Arch Linux User Repository

Leaving this as a permanent disclaimer:

It is illegal to redistribute any binaries created with this script because of the upstream project's licensing.

Cannot build when debug is enabled in makepkg.conf. It will try to run the test and find the file in debug src folder but cannot find the file.

Ah, the problem is that in line 26 of the PKGBUILD the file is always downloaded just under the package name, so if you have an older version still in cache the name doesn't change and it will find that.

@g3tchoo, please include the $pkgver variable in the download name in line 26 (before the ::). Then this issue can't occur.

old: still broken sha512sum, please fix

as @g3tchoo has written before, after cleaning the cache for paru via

 paru -Sau --clean 

solves the problem

Fetching origin
  -> Found mmc-stable-lin64.tar.gz
  -> Found modern-java.patch
  -> Found 0001-Readd-lin-system-and-LAUNCHER_LINUX_DATADIR.patch
  -> Found fix-jars.patch
  -> Found mmc-brand.patch
  -> Found multimc.desktop
==> Validating source files with sha512sums...
    multimc5 ... Skipped
    libnbtplusplus ... Skipped
    quazip ... Skipped
    mmc-stable-lin64.tar.gz ... FAILED
    modern-java.patch ... Passed
    0001-Readd-lin-system-and-LAUNCHER_LINUX_DATADIR.patch ... Passed
    fix-jars.patch ... Passed
    mmc-brand.patch ... Passed
    multimc.desktop ... Passed
==> ERROR: One or more files did not pass the validity check!
error: failed to download sources for 'multimc5-0.6.14-5':

The sources for this package do not pass validation.

@g3tchoo, please update the checksums.

melvyn2, sorry about the wait! i've been busy with finals and hadn't got a notification for the Launcher repo updating, so i thought the package was good :/

should build now though!

Arisa, i haven't been able to reproduce this :( maybe try to clean the $srcdir before the build by running this: paru -Sau --clean

or you can just build without an aur helper for a sanity check:

git clone https://aur.archlinux.org/multimc5.git
cd multimc5/
makepkg -siCcr

Couldn't compile it because of a combination of warning and -Werror

Used clang as compiler and mold as linker

/home/user/.cache/paru/clone/multimc5/src/multimc5/launcher/ui/instanceview/InstanceView.cpp:839:9: error: variable 'beginning_row' set but not used [-Werror,-Wunused-but-set-variable] int beginning_row = 0; ^ 1 error generated.

Please fix your hashes...

diff --git a/PKGBUILD b/PKGBUILD
index 2af89ca..d759c99 100755
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -36,7 +36,7 @@ source=("${pkgname}::git+https://github.com/MultiMC/Launcher#tag=${pkgver}"
 sha512sums=('SKIP'
             'SKIP'
             'SKIP'
-            'df53f8b13ceeddc7922b6553730f29fdf92662277d20c9fc97d56965df1fe2349dd724ad8f911b105b02653fad112785e72ffadbf5571458fa23a45daf473e11'
+            '333cf8f504385d9e24a06cec12f3eb944b27773900595e6116cdd8e372d9ce564d53e4cf88f99b5d935345d2cf2a46309ea2431d86da6869236be17acd035b87'
             'ede87db7e1ce67184409123720b48f65d35884bac5dfff34cc099c5d7520fd9173be00997df82ad29da4640928134a7199ffdd1cbaa83a72101dbed72736508b'
             'd0e3152b885ec73b019dc0269d46e1ce0c43211bbcf05776a1e97f10fdc29f442930cc2d5144c0f979133e6aec68bf68474b924c746d1fa6b8c7f604af6eb83f'
             'df077c24a235117cc391ac515dd1e28c10b2082093c80aac5dcf3d6d8feaf07acacc79bfd5f111e389e7972e6f0a3b6eb217202d13e362219ed22a5437cd49c7'

@Anaeijon

I encountered the same problem. It did not end up being a problem with the package (modern-java.patch ensures it works). You need to be sure that archlinux-java is set to a version you have a jdk for. To test if this is your issue, try running file /usr/bin/javac. If it gives an error, you either have nothing or a JRE set in archlinux-java.

Building this with jdk-openjdk and jdk11-openjdk installed produced following error:

CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to
  the full path to the compiler, or to the compiler name if it is in the
  PATH.

Installing jdk8-openjdk fixed the problem and this AUR built and installed just fine.

I don't have much insight into JDK and don't know why using JDK >8 was a problem here, but maybe someone needs this advice.

Maybe the java-environment dependency should be more specific?

On multimc 0.6.14-3 I get an error

❯ multimc multimc: error while loading shared libraries: libLauncher_quazip.so: cannot >>open >shared object file: No such file or directory I assume it's related to https://github.com/MultiMC/Launcher/issues/4371

had the same issue... I just did multimc-bin and it worked... not a solution but a workaround !

On multimc 0.6.14-3 I get an error

❯ multimc
multimc: error while loading shared libraries: libLauncher_quazip.so: cannot open shared object file: No such file or directory

I assume it's related to https://github.com/MultiMC/Launcher/issues/4371

I had to remove the LIBRARY_DEST_DIR line from the PKGBUILD to get this to run

Leaving this as a permanent disclaimer:

It is illegal to redistribute any binaries created with this script because of the upstream project's licensing.

Hi all, thanks to more help from LennyLennington, this package should be able to be run without any extra flags or .desktop trickery, as well as having the previous feature parity with upstream that was in the last release (meaning microsoft accounts and everything should work fine). I also fixed the issue with the patches, as well as updated the formatting of them. Thank you everyone for your help and I hope you all enjoy!

g3tchoo: You need to change -DMultiMC_* to -DLauncher_* in the CMake defines since they were changed by upstream and the MultiMC_* will have no effect on the build, but since you have it working and those flags have no effect anyway you could just remove them.

I would recommend using the lin-system define though, that way you don't rely on the .desktop wrapper to make it function properly with the config files.

Yes, that's right. My problem is that I use bemenu to launch applications -- no .desktop files are used.

For those who use dmenu or bemenu as their application launcher, simply write a wrapper script to launch multimc with -d flag.

