A final 'force push' to PR #15574 implemented functionality to make the banner
size depend on the actual 'startup list' instead of the configured maximum
startup list size. However, that 'new' functionality uses
`spacemacs-buffer//do-insert-startupify-lists` which activates `recentf-mode`,
and unfortunately, activating `recentf-mode` this early in the startup process
somehow breaks recentf functionality, see comments at
a485b5a84b.
This commit 'reverts' the functionality by making the banner size depend on the
configured maximum startup list size.
The default logo is rather big. Users might prefer to use the screen size for
other information (while still showing a banner).
This commit adds a configuration variable `dotspacemacs-startup-banner-scale`
to the dotfile
The script used to identify and update the change is added into the GitHub
workflows script directory. A workflow action can be created to trigger the
script to update the headers on the first of every new year. Possibly a task for
a consequent PR.
with this curent (package :location local) can be replaced with (package
:location (recipe :fetcher local)) then quelpa will install local pacakge. We
have the benefit of bytecompile and autoload
[packages] use new fetcher for local
- Remove the two functions that add directories to the load path.
- Inline the body forms of the functions to be removed.
- Rename the variables to be more explicit.
Fix let varlist in core-load-path.el
- When reporting issues on GitHub, it will copy the issue body to clipboard and
ask user to manually paste the content to the web browser, when the resulting
URL become too long.
- Cleaning up.
- Suppressed all compiler warnings.
- Re-ordered all top-level forms into various sections according to its functionality.
First, change read syntax for anonymous functions currently written '(lambda to
instead just be (lambda; otherwise this raises a warning.
Next, while we're at it, change all instances of #'(lambda to just (lambda,
since the two are equivalent and the latter is more concise (excepting anything
in core/libs or in /local/ subdirs of layers)
https://www.gnu.org/software/emacs/manual/html_node/elisp/Anonymous-Functions.html#index-_0023_0027-syntax
in #15421 core cannot find a certain text in the home buffer
to start adding additional elements. This did fail the
start up process.
Now if the starting point is not found the cursor is set
to the end of the buffer instead and no error is signaled.
This allows to still have the home buffer work as expected
even when the starting point has been somehow removed.
- Fixed a bug that when `all-the-icons` is excluded by user, Emacs reports
that using undefined `all-the-icons` functions.
- Changed the `spacemacs-buffer-mode` that now it derives from `special-mode`.
- Also defined a new command `spacemacs-buffer/return` which binds to `RET`
key in `spacemacs-buffer-mode`. It opens the button on the current line if
there's any, or move the cursor to next line.
For example, in `rust-mode`, `(spacemacs//get-jump-handlers)` returns
`((lsp-ui-peek-find-definitions :async t) evil-goto-definition)`.
So `-handler` is bound to the first one, i.e. `(lsp-ui-peek-find-definitions :async t)`.
We need to figure out whether it's asynchronous by checking the `:async` keyword argument.
But before this commit, we wrongly checks the second handler in the handlers list.
Even when multiple handlers do exist and they are all async, the typo made it sub-optimal since it uses not the best handler.
This commit fixes that.
This commit locks the spacemacs home buffer in read-only mode.
Before this, users may use `SPC b w` to toggle it to writable but
this buffer really should not be modified by users.
Note that, when we need to modify the home buffer in LISP program,
we can always set `inihibit-read-only` to nil. Thus this commit
won't disallow us to update the home buffer when it's needed.
Move closing parens on all dotspacemacs/user-* functions to a new line so the
function documentation and closing paren are not part of the config diff when a
user runs `SPC f e D`. This should make it faster to check config changes versus
the template.
