/i3</a>.</p></div>
\r
1508 <div class="paragraph"><p>Do not send emails to the mailing list or any author directly, and don’t submit
\r
1509 them in the bugtracker, since all reviews should be done in public at
\r
1510 <a href="https://github.com/i3/i3">https://github.com/i3/i3</a>. In order to make your review go as fast as possible, you
\r
1511 could have a look at previous reviews and see what the common mistakes are.</p></div>
\r
1513 <div class="sect2">
\r
1514 <h3 id="_which_branch_to_use">20.2. Which branch to use?</h3>
\r
1515 <div class="paragraph"><p>Work on i3 generally happens in two branches: “master” and “next” (the latter
\r
1516 being the default branch, the one that people get when they check out the git
\r
1517 repository).</p></div>
\r
1518 <div class="paragraph"><p>The contents of “master” are always stable. That is, it contains the source code
\r
1519 of the latest release, plus any bugfixes that were applied since that release.</p></div>
\r
1520 <div class="paragraph"><p>New features are only found in the “next” branch. Therefore, if you are working
\r
1521 on a new feature, use the “next” branch. If you are working on a bugfix, use the
\r
1522 “next” branch, too, but make sure your code also works on “master”.</p></div>
\r
1524 <div class="sect2">
\r
1525 <h3 id="_how_to_build">20.3. How to build?</h3>
\r
1526 <div class="paragraph"><p>You can build i3 like you build any other software package which uses autotools.
\r
1527 Here’s a memory refresher:</p></div>
\r
1528 <div class="literalblock">
\r
1529 <div class="content">
\r
1530 <pre><tt>$ autoreconf -fi
\r
1531 $ mkdir -p build && cd build
\r
1533 $ make -j8</tt></pre>
\r
1535 <div class="paragraph"><p>(The autoreconf -fi step is unnecessary if you are building from a release tarball,
\r
1536 but shouldn’t hurt either.)</p></div>
\r
1537 <div class="sect3">
\r
1538 <h4 id="_build_system_features">20.3.1. Build system features</h4>
\r
1539 <div class="ulist"><ul>
\r
1542 We use the AX_ENABLE_BUILDDIR macro to enforce builds happening in a separate
\r
1543 directory. This is a prerequisite for the AX_EXTEND_SRCDIR macro and building
\r
1544 in a separate directory is common practice anyway. In case this causes any
\r
1545 trouble when packaging i3 for your distribution, please open an issue.
\r
1550 “make check” runs the i3 testsuite. See docs/testsuite for details.
\r
1555 “make distcheck” (runs testsuite on “make dist” result, tiny bit quicker
\r
1556 feedback cycle than waiting for the travis build to catch the issue).
\r
1561 “make uninstall” (occasionally requested by users who compile from source)
\r
1566 “make” will build manpages/docs by default if the tools are installed.
\r
1567 Conversely, manpages/docs are not tried to be built for users who don’t want
\r
1568 to install all these dependencies to get started hacking on i3.
\r
1573 non-release builds will enable address sanitizer by default. Use the
\r
1574 --disable-sanitizers configure option to turn off all sanitizers, and see
\r
1575 --help for available sanitizers.
\r
1580 Support for pre-compiled headers (PCH) has been dropped for now in the
\r
1581 interest of simplicity. If you need support for PCH, please open an issue.
\r
1586 Coverage reports are now generated using “make check-code-coverage”, which
\r
1587 requires specifying --enable-code-coverage when calling configure.