Hi aleck099, your problem was discussed just a few comments ago. MultiMC does not use $HOME/.local/share/multimc as its data folder by default anymore, you now have to specify it with the -d flag. Your old data is definitely not gone. This problem is fixed with the .desktop file in the newest AUR release. If it still doesn’t work, maybe your DE caches the desktop entries. In the case you launch the binary directly, you have to set the option yourself.

DO NOT INSTALL THIS ON YOUR PC!!! INSTALL multimc-bin INSTEAD!!!

By 2021-12-18

The package multimc5 is unable to launch. It prompts "The launcher couldn't create a log file - the data folder is not writable.". See this issue for more information.

What I found is that it tries to create log files in /usr/bin/ and get an error. The error still remains even if I comment it out. All your Minecraft instances and accounts went missing -- the broken multimc is unable to read them, saying permission denied.

That's definitely not normal. Looking at the patches, it looks like mmc-brand.patch tries to change a line dataPath = xdgDataHome + "/devlauncher"; to dataPath = xdgDataHome + "/multimc";, but that line was added in 0001-Readd-lin-system-and-LAUNCHER_LINUX_DATADIR.patch and already uses /multimc instead of /devlauncher. Entering y (or patching with -R) changes that line to use /devlauncher instead, because it makes patch undo the hunk.

hey autumnontape, you can just type y, then confirm to apply the patch and it should work fine. I don't know much about how the patch command works honestly, but I'll try to fix it in the next release.

I'm getting a prompt during the prepare step since the latest revision:

==> Starting prepare()...
patching file CMakeLists.txt
patching file launcher/Application.cpp
Hunk #1 succeeded at 308 (offset 4 lines).
patching file libraries/javacheck/CMakeLists.txt
patching file libraries/launcher/CMakeLists.txt
patching file launcher/Application.cpp
Hunk #1 succeeded at 530 (offset 6 lines).
patching file launcher/Application.cpp
Reversed (or previously applied) patch detected!  Assume -R? [n]

Thank you for packaging such tough software!!

Just updated the PKGBUILD so that the .desktop file execs multimc with the -d flag like gregthwuen suggested (thanks again!) as well as adding the patches that LennyLennington mentioned from multimc-git. Since this package is using those patches now, I have to leave this disclaimer:

It is illegal to redistribute any binaries created with this script because of the upstream project's licensing.

With these patches though, this package should be at feature parity with upstream builds - which it should be anyways since it's an open source project, but upstream seems to be a bit adversarial towards other distributions and repositories packaging multimc, so it's been a bit hard getting this out. Also because of upstream's attitudes towards repackaging, issues should be directed here as they don't give support to third party binaries. Enjoy!

Fortunately, there’s an easy fix to the “data folder is not writable” issue: MultiMC still supports the -d/--dir flag for setting the application root. Just use this to specify your multimc directory, i.e., $HOME/.local/share/multimc. You can even set the option in the desktop entry, though that doesn’t support the $HOME environoment variable I think. (edit: You can just specify a path relative to your home I realized.)

For everyone saying that it doesn't work due to the data folder not being writable: https://github.com/MultiMC/Launcher/issues/2004 The main developer doesn't want anybody using their distro's native package management utilities, and has essentially added code to break the application when not run through a loose binary in a writable folder. They're not going to change this behavior, although multimc-bin currently works.

Error:

" The launcher couldn't create a log file - the data folder is not writable.

Make sure you have write permissions to the data folder.

The Launcher cannot continue until you fix this problem."

Solution for me was to switch to the multimc-bin package. Data was still present after switch.

Error:

" The launcher couldn't create a log file - the data folder is not writable.

Make sure you have write permissions to the data folder.

The Launcher cannot continue until you fix this problem.

"

sup wit dat? how fix? what do? much sad.

Check https://aur.archlinux.org/packages/multimc-git for patches to fix the issues that it currently has, and for part of the PKGBUILD to restore the old icons and make Microsoft accounts work by getting the client ID of the official build.

iio7, cmake is already in the makedepends; so if you're building the package with makepkg or a half decent aur helper, it should already be installed while building. make and gcc should also already be installed if you have have installed base-devel like the wiki tells you to if you ever use the AUR. So these errors shouldn't be showing up unless you have no clue what you're doing - and if they are, please read more about the AUR, the ABS, and makepkg.

In case anyone is getting errors about CMake being unable to find a build program, like this:

CMake Error: CMake was unable to find a build program corresponding to "Unix Makefiles".  CMAKE_MAKE_PROGRAM is not set.  You probably need to select a different build tool.
CMake Error: CMAKE_C_COMPILER not set, after EnableLanguage
CMake Error: CMAKE_CXX_COMPILER not set, after EnableLanguage
-- Configuring incomplete, errors occurred!

You need to install the following:

# pacman -S make cmake gcc

made a new release to fix the build issues mentioned by segaja and because minecraft 1.18 requires java 17+

The feature of logging into Microsoft accounts introduced in MultiMC 0.16.3 will not be avalible for this package. So while the current build is 0.16.3, it does not have that feature due to MSA IDs being required at compile time and upstream cannot redistribute it. You can read more about it in this github issue I made.

(Thanks sleuth56 for bringing this to my attention. I forgot to post the issue here when I first made it)

edit: If you want to be able to use Microsoft accounts easily, you can use multimc-bin or multimc-native - the latter for if you don't like extracting .deb packages. Otherwise, you could in theory make a patch to modify the file notsecrets/Secrets.cpp with your own MSA ID, and then edit the PKGBUILD to apply it during prepare().

setting archlinux-java to java-17-openjdk helps and lets the build work.

kb1000, I do think Firefox is hostile. It's not the right way to handle FOSS projects.

For not having a updater, that is the idea of a package manager. A package should not have it's own updater or script. Updates should be handled by the packager, not by the user (if you disagree then don't use pacman). The multimc-bin package does not uninstall the software when package is removed (still in .local). I don't mind if there are secrets you cannot distribute. That is fairly reasonable, but PKGBUILDs themselves specifically do not contain source code to avoid trademark and copyright issues so forbidding packaging scripts makes no sense. If you don't like them, feel free to not support them. Just let people know why in a obvious place instead of closing every ticket within hours with barely any explanation.

I just used jdk-openjdk for the builddep and it worked fine. The modern-java.patch only sets cmake to target java 8 - which should allow this package to build with any version of java >= 8 - so i'm not sure why java 17 builds it and not java 11, but I'll do some testing on it and release an update if I can figure it out. For now though, using jdk-openjdk should work fine for the java-enviorment.

