- Add a 27.x check, recommending that users upgrade to 28.1.
- Removed the check for <27. Loading doom.el with sub-27 will already
fail with a detailed error about supported versions.
- Now detects development (.50) and pre-release (.9x) builds of Emacs,
and warns the user of their dangers.
- Revises the warning for 29+ or .50|.9x users.
It's not useful in those scenarios, and is more likely to throw up
unrecoverably keybind conflict errors, when load order is determined by
an over-eager byte-compiler, rather than Doom's standard startup
process.
The debugger appears to display uninterned symbols properly, and since
these symbols should never be touched/referenced by users, there's no
reason to pollute the obarray with these transient symbols.
This file used to live in /.github/ci.el. It was later moved to
/.doomrc (in 9b8ed39), but I forgot to push this change to update the
relative path to Doom's modules. This caused the linter to complain that
all module scopes were invalid.
Also, I removed the $DOOMDIR/modules/ check, as modules in the user's
config shouldn't be factored into the list of valid scopes.
Amend: 9b8ed397e8
Rather than reimplement its face lookup (and have two versions of
doom-print-ansi-alist -- one for 27.x and one for 28+), let's just rely
on ansi-color.
This adds the basic framework of docs/examples.org, including the former
contents of demo.org in :lang emacs-lisp. elisp-demo has also been
reconfigured to search it instead.
Keep in mind that examples.org references a few things in as-of-yet
published documentation. This will be rectified soon.
This was adapted from
https://www.reddit.com/r/emacs/comments/d7x7x8/finally_fixing_indentation_of_quoted_lists/.
It fixes the indentation of quoted data (and plist keywords) so they're
indented like data, rather than function arguments, like so:
BEFORE:
`(foo bar baz
doom emacs)
'(:foo 1
:bar 2
:baz 3)
'(:foo 1
2
3
:bar 4)
(:foo 1
:bar 2)
(:foo 1
;; test comment
:bar 2)
(:foo 1
2
:bar 3)
AFTER:
`(foo bar baz
doom emacs)
'(:foo 1
:bar 2
:baz 3)
'(:foo 1
2
3
:bar 4)
;; only align unquoted keywords if keywords start each line:
(:foo 1
:bar 2)
(:foo 1
;; test comment
:bar 2)
(:foo 1
2
:bar 3)
Also, I added a way to declare that plists in an macro's arguments
should be indented like data:
(put 'map! 'indent-plists-as-data t)
BEFORE:
(map! :localleader
:map emacs-lisp-mode-map
(:prefix ("d" . "debug")
"f" #'+emacs-lisp/edebug-instrument-defun-on
"F" #'+emacs-lisp/edebug-instrument-defun-off))
AFTER:
(map! :localleader
:map emacs-lisp-mode-map
(:prefix ("d" . "debug")
"f" #'+emacs-lisp/edebug-instrument-defun-on
"F" #'+emacs-lisp/edebug-instrument-defun-off))
There was a third improvement I was hoping to include, namely,
proper indentation of interpolated forms:
BEFORE:
`(foo
bar
,(if t
'baz
'boo))
`(foo
bar
(if t
baz
boo))
AFTER:
`(foo
bar
,(if t
'baz
'boo))
`(foo
bar
(if t
baz
boo))
But this was removed because it breaks indentation for quoted macro
forms (or dynamic elisp programming):
BEFORE: (good)
`(with-temp-buffer
(if (always)
(message
"Hello %s"
user-login-name)
(message
"Goodbye %s"
user-login-name)))
AFTER: (bad)
`(with-temp-buffer
(if (always)
(message
"Hello %s"
user-login-name)
(message
"Goodbye %s"
user-login-name)))
Ref: https://www.reddit.com/r/emacs/comments/d7x7x8/finally_fixing_indentation_of_quoted_lists/'
These only benefit interactive sessions, and doom-start's responsibility
is to configure interactive sessions; it doesn't make sense to keep
these in core.
This refactors how Doom captures and redirects its output (to stdout and
stderr) into a more general with-output-to! macro, and:
- Simplifies the "print level" system. The various doom-print-*-level
variables have been removed.