\r
1595 <div class="sect1">
\r
1596 <h2 id="_thought_experiments">21. Thought experiments</h2>
\r
1597 <div class="sectionbody">
\r
1598 <div class="paragraph"><p>In this section, we collect thought experiments, so that we don’t forget our
\r
1599 thoughts about specific topics. They are not necessary to get into hacking i3,
\r
1600 but if you are interested in one of the topics they cover, you should read them
\r
1601 before asking us why things are the way they are or why we don’t implement
\r
1603 <div class="sect2">
\r
1604 <h3 id="_using_cgroups_per_workspace">21.1. Using cgroups per workspace</h3>
\r
1605 <div class="paragraph"><p>cgroups (control groups) are a linux-only feature which provides the ability to
\r
1606 group multiple processes. For each group, you can individually set resource
\r
1607 limits, like allowed memory usage. Furthermore, and more importantly for our
\r
1608 purposes, they serve as a namespace, a label which you can attach to processes
\r
1609 and their children.</p></div>
\r
1610 <div class="paragraph"><p>One interesting use for cgroups is having one cgroup per workspace (or
\r
1611 container, doesn’t really matter). That way, you could set different priorities
\r
1612 and have a workspace for important stuff (say, writing a LaTeX document or
\r
1613 programming) and a workspace for unimportant background stuff (say,
\r
1614 JDownloader). Both tasks can obviously consume a lot of I/O resources, but in
\r
1615 this example it doesn’t really matter if JDownloader unpacks the download a
\r
1616 minute earlier or not. However, your compiler should work as fast as possible.
\r
1617 Having one cgroup per workspace, you would assign more resources to the
\r
1618 programming workspace.</p></div>
\r
1619 <div class="paragraph"><p>Another interesting feature is that an inherent problem of the workspace
\r
1620 concept could be solved by using cgroups: When starting an application on
\r
1621 workspace 1, then switching to workspace 2, you will get the application’s
\r
1622 window(s) on workspace 2 instead of the one you started it on. This is because
\r
1623 the window manager does not have any mapping between the process it starts (or
\r
1624 gets started in any way) and the window(s) which appear.</p></div>
\r
1625 <div class="paragraph"><p>Imagine for example using dmenu: The user starts dmenu by pressing Mod+d, dmenu
\r
1626 gets started with PID 3390. The user then decides to launch Firefox, which
\r
1627 takes a long time. So they enter firefox into dmenu and press enter. Firefox
\r
1628 gets started with PID 4001. When it finally finishes loading, it creates an X11
\r
1629 window and uses MapWindow to make it visible. This is the first time i3
\r
1630 actually gets in touch with Firefox. It decides to map the window, but it has
\r
1631 no way of knowing that this window (even though it has the _NET_WM_PID property
\r
1632 set to 4001) belongs to the dmenu the user started before.</p></div>
\r
1633 <div class="paragraph"><p>How do cgroups help with this? Well, when pressing Mod+d to launch dmenu, i3
\r
1634 would create a new cgroup, let’s call it i3-3390-1. It launches dmenu in that
\r
1635 cgroup, which gets PID 3390. As before, the user enters firefox and Firefox
\r
1636 gets launched with PID 4001. This time, though, the Firefox process with PID
\r
1637 4001 is <strong>also</strong> member of the cgroup i3-3390-1 (because fork()ing in a cgroup
\r
1638 retains the cgroup property). Therefore, when mapping the window, i3 can look
\r
1639 up in which cgroup the process is and can establish a mapping between the
\r
1640 workspace and the window.</p></div>
\r
1641 <div class="paragraph"><p>There are multiple problems with this approach:</p></div>
\r
1642 <div class="olist arabic"><ol class="arabic">
\r
1645 Every application has to properly set <tt>_NET_WM_PID</tt>. This is acceptable and
\r
1646 patches can be written for the few applications which don’t set the hint yet.
\r
1651 It does only work on Linux, since cgroups are a Linux-only feature. Again,
\r
1652 this is acceptable.
\r
1657 The main problem is that some applications create X11 windows completely
\r
1658 independent of UNIX processes. An example for this is Chromium (or
\r
1659 gnome-terminal), which, when being started a second time, communicates with
\r
1660 the first process and lets the first process open a new window. Therefore, if
\r
1661 you have a Chromium window on workspace 2 and you are currently working on
\r
1662 workspace 3, starting <tt>chromium</tt> does not lead to the desired result (the
\r
1663 window will open on workspace 2).
\r
1667 <div class="paragraph"><p>Therefore, my conclusion is that the only proper way of fixing the "window gets
\r
1668 opened on the wrong workspace" problem is in the application itself. Most
\r
1669 modern applications support freedesktop startup-notifications which can be
\r
1670 used for this.</p></div>
\r
1675 <div id="footnotes"><hr /></div>
\r
1676 <div id="footer" lang="de">
\r
1677 © 2009-2011 Michael Stapelberg, <a href="/impress.html">Impressum</a>
\r
projects / i3 / i3.github.io / blob