I have cmake installed and archlinux-java is set to java-11-openjdk.

Which version of java is needed to build multimc?

According to this issue, it says to use sudo archlinux-java set JAVA, where JAVA is your respective java version, ie. sudo archlinux-java set java-11-openjdk

When trying to install 0.6.13-2:

==> Starting build()...
-- The C compiler identification is GNU 11.1.0
-- The CXX compiler identification is GNU 11.1.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Git commit: e2355eb276bf355ca4acf526a0f3cc390aa88f8b
-- Git refspec: refs/heads/makepkg
-- Compiling for linux system with share/multimc and MULTIMC_LINUX_DATADIR
Using override nbt++ name: MultiMC_nbt++
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY - Success
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY - Success
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR - Success
/home/segaja/.cache/paru/clone/multimc5/src/multimc5/cmake
CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to
  the full path to the compiler, or to the compiler name if it is in the
  PATH.


-- Configuring incomplete, errors occurred!
See also "/home/segaja/.cache/paru/clone/multimc5/src/multimc5/build/CMakeFiles/CMakeOutput.log".
==> ERROR: A failure occurred in build().
    Aborting...
error: failed to build 'multimc5-0.6.13-2':
error: packages failed to build: multimc5-0.6.13-2

Surely you'd call the Firefox upstream hostile as well then. It just happens to be the case that peterix doesn't have the time to validate the quality of third-party packages (besides that the patches and missing API keys of this package do make the quality inacceptable, sharing MultiMC's keys is not possible because of liability and other clauses in the MSA agreements), and that MultiMC has to verify that you have a valid license for the game. Also, I really don't see how multimc-native is better than multimc-bin. After all, it doesn't have a working updater and doesn't allow switching update channels.

I have created a multimc-native package to not use the non-native update script. It uses same folder as multimc5 and is official binary.

This is the most hostile upstream I've seen since Oracle and OpenOffice.

I edited the pkgbuild with the right checksum but then another error comes up:

==> Überprüfe source Dateien mit sha512sums...
    multimc5-0.6.13.tar.gz ... Erfolg
    multimc-0.6.1.tar.gz ... Erfolg
    multimc-3.tar.gz ... Erfolg
    quazip-fix-build-with-qt-511.patch ... Erfolg
    modern-java.patch ... Erfolg
==> Entferne existierendes $srcdir/ Verzeichnis...
==> Entpacke Quellen...
  -> Entpacke multimc5-0.6.13.tar.gz mit bsdtar
  -> Entpacke multimc-0.6.1.tar.gz mit bsdtar
  -> Entpacke multimc-3.tar.gz mit bsdtar
==> Beginne prepare()...
/home/pietro/.cache/yay/multimc5/PKGBUILD: Zeile 38: cd: 
/home/pietro/.cache/yay/multimc5/src/MultiMC5-0.6.13: Datei oder Verzeichnis nicht gefunden
==> FEHLER: Ein Fehler geschah in prepare().
    Breche ab...
error making: multimc5

sha512 failing for multimc5-*.tar.gz , please update thanks.

Once debranding is fully finished and pushed to a release, this will no longer build as MultiMC, it will instead build as DevLauncher. Package name should probably change to reflect this. see: https://github.com/MultiMC/Launcher/blob/develop/notsecrets/README.md

They added a file to put the secret in, along with a description and link to register for one, if you'd like to register and put one in (or read a config file/env for users to provide their own)

https://github.com/MultiMC/MultiMC5/blob/develop/notsecrets/Secrets.cpp

Welp someone will have to adopt this, My sister and I switched from this to the multimc-bin well before this went down because of a issue we had with the bin version but that got fixed some time ago.

There is no infringement here - This is not distributing MultiMC, this is distributing a build script for users on their machines. The developer is blowing a bunch of hot air. Since users find value with this package, I'm denying the deletion request.

Those that care about this package, consider piping up to the developer and letting them know that they are mistaken. Also maybe kindly remind them that they're being a jerk. :)

Interesting that NixOS is able to distribute a similar build script

Whats hostile about it exactly? it's it really annoying to maintain or something?

Eh, screw it, I filed a deletion request. I have no interest in using or maintaing software with such a hostile upstream.

@jotoho: as far as I can tell, "MultiMC" is not a registered trademark anywhere, so there's nothing to infringe upon.

why use modern-java.patch, breaking compat with java versions <8 which are useful for alphas and betas, over setting JAVA_HOME via env variable to /usr/lib/jvm/java-8-openjdk(e.g. JAVA_HOME=/usr/lib/jvm/java-8-openjdk cmake ...) and adding java-8-openjdk as a makedepends?

I ran into this issue as a NixOS developer. It's important to note that MultiMC only requires a client ID, and not a client secret. The documentation explicitly states that the client ID can be public, and the Terms of Service explicitly states that it's confidential. I opened an issue about this, but haven't received a response: https://github.com/MicrosoftDocs/azure-docs/issues/80905

The author is now threatening that any "unofficial" builds are "a trademark infringement" and "illegal". https://github.com/MultiMC/MultiMC5/issues/4087#issuecomment-925005124

Thank you @jotoho for your explanation. Of course, as @xiretza pointed out, it would be nice to be able to compile the package with ms login support, but I see why this isn't viable. Nor I can think of ways to use a unique secret and compiling, if not by providing at least part of a binary. I guess we'll add it to the list of things we blame Microsoft for.

A -bin package can never supercede a package that builds from source just by virtue of existing. Especially for anyone that doesn't need Microsoft integration, building from source is obviously the better choice than using some binary off the internet that presumably even contains binary blobs to obfuscate secrets.

There's a patch that adds the MSA client ID: https://github.com/AOSC-Dev/aosc-os-abbs/pull/3445

Any chance to make the Microsoft login work in this package? It only works in the prebuilt one (multimc-bin) atm.

See related GitHub issue.

