| Git Clone URL: | https://aur.archlinux.org/ashuffle.git (read-only, click to copy) |
|---|---|
| Package Base: | ashuffle |
| Description: | Automatic library-wide shuffle for mpd. |
| Upstream URL: | https://github.com/joshkunz/ashuffle |
| Licenses: | MIT |
| Submitter: | oi_wtf |
| Maintainer: | oi_wtf |
| Last Packager: | oi_wtf |
| Votes: | 7 |
| Popularity: | 0.000000 |
| First Submitted: | 2019-09-20 18:25 (UTC) |
| Last Updated: | 2024-08-23 10:30 (UTC) |
If I go to the git archive and build according to the instructions, I get a fully built binary:
git clone https://github.com/joshkunz/ashuffle.git
git checkout v3.5.0
git submodule update --init --recursive
meson -Dbuildtype=release build
ninja -C build
I think the PKGBUILD file needs to be adjusted somehow.
I finally was able to get the pre-requisites set up to build this. I am trying to build for RPi (armv6h and armv7h). I was able to build the 2.0.1, but cannot build 3.4.0. I also tried 3.5.0 (latest release from the git repository) and get the same result. In case anyone is curious, I had to build the abseil-cpp on x86-64 (it is not arch dependent and requires more memory to build than the RPIb3+ has) and build and install the 2 packages generated by g-test. When I try to build this, though, I get this:
Found ninja-1.10.0 at /usr/bin/ninja
ninja: Entering directory `builddir'
[19/49] Linking target rule_test
FAILED: rule_test
c++ -o rule_test 'rule_test@exe/src_ashuffle.cc.o' 'rule_test@exe/src_load.cc.o' 'rule_test@exe/src_args.cc.o' 'rule_test@exe/src_getpass.cc.o' 'rule_test@exe/src_rule.cc.o' 'rule_test@exe/src_shuffle.cc.o' 'rule_test@exe/t_rule_test.cc.o' -Wl,--as-needed -Wl,--no-undefined -pie -march=armv7-a -mfloat-abi=hard -mfpu=vfpv3-d16 -O2 -pipe -fstack-protector-strong -fno-plt -Wl,-O1,--sort-common,--as-needed,-z,relro,-z,now -Wl,--start-group -labsl_raw_logging_internal -labsl_int128 -labsl_str_format_internal -labsl_strings_internal -labsl_strings -labsl_hash -labsl_city -pthread /usr/lib/pkgconfig/../../lib/libgtest.so -lpthread /usr/lib/pkgconfig/../../lib/libgmock.so /usr/lib/pkgconfig/../../lib/libgmock_main.so -Wl,--end-group
/usr/bin/ld: /usr/lib/gcc/armv7l-unknown-linux-gnueabihf/9.3.0/../../../libabsl_str_format_internal.a: error adding symbols: file format not recognized
collect2: error: ld returned 1 exit status
distcc[8073] ERROR: compile (null) on localhost failed
[24/49] Compiling C++ object 'shuffle_test@exe/t_shuffle_test.cc.o'
ninja: build stopped: subcommand failed.
==> ERROR: A failure occurred in build().
Aborting...
It is possible if I started using VCS sources. But when the release tarball is used, it is not that easy.
Since the current approach is not broken and libtap does not update very often, I wont change the source from the tarball to git for this package. (If you want git, take a look at ashuffle-git )
libtap could be included via a git submodule instead, see: https://wiki.archlinux.org/index.php/VCS_package_guidelines#Git_Submodules
Pinned Comments
oi_wtf commented on 2023-02-25 13:46 (UTC) (edited on 2023-02-25 13:49 (UTC) by oi_wtf)
As the upstream Release Notes state, most recent releases were rebuilds of the prebuilt upstream binaries only, so they get updated versions of the staticly linked dependencies. This PKGBUILD does not use those binaries, it builds from source and links the dependencies shared, so it does not need to be updated every time abseil, yaml-cpp or mpdclient or their dependencies get minor updates.
So that's usually the reason for not updating this PKGBUILD to every release done upstream. It's to avoid unnecessary rebuilds. I'll update when actual source changes are done or if someone explicitly flags o-o-d.
Please do feel free to flag (or comment), even if it may be unnecessary, if you want me to update the PKGBUILD anyway or if you think I've overlooked some changes. I've got no problem with that.