- Adds a new print level: notice, which will be the default level for
all standard output (from print!, doom-print, prin[ct1], etc).
- Adds a with-output-to! macro for capturing and redirecting
output to multiple streams (without suppressing it from stdout). It
can also be nested.
- Changes the following about doom-print:
- Default :format changed to nil (was t)
- Default :level changed to t (was `doom-print-level`)
- No longer no-ops if OUTPUT is only whitespace
This commit reduces the debug log noise, makes it easier to
read/parse/search, and soft-introduces a convention for doom-log
messages, where they are prefixed with a unique identifier loosely named
after it's running context or calling function.
I haven't enforced it everywhere doom-log is used yet, but this is a
start.
BREAKING CHANGE: This removes the doom-incremental-load-immediately
variable. Instead, set doom-incremental-first-idle-timer to 0 to force
all iloaded packages be eagerly loaded at startup. This is already the
default behavior for daemon sessions.
BREAKING CHANGE: If anyone is using Doom's CLI framework and are
defining their own CLIs with any of the following macros, they'll need
to be updated to their new names:
- defautoload! -> defcli-autoload!
- defgroup! -> defcli-group!
- defstub! -> defcli-stub!
- defalias! -> defcli-alias!
- defobsolete! -> defcli-obsolete!
These were renamed to make their relationship with CLIs more obvious;
they were too ambiguous otherwise.
BREAKING CHANGE: This performs the following backwards-incompatible
changes:
- Replaces `+emacs-lisp-reduce-flycheck-errors-in-emacs-config-h` with a
`+emacs-lisp-non-package-mode` minor-mode.
- Removed the `+emacs-lisp-disable-flycheck-in-dirs` variable, as this
mechanism no longer checks a directory list to detect a "non-package".
If you've referenced either of these symbols, you'll need to
update/remove them from your config. No extra config is needed
otherwise.
Why: Doom has always tried to reduce the verbosity of Flycheck when
viewing elisp config files or scripts (i.e. non-packages). These are so
stateful that the byte-compiler, package-lint, and checkdoc inundate
users with false positives that are more overwhelming than helpful.
The heuristic for this has always been a simple "is this file in
$DOOMDIR or $EMACSDIR", but this wasn't robust enough, especially in
cases where symlinking was involved, so I've employed a new, more
general heuristic for detecting non-package files:
- The file isn't a theme in `custom-theme-load-path`,
- The file doesn't have a (provide ...) or (provide-theme ...)
statement whose first argument matches the file name,
- The file lives in a project with a .doommodule file (doom modules
never have convention package files in them),
- Or the file is a dotfile (like .dir-locals.el or .doomrc).
I've also tweaked byte-compile-warnings to yield a little more output,
but not by much. Whether this is too permissive or not will require
further testing to determine.
What's more, I've updated this to reflect recent changes to Doom's
startup process (in c05e615).
Ref: c05e61536e
While fixing #6772, we also address several other issues found by
sh-shellcheck:
L47C39 SC1007:Remove space after = if trying to assign a value
(for empty string, use var='' ... ).
L47C46 SC2046:Quote this to prevent word splitting.
L62C20 SC2086:Double quote to prevent globbing and word splitting.
L82C9 SC2154:tmpdir is referenced but not assigned
(did you mean 'TMPDIR'?).
When trying to establish the value of EMACSDIR, the proper fallback
when BASH_SOURCE is undefined is $0 rather than 0, in correct shell-
scripting parlance.
With mu4e came a new face for the header view. While this is nice for
the text, it had the consequence of making the account stripe (the
vertical bar) a bit jarring. This patch fixes that behavior to make the
stripe non-italic by append `'default` to the face.
The nix-mode package already does this and it shadows other entries for the .drv files in auto-mode-alist (namely Guix derivations).
Ref: 34d51e2731/nix-drv-mode.el (L48)
Necessitated by 7e0c2ed, and missed in 45a66cd. This would indirectly
cause "No :repo specified for package 'X'" errors for packages with a
`:local-repo` relative to their packages.el file.