@gameslayer: You probably have jre-openjdk, jre8-openjdk and jdk8-openjdk installed but not jdk-openjdk. It builds fine with jdk-openjdk 16.0.2.u7-1

had to set my java version to java-8-openjdk to make it build successfully otherwise I ran into this issue

Preparing...
Cloning multimc5 build files...
Checking multimc5 dependencies...
Checking brave-bin dependencies...
Synchronizing package databases...
Warning: manjaro-hello: local (0.6.7-2) is newer than extra (0.6.6-9)
Resolving dependencies...
Checking inter-conflicts...
Building multimc5...
==> Making package: multimc5 0.6.13-1 (Fri 10 Sep 2021 14:58:31)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Found multimc5-0.6.13.tar.gz
  -> Found multimc-0.6.1.tar.gz
  -> Found multimc-3.tar.gz
  -> Found quazip-fix-build-with-qt-511.patch
  -> Found modern-java.patch
==> Validating source files with sha512sums...
    multimc5-0.6.13.tar.gz ... Passed
    multimc-0.6.1.tar.gz ... Passed
    multimc-3.tar.gz ... Passed
    quazip-fix-build-with-qt-511.patch ... Passed
    modern-java.patch ... Passed
==> Removing existing $srcdir/ directory...
==> Extracting sources...
  -> Extracting multimc5-0.6.13.tar.gz with bsdtar
  -> Extracting multimc-0.6.1.tar.gz with bsdtar
  -> Extracting multimc-3.tar.gz with bsdtar
==> Starting prepare()...
patching file libraries/javacheck/CMakeLists.txt
patching file libraries/launcher/CMakeLists.txt
patching file CMakeLists.txt
==> Removing existing $pkgdir/ directory...
==> Starting build()...
-- The C compiler identification is GNU 11.1.0
-- The CXX compiler identification is GNU 11.1.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Git commit: e0ec5ef3944095cc331082cc52c0726c5e7191fc
-- Git refspec: refs/heads/master
-- Compiling for linux system with share/multimc and MULTIMC_LINUX_DATADIR
Using override nbt++ name: MultiMC_nbt++
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY - Success
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY - Success
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR - Success
/var/tmp/pamac-build-corey/multimc5/src/MultiMC5-0.6.13/cmake
CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to
  the full path to the compiler, or to the compiler name if it is in the
  PATH.


-- Configuring incomplete, errors occurred!
See also "/var/tmp/pamac-build-corey/multimc5/src/MultiMC5-0.6.13/build/CMakeFiles/CMakeOutput.log".
==> ERROR: A failure occurred in build().
    Aborting...

@ISSOtm: thanks, I never come across such problems because I always build in clean chroots. I ended up using symlinks instead, you might have to remove the leftover directories once (or just use makepkg -C).

The package does not support building on top of an already-built repo; this patch fixes it.

diff --git a/PKGBUILD b/PKGBUILD
index f8f31a0..2f8c981 100755
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -45,8 +45,8 @@ prepare() {
   patch -p1 < "${srcdir}/add-limits-header.patch"
   patch -p1 < "${srcdir}/konamicode-fix-type-mismatch.patch"

-  rmdir "libraries/libnbtplusplus"
-  rmdir "libraries/quazip"
+  rm -rf "libraries/libnbtplusplus"
+  rm -rf "libraries/quazip"
   cp --recursive "${srcdir}/libnbtplusplus-${__pkgver_libnbtplusplus}/" \
     "libraries/libnbtplusplus"
   cp --recursive "${srcdir}/quazip-${__pkgver_quazip}/" \

just in case. fails to build with java-15-openjdk. archlinux-java set java-11-openjdk and back to 15 after install

@katt: many thanks for the patch and sorry for the wait.

Needs this patch applied now that GCC 11 is in core: https://github.com/MultiMC/MultiMC5/pull/3747

This gets it to 80%, then it needs this applied as well: https://github.com/MultiMC/MultiMC5/commit/3a1abb555b844c79a80615727b11019403767ce7.patch

diff --git a/PKGBUILD b/PKGBUILD
index b12c385..fef52c2 100755
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -5,7 +5,7 @@

 pkgname=multimc5
 pkgver=0.6.12
-pkgrel=1
+pkgrel=2
 __pkgver_libnbtplusplus=multimc-0.6.1
 __pkgver_quazip=multimc-3
 pkgdesc="Minecraft launcher with ability to manage multiple instances."
@@ -26,16 +26,22 @@ source=("$pkgname-$pkgver.tar.gz::https://github.com/MultiMC/MultiMC5/archive/${
         "https://github.com/MultiMC/quazip/archive/${__pkgver_quazip}.tar.gz"
         "quazip-fix-build-with-qt-511.patch"
         "modern-java.patch"
+        "multimc5-add-limits-header.patch::https://github.com/MultiMC/MultiMC5/commit/c2fd714f8d77db3223c5cb1a2d7f902d3e3e73fd.patch"
+        "multimc5-konamicode.patch::https://github.com/MultiMC/MultiMC5/commit/3a1abb555b844c79a80615727b11019403767ce7.patch"
 )
 sha512sums=('334b0996b7b49a0f4a46a3e5b08d50d5f7aa398b8d5871772c751c1f98438d9584ffcdaf27ad6d2e84b2776415bd38910dffb8e746c4b343916c6849df2146da'
             '81a1640a069d88df7ba0abf72089aecbe1e9d791c88acaaa7e70c8f0bcd0512cf8698178342657e363524ce8488dd072368a0aa8cc091a24912d6f8b6b0f4f2d'
             '2e9074203c67bc7ad98621c551047e5367f06e54cacfecc755a5bf2c9f99266eab42ad972f86ae28ed7e1507f6d27d8d2680a87ce9fd5b1e93a18bcb627ec3f0'
             'ca7a350bdeecf65dbca7de8d6912c935c6ba603edcddcd4ffe71d8997e50e4046335dde6d1d7c629d35025073d18be4d112a960d43a8801de979687bc26e46d4'
-            'ede87db7e1ce67184409123720b48f65d35884bac5dfff34cc099c5d7520fd9173be00997df82ad29da4640928134a7199ffdd1cbaa83a72101dbed72736508b')
+            'ede87db7e1ce67184409123720b48f65d35884bac5dfff34cc099c5d7520fd9173be00997df82ad29da4640928134a7199ffdd1cbaa83a72101dbed72736508b'
+            'e58cc72194b69999f96f08585a41470c0d1c074bc03992a9b41ac3d1bf3f676e526c2b42191e4d0c511281467445b13c1da82a26aeed271e05eb79efad30459e'
+            '44354df2fafcc9c08901a94d3d3eb8fe0a818d493b438799bdb75339c3fd9db91ccf9be2073271c1381496789819d74daf9353b28ab7e6a381db0d4f5ee41939')
 prepare() {
   cd "${srcdir}/MultiMC5-${pkgver}"

   patch -p1 < "${srcdir}/modern-java.patch"
+  patch -p1 < "${srcdir}/multimc5-add-limits-header.patch"
+  patch -p1 < "${srcdir}/multimc5-konamicode.patch"

   rmdir "libraries/libnbtplusplus"
   rmdir "libraries/quazip"

Tested and works

@drive_desaster: gcc is in base-devel, so it doesn't belong in makedepends. Either way, since gcc-git provides gcc, this wouldn't change anything.

