Arch Linux User Repository

New version prefers system packages for libfmt and CLI11

when available/found, mu now (by default) uses the system versions of CLI11 and libfmt, rather than mu's "vendored" versions.

Please add them to dependencies.

diff --git a/PKGBUILD b/PKGBUILD
index 198f6a3..53f5ccc 100644
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -9,8 +9,8 @@ pkgdesc="Maildir indexer/searcher and Emacs client (mu4e)"
 arch=("x86_64" "armv7h" "aarch64")
 url="http://www.djcbsoftware.nl/code/mu"
 license=("GPL-3.0-or-later")
-depends=("gmime3" "xapian-core" "readline" "glib2")
-makedepends=("emacs" "meson" "guile")
+depends=("gmime3" "xapian-core" "readline" "fmt" "glib2")
+makedepends=("emacs" "meson" "guile" "cli11")
 optdepends=(
        "emacs: mu4e support"
        "guile: guile bindings"

done

@mroethke, could you add aarch64 to the arch array? Much appreciated.

New version update:

pkgname=mu pkgver=1.12.11 pkgrel=1 pkgdesc="Maildir indexer/searcher and Emacs client (mu4e)" arch=("x86_64" "armv7h") url="http://www.djcbsoftware.nl/code/mu" license=("GPL-3.0-or-later") depends=("gmime3" "xapian-core" "readline" "glib2") makedepends=("emacs" "meson" "guile") optdepends=( "emacs: mu4e support" "guile: guile bindings" ) source=("https://github.com/djcb/mu/releases/download/v${pkgver}/mu-${pkgver}.tar.xz") sha256sums=('c6e463f4936edc7f2c0a0b0d038183783b35430b35887b79301072bf1d5e9daf')

aarch64 should also be added to the list of architecture as I've been using mu on an ampere processor for years.

Oh I did not notice that guile3 landed in the official repos and it has only been like 2 years … Thanks for the hint!

Will do a quick check myself and push an update if everything goes well.

FYI, out of sheer curiosity, I tested what happens if I re-enable guile, and it did actually compile successfully. I haven't tested whether any guile-related function works or anything, but previously (back in version 1.8) it wouldn't even compile. I'm not trying to suggest that you necessarily should reenable it in the PKGBUILD, but doing so doesn't look like it'd cause problems anymore (although note that the repo package is now guile and not guile3, and of course I compile in a clean chroot).

Thanks, it worked pretty well,

but now I get this error, yesterday I updated the sha and worked

``` [154/212] Generating man/mu-bookmarks.5.org-to-man with a custom command FAILED: man/mu-bookmarks.5 /usr/bin/emacs --no-init-file --batch /home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-bookmarks.5.org --eval '(progn (require '"'"'ox-man) (setq org-export-with-sub-superscripts '"'"'{}) (org-export-to-file '"'"'man "/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-bookmarks.5"))' Select coding system (default utf-8): Error: end-of-file ("Error reading from stdin") mapbacktrace(#f(compiled-function (evald func args flags) #<bytecode 0x1297608829ae86a5>)) debug-early-backtrace() debug-early(error (end-of-file "Error reading from stdin")) completing-read-default("Select coding system (default utf-8): " (("georgian-academy") ("georgian-academy-mac") ("georgian-academy-dos") ("georgian-academy-unix") ("georgian-ps") ("georgian-ps-mac") ("georgian-ps-dos") ("georgian-ps-unix") ("cp1256") ("cp1256-mac") ("cp1256-dos") ("cp1256-unix") ("windows-1256") ... ... select-safe-coding-system-interactively(1 68 (utf-8 chinese-gb18030 utf-7 utf-16 utf-16be-with-signature utf-16le-with-signature utf-16be utf-16le compound-text-with-extensions compound-text iso-2022-7bit utf-8-auto utf-8-with-signature emacs-mule raw-text iso-2022-8bit-ss2 iso-2022-7bit-lock utf-7-imap utf-8-emacs prefer-utf-8 no-conversion ctext-no-compositions iso-2022-7bit-ss2) (iso-latin-1-unix) nil utf-8) select-safe-coding-system(1 68 iso-latin-1-unix nil) org-persist--normalize-associated(#<buffer *temp*-444396>) org-persist-read(org-element--cache #<buffer *temp*-444396> t t) org-persist-load(org-element--cache #<buffer *temp*-444396> t) org-mode() org-export--prepare-file-contents("/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/macros.inc" nil 0 1 1 #<hash-table equal 0/65 0x1571dbe4f6b9> "/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1.org") org-export-expand-include-keyword() org-export-as(man nil nil nil (:output-file "/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1")) org-export-to-file(man "/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1") (progn (require 'oEnd of file during parsing: Error reading from stdin x-man) (setq org-export-with-sub-superscripts '{}) (org-export-to-file 'man "/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1")) command-line-1(("/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1.org" "--eval" "(progn (require 'ox-man) (setq org-export-with-sub-superscripts '{}) (org-export-to-file 'man \"/home/toni/learn/setup/mu/src/mu-1.12.6/build/man/mu-info.1\"))")) command-line() normal-top-level() [165/212] Compiling C++ object mu/mu.p/mu-options.cc.o ninja: build stopped: subcommand failed. ==> ERROR: Ha ocorregut una fallada a build(). Es cancel·la..

sorry, should be fixed now.

Hi, thanks for maintaining this package, but I have one problem and I think that the shasum256 is not correct with the current mu package 1.12.6

https://github.com/djcb/mu/releases/

do not have coincidence in PKGBUILD

sha256sums=('770b92e3072adae8777bd4e985c8b027c66ce0b7ef7cbb8eab4497b92ba68acb')

and the one in the release

f8a539b687c999678fd7cd37cc4ab15ee5e87801027d982ba195b3a9cb53b761 *mu-1.12.6.tar.xz

and give this error:

    ~/learn/setup/mu    master ?1  LC_ALL=c makepkg -sic                                       INT ✘ 
/usr/bin/bash: warning: setlocale: LC_ALL: cannot change locale (c)
==> Making package: mu 1.12.6-1 (Mon Jul 29 09:19:23 2024)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Found mu-1.12.6.tar.xz
==> Validating source files with sha256sums...
    mu-1.12.6.tar.xz ... FAILED
==> ERROR: One or more files did not pass the validity check!