BASH_ENV causes all non-interactive shells called via eg. /etc/bashrc to
remove the rc-file before the main shell gets to run it. Completion
scripts will often do this. Fixes#976.
Adapted from and fixes#1034.
* exwm-floating.el (exwm-floating--set-floating):
* exwm-workspace.el (exwm-workspace-move-window):
Buffers visible on other frames should be treated as invisible. One
side effect is visible buffers on the current frame is also taken into
account.
This is a small change that improves the behavior of
`exwm-workspace-move-window` in the following situation:
0. `exwm-workspace-show-all-buffers` and `exwm-layout-show-all-buffers`
are `nil`*.
1. On active workspace `i`, there is X window `a` in the selected Emacs
window.
2. On workspace `j`, there is X window `b` in the selected Emacs window
on that workspace frame.
3. While workspace `i` is active, use `exwm-workspace-move-window` to
move `a` to workspace `j`.
4. Switch to workspace `j` and use `exwm-workspace-move-window` to move
`a` back to workspace `i`.
Expected behavior: X window `a` is once again shown in the selected
Emacs window on workspace `i` and X window `b` is once again shown in
the selected Emacs window on workspace `j`.
What is observed: `a` is OK but the selected Emacs window on workspace
`j` does not show `b`. However, `b` is the first candidate when doing a
`switch-to-buffer` in that Emacs window on workspace `j`.
I'm not sure if this is the correct and complete change required, but it
is working well so far.
*The expected behavior is observed with EXWM 0.10 if
exwm-{workspace,layout}-show-all-buffers are non-nil.
This fixes an assertion failure in "assert(goal);" in
Worker::waitForInput() after a substitution goal is cancelled by the
termination of another goal. The problem was the line
//worker.childTerminated(shared_from_this()); // FIXME
in the SubstitutionGoal destructor. This was disabled because
shared_from_this() obviously doesn't work from a destructor. So we now
use a real pointer for object identity.
The implementation of "partition" in Nixpkgs is O(n^2) (because of the
use of ++), and for some reason was causing stack overflows in
multi-threaded evaluation (not sure why).
This reduces "nix-env -qa --drv-path" runtime by 0.197s and memory
usage by 298 MiB (in non-Boehm mode).
Normally it's impossible to take a reference to the function passed to
callFunction, so some callers (e.g. ExprApp::eval) allocate that value
on the stack. For functors, a reference to the functor itself may be
kept, so we need to have it on the heap.
Fixes#1045
The inner lambda was returning a SQLite-internal char * rather than a
std::string, leading to Hydra errors liks
Caught exception in Hydra::Controller::Root->narinfo "path âø£â is not in the Nix store at /nix/store/6mvvyb8fgwj23miyal5mdr8ik4ixk15w-hydra-0.1.1234.abcdef/libexec/hydra/lib/Hydra/Controller/Root.pm line 352."
* exwm-input.el (exwm-input-toggle-keyboard): New command for toggling
keyboard mode.
* exwm-layout.el (exwm-layout-toggle-fullscreen): New command for
toggling fullscreen mode.
* exwm-core.el (exwm-mode-menu, exwm-mode-map): Use them.
That is, unless --file is specified, the Nix search path is
synthesized into an attribute set. Thus you can say
$ nix build nixpkgs.hello
assuming $NIX_PATH contains an entry of the form "nixpkgs=...". This
is more verbose than
$ nix build hello
but is less ambiguous.