I had problems compiling with gcc-git, so maybe make gcc in the current version a make dependency. It could also be that I made mistakes while installing gcc-git, I'll update this, as soon, as I reinstalled gcc-git and retried.

Since version 0.6.12 MultiMC allows using system's openal and glfw libraries. See the release notes at https://multimc.org/posts/0-6-12-update.html. Can you maybe add these as optional dependencies?

I fixed the issue by installing jdk-openjdk, previously I only had java 8 installed because that was what I was running minecraft with.

yeah looks like java-15-openjdk was needed to build because that fixed it

@xiretza I have the same issue as @gameslayer. I'm using java-8-openjdk/jre in archlinux-java, which is also the one I've been using until now.

I'm building via yay

EDIT: using archlinux-java to set the version to java-15-openjdk allowed me to build, I think there's a change in the build requirements...

I have java-11-openjdk set as default

Seems to be a issue with the quazip patch file

@xiretza I have the same issue as @gameslayer. I'm using java-8-openjdk/jre in archlinux-java, which is also the one I've been using until now.

I'm building via yay

@gameslayer: you probably selected an old java version with archlinux-java. Either change that or just build in a clean chroot to avoid any such issues with dirty host configurations.

Ran into a issue building the latest version

Preparing...
Cloning multimc5 build files...
Checking multimc5 dependencies...
Synchronizing package databases...
Resolving dependencies...
Checking inter-conflicts...
Resolving dependencies...
Checking inter-conflicts...
Download of manjaro-system (20210321-2) started
Download of manjaro-system (20210321-2) finished
Checking keyring...
Checking integrity...
Loading packages files...
Checking file conflicts...
Checking available disk space...
Upgrading manjaro-system (20210302-1 -> 20210321-2)...
==> Checking for 'os-prober' setup ...
    We will reenable 'os-prober' for you ...
Generating grub configuration file ...
Found theme: /usr/share/grub/themes/manjaro/theme.txt
Found linux image: /boot/vmlinuz-5.4-x86_64
Found initrd image: /boot/intel-ucode.img /boot/initramfs-5.4-x86_64.img
Found initrd fallback image: /boot/initramfs-5.4-x86_64-fallback.img
Warning: os-prober will be executed to detect other bootable partitions.
It's output will be used to detect bootable binaries on them and create new boot entries.
Found Windows Boot Manager on /dev/sda1@/efi/Microsoft/Boot/bootmgfw.efi
Adding boot menu entry for UEFI Firmware Settings ...
Found memtest86+ image: /boot/memtest86+/memtest.bin
/usr/bin/grub-probe: warning: unknown device type nvme0n1.
done
Resolving dependencies...
Checking inter-conflicts...
Download of java-environment-common (3-3) started
Download of java-environment-common (3-3) finished
Download of jre-openjdk (15.0.2.u7-1) started
Download of jre-openjdk-headless (15.0.2.u7-1) started
Download of jre-openjdk (15.0.2.u7-1) finished
Download of jdk-openjdk (15.0.2.u7-1) started
Download of jre-openjdk-headless (15.0.2.u7-1) finished
Download of jdk-openjdk (15.0.2.u7-1) finished
Checking keyring...
Checking integrity...
Loading packages files...
Checking file conflicts...
Checking available disk space...
Installing jre-openjdk-headless (15.0.2.u7-1)...
Default Java environment is already set to 'java-11-openjdk'
See 'archlinux-java help' to change it
Installing jre-openjdk (15.0.2.u7-1)...
Default Java environment is already set to 'java-11-openjdk'
See 'archlinux-java help' to change it
when you use a non-reparenting window manager,
set _JAVA_AWT_WM_NONREPARENTING=1 in /etc/profile.d/jre.sh
Installing java-environment-common (3-3)...
Installing jdk-openjdk (15.0.2.u7-1)...
Default Java environment is already set to 'java-11-openjdk'
See 'archlinux-java help' to change it
Running post-transaction hooks...
Arming ConditionNeedsUpdate...
Updating icon theme caches...
Updating the desktop file MIME type cache...

Building multimc5...
==> Making package: multimc5 0.6.12-1 (Mon 22 Mar 2021 13:29:31)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Downloading multimc5-0.6.12.tar.gz...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
100   124  100   124    0     0    351      0 --:--:-- --:--:-- --:--:--   351

  1 2333k    1 45089    0     0  59879      0  0:00:39 --:--:--  0:00:39 59879
100 2333k  100 2333k    0     0  2083k      0  0:00:01  0:00:01 --:--:-- 6256k
  -> Downloading multimc-0.6.1.tar.gz...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
100   137  100   137    0     0    353      0 --:--:-- --:--:-- --:--:--   354

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
100 47171    0 47171    0     0  60709      0 --:--:-- --:--:-- --:--:--  258k
  -> Downloading multimc-3.tar.gz...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
100   125  100   125    0     0    354      0 --:--:-- --:--:-- --:--:--   354

