Re: Has lexical binding declaration format changed in Emacs 30 or 31?

On Wed, Feb 25, 2026 at 12:50 PM <[email protected]> wrote:

> On Wed, Feb 25, 2026 at 12:17:47PM -0500, [email protected] wrote:
> > Pierre Rouleau <[email protected]> writes:
>
> > > ;;; gnu-elpa-keyring-update.el --- Update Emacs's GPG keyring for GNU
> ELPA
> > >  -*- lexical-binding: t; -*-
> > >
>

Note that the original line is not wrapped.  It was wrapped by email.


> - I see a newline in the Pierre's example. I assumed it was just MUA
> wrapping,
>   but worth checking anyway
>

There's no new line in the original line.  Its MUA wrapping induced as you
guessed.


> - Pierre's example has a semicolon after the value (i.e. t;). This should
> be
>   harmless (and local testing confirms that), since it'd be a separator for
>   more buffer-local variables, but who knows...
>

I tried removing the semicolon and byte compiling.  There were no warnings.
I added it back and byte compile again: still no warning.

The warning showed up when I first installed the file.  And now it does not
appear....

So I cd to the directory where the files were installed and
I performed the following command (show with the leading +):

 + emacs -Q --batch -L . --eval '(setq byte-compile-error-on-warn t)' -f
batch-byte-compile gnu-elpa-keyring-update.el
 Skipping update: Symbol’s function definition is void:
gnu-elpa-keyring-update

 + emacs -Q --batch -L . --eval '(setq byte-compile-error-on-warn t)' -f
batch-native-compile gnu-elpa-keyring-update.el
 Skipping update: Symbol’s function definition is void:
gnu-elpa-keyring-update

I'm now wondering why the compilations generate the "Skipping update:...."
message....

I did not write that file.
It's from https://elpa.gnu.org/packages/gnu-elpa-keyring-update.html

Perhaps Stephan can help?

-- 
/Pierre