When "something goes wrong" in cairo-land, the corresponding cairo
object goes into an error state. These errors are sticky. Thus, it is
enough to check for errors before destroying the context.
This commit adds a check in draw_util_surface_free() to check the cairo
context's status and print a log message if anything is wrong.
The idea here is to help debugging drawing issues. Instead of "nothing
visible", the corresponding log message hopefully helps debugging.
This code would have saved me lots of time in figuring out why my pull
request #4379 did not work.
Signed-off-by: Uli Schlachter <psychon@znc.in>
When adding/removing a monitor, the outputs are likely to be modified.
Send an IPC event "output", like when there is a screen configuration
change.
Signed-off-by: Vincent Bernat <vincent@bernat.ch>
Fixes#4464.
Inspired by changes from #4453 but I didn't end up using the original
code.
For ALIGN_LEFT and ALIGN_CENTER icon offset is calculated first and then
title offset is based on that. For ALIGN_RIGHT title offset is
calculated first but we make sure that the icon fits in the window.
Additionally, icons will now hide if there is not enough space to render
them + their padding. Realistically, this won't happen in normal
usecases.
Effectively, icons have higher precedence over text: draw_util_text will
hide text when it exceeds the max width and the code takes this into
consideration. Icons will only hide when they can't fit at all, not in
order to assure that the text can be displayed fully.
This is useful if we want to be able to start some services depending
on i3, notably some script using the socket or third-party bars like
polybar. To make use of this change, user is expected to use the
following stanza:
```
[Unit]
Description=i3 window manager
PartOf=graphical-session.target
[Service]
Type=notify
ExecStart=/usr/bin/i3
ExecStopPost=/bin/systemctl --user stop graphical-session.target
Restart=on-failure
```
Something similar is already possible using socket activation. For
example, we could use:
```
[Unit]
Description=i3 window manager socket
PartOf=graphical-session.target
[Socket]
ListenStream=%t/i3.sock
ExecStartPost=/bin/systemctl --user set-environment I3SOCK=%t/i3.sock
```
And other units could `Requires=i3.socket`. Unfortunately, not
everything is using I3SOCK. Notably, `i3 --get-socketpath` does not
and that's what i3ipcpp is doing. An alternative would be to patch `i3
--get-socketpath` to use I3SOCK if present, however, this may be a bit
risky. Should we check the environment variable first or the root
attribute?
Another alternative not requiring any modification is to have a
dedicated `i3-session.target`:
```
[Unit]
Description=i3 session
BindsTo=graphical-session.target
Wants=wallpaper.service
Wants=wallpaper.timer
Wants=polybar.service
Wants=i3-companion.service
```
And trigger it from i3:
```
exec_always --no-startup-id systemctl --user start --no-block i3-session.target
```
The proposed change being quite small, it seems harmless and
low-maintenance.
Signed-off-by: Vincent Bernat <vincent@bernat.ch>
Commands with more than 10 words would result in i3 crashing. The root
cause is that the stack which holds arguments with identifiers is
only 10 big, but generate-command-parser.pl was giving every word an
identifier of "".
This criterion matches all open windows, as a more readable (and
correct) version of the 'class=".*"' criterion (more correct because
it'll also match windows which don't have WM_CLASS set yet).
This feature defaults to off, and can be turned on for individual windows,
or (with for_window) for all new windows. See the userguide change.
This commit is partially based on work by:
• Marius Muja
• mickael9
• Esteve Varela Colominas
• Bernardo Menicagli
The implementation uses wordexp(3) just like sway:
https://github.com/i3/i3/issues/1197#issuecomment-226844106
Thanks to jajm for their implementation at
bb55709d0a
This required refactoring the config parser to be re-entrant
(no more global state) and to return an error instead of dying.
In case a file cannot be opened, i3 reports an error but proceeds with the
remaining configuration.
Key bindings can be overwritten or removed using the new --remove flag of the
bindsym/bindcode directive.
All files that were successfully included are displayed in i3 --moreversion.
One caveat is i3 config file variable expansion, see the note in the userguide.
fixes#4192
This helps write correct descriptions of the JSON schema for strongly
typed languages that support deserializing to native types (and
require a different type for null).
These seemed to be inconsistently specified.
Particularly, the "SYNC" message payload was not described anywhere.
Even when the payload is empty, it is helpful to specify that
explicitly, as it prevents the reader from having to guess whether if
it's really empty, or otherwise somehow implicitly obvious.
The "Reply format" section is now "Messages and replies", and covers
both the reply format and the format of sent messages.
- Make sure to place the description of the packet before the
description of the payload.
- Describe the relationship of messages and replies.
- Add note on pipelining.
xcb_flush() flushes xcb's output buffer. Thus, after this call, all
previous requests are surely written to the connection with the X11
server.
xcb_aux_sync() synchronises with the X11 server. It makes sure all
previous requests were sent to the X11 server and already processed. It
does this via free(xcb_get_input_focus_reply(xcb_get_input_focus())): It
sends a request and waits for its reply. It is guaranteed that the X11
server processes requests in-order.
This means that the sequence xcb_flush(); xcb_aux_sync() first writes
whatever is in the output buffer and then does another write for the
four bytes of the GetInputFocus request from xcb_aux_sync().
Put differently: xcb_flush(); xcb_aux_sync() doesn't do anything more
than just xcb_aux_sync(). In fact, it has slightly more overhead for the
same result.
Thus, this commit drops all calls to xcb_flush() immediately after
xcb_aux_sync().
Signed-off-by: Uli Schlachter <psychon@znc.in>
"Set" the wallpaper during startup only sometimes
Since commit 4f5e0e7, i3 would take a screenshot and set that as the
background pixmap of the root window during startup. This is the easy
part of setting a proper X11 wallpaper.
The code in question was added because something either set the
background pixmap of the root window to NONE or the X11 server was
started with "-background none". This is apparently done by default by
e.g. gdm to avoid some flickering while the X11 server starts up.
This commit makes this code conditional: Only when no wallpaper is
detected is a screenshot taken.
Since I could not find any way to query the background of a window, a
more direct approach is taken to detect this situation: First, we find
some part of the root window that is not currently covered. Then we open
a white window there, close it again and grab a screenshot. If a
wallpaper is set, the X11 server will draw this wallpaper after the
window is closed and something else will be visible in the screenshot.
However, the wallpaper could have a white pixel at the tested position.
Thus, this procedure is repeated with a black window.
Only when this procedure produces two different pixel values is a
screenshot taken and set as the wallpaper.
Fixes: https://github.com/i3/i3/issues/4371
Fixes: https://github.com/i3/i3/issues/2869
Signed-off-by: Uli Schlachter <psychon@znc.in>