 87  101k   87 90964    0     0   114k      0 --:--:-- --:--:-- --:--:--  114k
100  101k  100  101k    0     0   127k      0 --:--:-- --:--:-- --:--:--  721k
  -> Found quazip-fix-build-with-qt-511.patch
  -> Found modern-java.patch
==> Validating source files with sha512sums...
    multimc5-0.6.12.tar.gz ... Passed
    multimc-0.6.1.tar.gz ... Passed
    multimc-3.tar.gz ... Passed
    quazip-fix-build-with-qt-511.patch ... Passed
    modern-java.patch ... Passed
==> Removing existing $srcdir/ directory...
==> Extracting sources...
  -> Extracting multimc5-0.6.12.tar.gz with bsdtar
  -> Extracting multimc-0.6.1.tar.gz with bsdtar
  -> Extracting multimc-3.tar.gz with bsdtar
==> Starting prepare()...
patching file libraries/javacheck/CMakeLists.txt
patching file libraries/launcher/CMakeLists.txt
patching file CMakeLists.txt
==> Starting build()...
-- The C compiler identification is GNU 10.2.0
-- The CXX compiler identification is GNU 10.2.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Git commit: 938beab8a5e07af9377820eecc8faec56515f525
-- Git refspec: refs/heads/master
-- Compiling for linux system with share/multimc and MULTIMC_LINUX_DATADIR
Using override nbt++ name: MultiMC_nbt++
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_VISIBILITY - Success
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY
-- Performing Test COMPILER_HAS_HIDDEN_INLINE_VISIBILITY - Success
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR
-- Performing Test COMPILER_HAS_DEPRECATED_ATTR - Success
/var/tmp/pamac-build-corey/multimc5/src/MultiMC5-0.6.12/cmake
CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to
  the full path to the compiler, or to the compiler name if it is in the
  PATH.


-- Configuring incomplete, errors occurred!
See also "/var/tmp/pamac-build-corey/multimc5/src/MultiMC5-0.6.12/build/CMakeFiles/CMakeOutput.log".
==> ERROR: A failure occurred in build().
    Aborting...

xorg-xrandr should be marked as an optional dependency really. The package is only required for running Minecraft versions <=1.12.2

@ImagineBeingALEX: can't reproduce, use makechrootpkg before reporting a bug that may very well be your (unsupported) AUR helper's fault.

When ever I try to install Multi Mc through paru I get this error error: failed to run: pacman --sync -- extra/xcb-util-keysyms extra/xcb-util-renderutil extra/xcb-util-wm extra/xcb-util-image extra/tslib extra/double-conversion extra/md4c extra/qt5-base extra/qt5-x11extras extra/qt5-svg extra/xorg-xrandr: No such file or directory (os error 2)

@dotto: patch is in base-devel, which is assumed to be installed.

The package requires patch to build. it would be nice if you add it to the dependency list.

May I suggest to use the flag "-j" in order to build on multiple CPU core ? @mcmacker4 explain it well in the following comment: https://aur.archlinux.org/packages/multimc5/?O=10&PP=10#comment-727637 Thank you :)

May I suggest to make xorg-xrandr an optional dependency for those who play Minecraft <= 1.12.2?

i caint say for sure but this was a fresh install less then a week old and i def ran that several times (maybe even as tried as a fix for this) so i know it is up to date

@Smaug382: when did you last -Syu?

i didnt see this problem in any of the comments but im having an error with the cmake mirror list all mirror url's return error 404 (it then fails install bc if missing dependency)

$ sudo archlinux-java fix
[sudo] password for user: 
$

$ archlinux-java status  
Available Java environments:
  java-11-openjdk
  java-8-openjdk/jre (default)

$ yay -Syu aur/multimc5
[...]
CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.
[...]

(corrected with sudo archlinux-java set java-11-openjdk)

@spk44: oh, that could be it. That's some unfortunate placement and wording, but I guess I'll have to live with it.

I believe the reason you are getting "out of date" flags is because the launcher itself displays the last blog post for 0.6.7, and includes in the title "update required!"

I don't know where people keep getting this from, but please stop flagging the package out of date with links to old release announcements from a year ago. 0.6.11, from 2020-03-29, is the latest release as of right now.

About the Cmake-error:

If someone is confused about jdkX (whereas X = java environment you have), you should check with "sudo archlinux-java status" what kind of java environment you have. Mine was defaulted to java-8-openjdk/jre which unfortunately didn't work, so i installed only the jdk8-openjdk-package with "sudo pacman -S jdk8-openjdk" and the install worked for me after that. Im guessing the -jre is same problem for all java environments and this fix should work in the other environments too, but im not sure.

I did not remove the "jre"-jdk8 package: the package overrided it automatically.

Why can't you set the JAVA_HOME env variable in the build function, like is done in the multimc-git package (https://aur.archlinux.org/packages/multimc-git/). It seems to work fine like this, and could solve a lot of frustration.

@xiretza thanks, I set java to openjdk 8 and everything works nice now

@Wennadocta: first off, it builds fine in a clean chroot, which is the only setup that can be supported reasonably - AUR helpers in particular are completely out of scope, I suggest you at least use makepkg before reporting any package bugs in the future. Also, did you read the pinned comment? I think it might solve your problem.

Last version doesn't compile

/home/ayanami/.cache/yay/multimc5/src/MultiMC5-0.6.11/cmake
CMake Error at libraries/launcher/CMakeLists.txt:2 (project):
  No CMAKE_Java_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to
  the full path to the compiler, or to the compiler name if it is in the
  PATH.


-- Configuring incomplete, errors occurred!
See also "/home/ayanami/.cache/yay/multimc5/src/MultiMC5-0.6.11/build/CMakeFiles/CMakeOutput.log".
==> ERROR: A failure occurred in build().
    Aborting...
Error making: multimc5

@xirertza Yeah, that did the trick. Turned out I was using jdk 13, switched to version 8 and it worked like a charm.

@sybsfksks: "doesn't work" is about the least helpful error report possible. I just created a 1.10 instance, then installed forge 12.18.0.2000 from the "Version" menu of the instance, and it launches just fine using Java 8.

Edit: just saw you added a stack trace - the full log would probably show that the java version you're using is too new.

On the latest version, any version below 1.12.2 with forge doesn't work for some reason. Here's the full error: https://pastebin.com/nKCM7tjJ

@mcmacker4: please use MAKEFLAGS= in makepkg.conf for this, that way it applies to all packages you build and package maintainers don't have to implement it separately for every package.

Since the whole program is built locally and there is quite a lot of files to build, may i suggest adding the jobs flag to the make command in the build() function of PKGBUILD like so:

make -j $(nproc)

This will allow make to build files in parallel using as many threads as cores the computer has (as reported by the nproc command). By adding this flag (on my 4 core machine) the build time was reduced from 8m36s down to 2m52s.

