Please, provide an instruction how to build on Linux.

[muhammadyusuf-kurbonov]

I'm a Linux user (Fedora 41). And of course love C&C, want to play Native on Linux. I have done vcpkg install, but I don't know what to do next? I'm just Java and JS developer, not familiar with C/C++ ecosystem (sorry for that 😃 )

[sebastianPsm]

Hi,

you'll find GitHub workflow files for ubuntu here. That's exactly what you requested (for ubuntu; Not for fedora).

See the ubuntu.yml

• l:18..22 (installing dependencies on the system)
• l:28 (installing CMake and Ninja on the system)
• l:32 (installing vcpkg)
• l:35 (start the CMake dance)

The CMake dance: Typical workflow is:

• go into the local repo
• create a build directory (mkdir build)
• change into the build dir (cd build)
• Start cmake to generate a Ninja project (cmake .. -G "Ninja"). Something is missing here reagrding the presets (you need to find that out yourself)
• Start the build (cmake --build . -j)

[muhammadyusuf-kurbonov]

Ough. Thank you

[karstenBriksoft]

On Debian 12 (x64, as ARM64 is not supported yet):

apt install -y nasm autoconf automake libtool pkg-config libltdl-dev libx11-dev 
      libxft-dev libxext-dev libwayland-dev libxkbcommon-dev libegl1-mesa-dev 
      libibus-1.0-dev ninja-build git curl cmake clang zip python3-jinja2

Install vcpkg via git clone https://github.com/microsoft/vcpkg.git and then cd vcpkg and ./bootstrap-vcpkg.sh.
You need to add VCPKG to the environment via export VCPKG_ROOT=$PWD and export PATH=$PATH:$VCPKG_ROOT.

Then you can run cmake using the --preset default option:

mkdir build
cd build
cmake .. --preset default
# this will also generate a default folder, where all the cmake files go into, thus the next build command requires the default folder as argument
cmake --build default

That will automatically run vcpkg, too.
It may also download a more up to date version of cmake.

This seemed to have done it. But while it produced a lot of .o files i failed to find any executable. What the build script ended with was:

Linking CXX static library GeneralsMD/Code/GameEngineDevice/libGameEngineDevice.a

[karstenBriksoft]

with these settings it was even possible to partially build on macOS... what i'm wondering though: at the beginning of the first cmake call, the output are a lot of using ... ON or OFF messages... how can these features be enabled?

[sebastianPsm]

-D<MY_OPTION_NAME>:BOOL=ON (or OFF)

[WolfMad]

hi
Using the default preset produced a lot of .o object files, but you do not get a usable binary file.

Using preset linux64-deploy "- Linux Release Config" or linux64-testing "- Linux Test Config" in the folder you will find the RTS binary.
CnC_Generals_Zero_Hour/build/linux64-deploy/GeneralsMD/Code
CnC_Generals_Zero_Hour/build/linux64-testing/GeneralsMD/Code

But something is missing I can post the infomation for how I build Command & Conquer: Generals – Zero Hour. if it will help

[muhammadyusuf-kurbonov]

Don't you know, what should I do with libmss32.so.1.0?

[xpocho]

Hi, I'm not sure what you mean by that question, but if the problem is that the game can't find it during startup, then I assume you need to add it to library load path same as for the libdxvk_d3d8.so eg. by setting env var:
"export LD_LIBRARY_PATH=/path/to/your/build/_deps/dxvk-src/lib:/path/to/your/build/_deps/miles-build:$LD_LIBRARY_PATH"
(obviously need to change "/path/to/your/build/" to the path to your linux64-deploy or linux64-testing folder whichever you use).
I did not have problem with libmss32.so. But I did run into problem with the missing libdxvk_d3d8.so before so I checked for other libs in the "_deps" folder and found the mss32, so I have started adding both load paths before starting the game.

[misterhackerman]

will this give a linux executable? and does it support multiplayer?

[muhammadyusuf-kurbonov]

I've made a devcontainer (Docker container) and build the executable successfully. Now, when I start the app - got an error:

Authorization required, but no authorization protocol specified

It is because Docker container doesn't have XServer. I have mounted my host system XWayland. But no success. How can I turn on verbose mode to troubleshoot?

[Fighter19]

I got the same message yesterday, while trying to connect to a Raspberry Pi, without a running XServer.

Make sure to be able to start up an application like "vkcube".
And make sure it uses your GPU and not llvmpipe.

Tunneling via SSH is not supported. To my knowledge Docker should work, but often running graphical applications requires sharing resources from /dev and is quite a bit of pain.

I'll see if more is available with the appropriate flags. But I assume that message just means that the Xauthority file for the current session couldn't be found.