Amend: 45a66cda60
Ref: 7e0c2ed8a3
In v3, doom-module data is stored in symbol plists, but in v2, it's
stored in a hash table. Some v3 code snuck into 45a66cd, which made Doom
try to read module data from plists that hadn't been initialized yet, so
Doom could no longer see your module settings.
Fix: #6769
Amend: 45a66cda60
Otherwise, doom-module-from-path (and modulep!) would fail to detect the
module they're in, or at least, modulep! would incorrectly return nil,
even for enabled modules.
This issue is what would've caused the package list or the doctor to
include/consider packages in disabled modules or behind disabled flags.
Writing a debugger for Elisp is too much hassle. `debug` itself isn't
very customizable without a *lot* of boilerplate, so instead of writing
my own, it's more effective to advise debug instead. Certainly, I don't
do anything with it yet, but I will soon.
Emacs 27.x does not collapse consecutive slashes in a file path when
trying to load them, and instead discards everything before it and
treats the rest as an absolute path, e.g. "~/some//path/foo/" ->
"/path/foo". This is not the case in 28.1, but Doom's backport of
file-name-concat did not take this into account, so it's been modified
to trim trailing slashes.
Fix: #6766
Amend: 433c9e344d
This is caused by a bug in recent builds of Emacs 29, where
`loaddefs-generate` will activate emacs-lisp-mode to read a package's
autoloads, but does so without suppressing its mode hooks. Other
packages may add functions to this hook from their autoloads (like
overseer.el does). Calling these functions will initiate a chain
reaction where other packages will be loaded (plus their dependencies),
but aren't guaranteed to be available so early in the bootstrap process.
The result are file-missing errors about seemingly unrelated packages,
like pkg-info or dash.
Ref: emacs-mirror/emacs@0d383b592c
Fix: https://discourse.doomemacs.org/t/3149
Occurs when a site-file fails to be natively compiled, and Doom attempts
to write an error file in the same directory. On some systems, the
site-lisp directory is in a read-only tree/mount (like nix and guix).
This should suppress those attempts.
Relying on eval-after-load's compiler-macro magic (after 8b4f722) can
cause scope issues with nested macros (like file! and dir! running in
the context of a temp buffer), so best we use with-eval-after-load
directly.
Ref: 8b4f722fa3
To ensure that they're expanded at a file's top-level, while expanded,
where they're used. It also fixes a few inlined uses of the file!
macro (e.g. in `load!`, as reported in #6764), which was prematurely
committed ahead of this change.
Close: #6764
Amend: a179b8d262
b7f84bd introduced a nasty regression that caused an infinite loop and
runaway memory usage on some pgtk+native-comp builds of Emacs when it
attempted to perform deferred native compilation of your packages. This
would make Emacs unusable and, if left alone, could even crash your
system.
The only Emacs builds I'm certain are affected are derived from
flatwhatson/emacs (like emacs-pgtk-native-comp on Guix and Arch Linux in
particular). 28.1 stable and master (on emacs-mirror/emacs@e13509468b)
are unaffected.
It appears that some, earlier pgtk builds stack idle timers differently.
I'm not entirely sure how, because it doesn't manifest in more recent
builds of Emacs, and I'm already burnt out on debugging this, but here's
how Doom encountered it:
Doom has an incremental package loader; it loads packages, piecemeal,
after Emacs has been idle for 2s, then again every 0.75s until it
finishes or the user sends input (then it waits another 2s before
starting again). However, if at any time the iloader detected that
native-compilation is in progress, it waits 2s before trying
again (repeat, until native-comp is done). But here's the catch, given
the following:
(run-with-idle-timer
2 nil (lambda ()
(run-with-idle-timer 1 nil (lambda () (message "hi")))))
I had assumed "hi" would be emitted after 3 seconds (once idle), but
instead it is emitted after 2. Like this, Doom's iloader would elapse
one idle timer directly into another, ad infinitum, until Emacs was
forcibly killed.
By switching to run-at-time and employing my own rudimentary idle timer,
I avoid this issue. Also, the iloader no longer needs to be considerate
of native-comp, because the latter does its own rate-limiting controlled
by native-comp-async-jobs-number.
Amend: b7f84bdd01