@gameslayer: sounds like an AUR helper problem. Please use makechrootpkg when reporting bugs.

Cloning multimc5 build files... Checking multimc5 dependencies...

than just stops, my system is up to date..

Can't reproduce any build failures on either my live system or in a clean chroot (which is all that matters anyway). Also, don't use the out-of-date system for broken packages, post a detailed comment including a reproducer, logs and maybe even a workaround instead.

Build assumes that the Makefile generator is used by cmake. But from cmake 3.15 on the CMAKE_GENERATOR environment variable changes the default generator.

This can be circumvented by setting the Makefile generator explicitly or initiating the build using cmake so

cmake --build .
cmake --build . --target test
cmake -E env DESTDIR="${pkgdir}" cmake --build . --target install

Does not build https://haste.rys.pw/raw/mufelaquwa

sudo archlinux-java status -> Available Java environments: java-10-openjdk; java-8-openjdk/jre (default)

AUR/multimc-git does, it recognized I'm missing jdk8-openjdk and pulled it, and then this package built fine too.

I'm getting an error when building the NewLaunch.jar file:

warning: [options] bootstrap class path not set in conjunction with -source 6
error: Source option 6 is no longer supported. Use 7 or later.

Edit: Turns out I had multiple versions of Java installed and the build process used the wrong compiler. I simply removed the newer Java versions and it built OK.

just a friendly reminder even though this installs packages of jdk < 12 it will still attempt to use the default one which may be set to 12 like in my case.

You need to use sudo archlinux-java status and sudo archlinux-java set java-11-openjdk to figure out what you have and set something lower than 12 as default before attempting to install.

Just wanted to let you all know that 0.6.6-1 fixes the issue with quazip. The package compiled just fine for me.

0.6.5-2 fails to compile, git version did.

Having same/similar problem as @mjevans.

Edit: Strangely enough, though @mjevans's comment reports that they are unable to update multimc-git, I tried installing that and it compiled just fine.

I'm unable to update multimc-git (and unable to install multimc5) with multiple errors related to a deprecated function being treated as: error: ‘static QString QuaGzipFile::trUtf8(const char, const char, int)’ is deprecated [-Werror=deprecated-declarations]

The patch rasphino commented on 2019-05-23 11:47 looks like it might resolve this issue.

[ 16%] Building CXX object libraries/pack200/CMakeFiles/MultiMC_unpack200.dir/src/unpack.cpp.o ~/.cache/yay/multimc5/src/MultiMC5-0.6.5/libraries/quazip/quazip/quagzipfile.cpp: In member function ‘bool QuaGzipFilePrivate::open(FileId, QIODevice::OpenMode, QString&)’: ~/.cache/yay/multimc5/src/MultiMC5-0.6.5/libraries/quazip/quazip/quagzipfile.cpp:60:30: error: ‘static QString QuaGzipFile::trUtf8(const char, const char, int)’ is deprecated [-Werror=deprecated-declarations]
60 | error = QuaGzipFile::trUtf8("QIODevice::Append is not " | ^~~~~~ In file included from /usr/include/qt/QtCore/qobject.h:46, from /usr/include/qt/QtCore/qiodevice.h:45, from /usr/include/qt/QtCore/qfiledevice.h:43, from /usr/include/qt/QtCore/qfile.h:44, from /usr/include/qt/QtCore/QFile:1, from ~/.cache/yay/multimc5/src/MultiMC5-0.6.5/libraries/quazip/quazip/quagzipfile.cpp:25:

@RealGecko as @knowt mentions below: Removing java12 packages - jdk-openjdk, jre-openjdk jre-openjdk-headless and then, while installing choosing jdk8-openjdk did a trick for me

:: There are 5 providers available for java-environment:
:: Repository extra
    1) jdk-openjdk 2) jdk10-openjdk 3) jdk11-openjdk 4) jdk7-openjdk 5) jdk8-openjdk 

Enter a number (default=1): 5

https://paste.nomsy.net/ojikevolub Can someone tell me how to fix this, both packages do this

I get following error when try to compile 0.6.5-2:

/tmp/packerbuild-1000/multimc5/multimc5/src/MultiMC5-0.6.5/cmake

CMake Error at libraries/launcher/CMakeLists.txt:2 (project):

No CMAKE_Java_COMPILER could be found.

@jambon thanks, the patch is now applied by the package, and launching works again. Sorry for the late reply.

@xiretza This patch should fix the issue https://github.com/MultiMC/MultiMC5/commit/70ed30f9e6b47fbfb3151078ca61a2bb419feccb.patch

I failed to build multimc5, because it use -Werror in its cmakefile, while using tons of deprecated functions. Thus, I make a patch, which adds -Wno-error=deprecated-declarations to make multimc5 compile.

https://paste.ubuntu.com/p/T5R4FcjqZP/

If you encounter the same problem, you can give this patch a try.

The segfault when launching an instance is a known issue and will be fixed in the next MultiMC release. If you don't want to wait for that, either use multimc-git or post a patch here, I'll include it in the package.

Crashes and does the Segmentation fault (core dumped) error when I launch any version of Minecraft

work around for me;

removed all java via Add/Remove software under AUR and Repo. Download openjdk-8

https://www.archlinux.org/packages/extra/x86_64/jdk8-openjdk/download/

and install it. retry MultiMC from AUR and it should compile and install.

this is for Manjaro, but should work either way.

@radiomike: Thanks for the heads up! Literally spent half an hour looking at the log file trying to find what happened.

I think the java dependency should be java-runtime=8 instead of just java-runtime. Also java-environment=8 instead of just java-environment

Currently broken with latest qt5 and icu updates. Fixed upstream already.

@ShayBox: I found at least a temporary solution as I was having the same problem. I used archlinux-java to change my Java environment back to a JDK. In my case I was using java-8-jre/jre instead of java-8-openjdk or java-8-jdk.

Here's what I did since I had jdk8 (http://aur.archlinux.org/packages/jdk8) installed. jdk8-openjdk would also probably work:

$ archlinux-java status
Available Java environments:
  java-8-jdk
  java-8-jre/jre (default)
  java-8-openjdk
$ sudo archlinux-java set java-8-jdk

This solved the problem for me.

@ShayBox: it builds in a clean chroot, and I honestly can't be bothered to try adding special support for multiple java versions to the PKGBUILD, sorry. Feel free to post a good, working patch though.

