- move the magit, helm and other dependencies to their own init
functions
- remove all nested use-package forms and enable correct lazy-loading
of eshell
- import interesting features from trishume like auto-completion and
automatic jump to prompt when switching to insert state
- new layer variable `shell-enable-smart-eshell` to conditionally enable
`em-smart` (default is not enabled)
- document all the last additions to the layer in README.org
magit-log-all has the same with the old magit-log: show the commit log
immediately without asking for revision. We should use magit-git-log for
smoother transition to magit 2.1.
Calling `magit-blame-mode` directly is not advised as of Magit 2.1.0,
and will leave a buffer in read-only mode upon exit. `magit-blame` also
allows recursive blame upon repeated usage.
Rename the layer to idris-lang because idris-packages variable is used
by the package
Update key bindings to better fit spacemacs conventions
Convert markdown README to org
This removes idris-mode from the extra-langs layer and moves it to its
own layer with custom evil-leader keybindings. It also removes
documentation in extra-langs on other removed languages.
There are two major complications with idris-mode: it defines its own
evil-leader bindings which overwrite Spacemacs defaults, and it uses
the variable idris-packages, which Spacemacs uses to include the
idris-mode package, to define Idris packages to load for the current
buffer.
- Use compile instead of shell for cargo. This allow Emacs to parse the
output, including all the errors :)
- Added support for Racer (it works for me at least... hopefully for
others too)
Signed-off-by: JP-Ellis <coujellis@gmail.com>
This commit drastically simplify the previous commit by removing
the layer variable syntax-checking-check-on-buffer-change and
the associated usage of it.
This effectively make spacemacs use the default flycheck setting.
If the user wants to customize this behavior then she can modify
directly the flycheck variable.