Why is the java-environment package necessary? The runtime alone seems to work well enough.

@CodingKoopa, you need Java 8. Vanilla Minecraft can run with an older version, but modded specifically demands 8.

ninja edited for formatting

I'm getting this when building this package: "The dependency jdk-openjdk demands version <10, but the most recent version is 11.0.2.u7-1.". I have both the "jdk-openjdk" and "jdk10-openjdk" packages installed, with the Java 11 runtime set as the default with the "archlinux-java" script.

@xiretza well, you were right... I forgot the backslash and now that I have placed it correctly, I realize I have not "fixed" the issue and I am back to square one.

@Monkeyfume: this is why I asked for a diff, you probably messed up the cmake command while "fixing" your broken java path. The last argument to cmake has to be .., you probably forgot a backslash to escape the newline at the end of your variable definition.

@xiretza here is my output before changing the PKGBUILD: https://pastebin.com/6qss3UX3

@ender4's suggestion from 2017-10-30 was "I was able to compile it with openjdk-9 installed by changing the cmake line of the install script to include "-DJAVA_HOME=/usr/lib/jvm/java-8-openjdk" ". I followed this and the No CMAKE_Java_COMPILER could be found. ... error went away. However now I am left with the second error, CMake Error: The source directory "/tmp/yaourt-tmp-monkeyfume/aur-multimc5/src/MultiMC5-0.6.2/build" does not appear to contain CMakeLists.txt.

@Monkeyfume: "I changed the PKGBUILD and now it won't build anymore" - sorry, can't really help you with that, at least not without a diff.

I was able to solve my previous issue from a few comments down, and compiled using ender4's suggestion of editing the PKGBUILD.

However, I now receive the error: CMake Error: The source directory "/tmp/yaourt-tmp-monkeyfume/aur-multimc5/src/MultiMC5-0.6.2/build" does not appear to contain CMakeLists.txt.

I have found the reason! I'm using proxychains to bypass some network issues in China, but I wasn't excluding localhost in the configuration. So the connection to localhost in the test program times out. Problem resolved when I set up proxychains properly.

@Towdium: can't reproduce, neither in clean chroot nor on an actual fresh install.

It seems I'm still getting problem with quazip. On a fresh installed system, I get test quazip_testsuite failed: https://pastebin.com/kP3Nyuud

@Monkeyfume: looks good, not sure what could be wrong. Try archlinux-java fix, if that doesn't work, you'll have to go fix your $PATH manually.

@xiretza: base-devel is installed. Output of $ archlinux-java status is:

[monkeyfume@computer ~]$ archlinux-java status Available Java environments: java-10-openjdk java-7-openjdk java-8-openjdk/jre (default)

@Monkeyfume: did you follow the tips that were posted back then? Namely:

• do you have base-devel installed?

• what's the output of archlinux-java status?

Installation failed. I'm getting the same issue @Aurelio experienced two years ago--

"No CMAKE_Java_COMPILER could be found.

Tell CMake where to find the compiler by setting either the environment variable "JAVA_COMPILER" or the CMake cache entry CMAKE_Java_COMPILER to the full path to the compiler, or to the compiler name if it is in the PATH."

Tested with pacaur and yaourt. Not sure how to solve this issue. Any help?

I adopted this package and manually applied the Qt5.11 patches from the github issues (MultiMC5#2304, quazip#1). Let me know if there's any other problem

Can't compile with Qt 5.11 due to the removal of the deprecated qt5_use_modules. These patches work on the current version, and are based on the fixes in https://github.com/MultiMC/MultiMC5/pull/2304:

• PKGBUILD: https://wank.party/uAxE.pkgbuild
• multimc_use_modules.patch: https://wank.party/U5LR.patch
• quazip_use_modules.patch: https://wank.party/rVyh.patch

I had hoped that I fixed it with:

makedepends=('java-environment>=8' 'java-environment<10')

Doesn't look like it though. It looks like it just makes certain that you have it installed. I might try erroring out if java >= 10 before build sometime and print an error message on the next update.

At first I have exactly the same problem like @SuperGeniusZeb.

My Enviroment:

archlinux-java status

Available Java environments:

java-10-jdk

java-10-openjdk (default)

java-8-openjdk

But it worked with the tip from @ender4 (change the cmake line).

This package does not install properly when OpenJDK 10 is being used. (It worked fine with OpenJDK 9, if I remember correctly.)

Here is the log for when I tried to install the package: https://pastebin.com/3uPkUFdC

@Derriick: Since Minecraft is not working on your setup without xorg-xrandr, and Minecraft is what MultiMC is for, I'm moving xorg-xrandr to the depends array.

Thanks for the pastebin. New version should be up now.

EDIT: Minecraft is core functionality for MultiMC -- if it doesn't run, then MultiMC is not running as intended.

@vorpalblade: Thanks for your answer. I just confirm that this problem occurs when I try to run vanilla minecraft (1.12.2 and 1.11.2, without and with optifine), maybe because of my setup (wayland + sway).

I post here the bug report in case you still want to see it: https://pastebin.com/maJ6jLmx

However, I understand xorg-xrandr can't be set as required, as multimc doesn't need it directly to run.

@Derriick: xorg-xrandr is optional -- I have run vanilla minecraft without it before.

That said, making xorg-xrandr mandatory isn't a bad idea, but I did specify xorg-xrandr as something that may be required for some modpacks. This fits with the wiki model for dependencies[1].

Even though it isn't required, I would still highly recommend xorg-xrandr. But xorg-xrandr isn't absolutely required for Minecraft to run, so I'll leave it as an optdepend.

I may at some point try to nail down what, exactly, requires xorg-xrandr in modpacks, but it isn't high on my priority list right now. TBH, I kind of suspect Minecraft Forge. Either that, or one of the various "make Minecraft faster" mods.

[1] https://wiki.archlinux.org/index.php/PKGBUILD#Dependencies

For Wayland users, xorg-xrandr (which is indicated as "optional") is necessary. I think it's automatically installed for all Xorg users.

Could you add it as a required dependency (like in minecraft package), to avoid future users any issues ?

@JuniorJPDJ Thanks -- I intended to increase pkgrel, not pkgver.

There is no tag for 0.6.2, only 0.6.1