DIRTSAND is a full featured MOULa-compatible server platform for POSIX-compliant operating systems, written in C++ and released under the AGPL version 3+. Currently, it has only been tested on Linux, but in theory it should work on other Unixes as well. There are, however, currently no plans for Windows development or support.
However, DIRTSAND may be set up on Windows (or other architectures) using Docker in order to be used as a local testing server. For instructions, go to the DOCKERSAND section directly.
1) Create a user and path for the DIRTSAND server to run from:
$ sudo useradd dirtsand -d /opt/dirtsand -m -p <password> -s /bin/bash
$ su dirtsand
2) Check out a copy of the source:
$ git clone https://github.com/H-uru/dirtsand.git
$ cd dirtsand
3) Compile it for your system:
$ mkdir build && cd build
$ cmake -DCMAKE_INSTALL_PREFIX=/opt/dirtsand ..
$ cmake --build . && sudo cmake --build . --target install
$ cd ..
If you run into any errors about finding libraries or headers, make sure
you have the development versions of all of the required libraries, and
that they are in your path. You can also use the cmake-gui
to help cmake
locate the missing paths and files.
You will need a working Postgres server which DIRTSAND can use to store its data. Although you don't need superuser access to the postgres server, you will need to have a database which you can add schemas, tables, etc into.
Setting up postgres functionality without a superuser account is currently beyond the scope of this document... If you need to do this, keep in mind that you may also have to edit the database initialization scripts to work against an existing user and/or database.
For the default installation, the provided scripts will create a dirtsand database and set its ownership to a 'dirtsand' database user, which can directly map the system dirtsand user created in step 1 of the "Building the code" instructions above. For better security, it is recommended to use a password (as shown in the steps below), which can be configured in the server settings as described in the "configure dirtsand" step.
1) Set up the postgres user:
$ sudo -u postgres psql -d template1
template1=# CREATE USER dirtsand WITH PASSWORD '<password>';
template1=# CREATE DATABASE dirtsand WITH TEMPLATE = template0 ENCODING = 'UTF8';
template1=# ALTER DATABASE dirtsand OWNER TO dirtsand;
template1=# \q
2) Install the UUID functionality:
This may be provided by your OS distribution. In Ubuntu, simply install
the postgresql-contrib
package to provide the necessary libraries and
installation scripts. If your distribution does not provide a contrib
or uuid-ossp bundle, you can get it and build it yourself from the
sources provided at: http://www.ossp.org/pkg/lib/uuid/
In versions of Postgres older than 9.x, you will need to use the import script to add the uuid extension to the dirtsand database:
$ sudo -u postgres psql -d dirtsand < /path/to/uuid-ossp.sql
# (Example for Ubuntu 10.10 with PostgreSQL 8.4)
$ sudo -u postgres psql -d dirtsand < /usr/share/postgresql/8.4/contrib/uuid-ossp.sql
In PostgreSQL 9.x and newer, simply use the "CREATE EXTENSION" syntax to add the uuid-ossp extension it to the Dirtsand database:
$ sudo -u postgres psql -d dirtsand
dirtsand=# CREATE EXTENSION "uuid-ossp";
dirtsand=# \q
3) Set up the dirtsand database:
$ psql -d dirtsand < db/dbinit.sql
$ psql -d dirtsand < db/functions.sql
If there were no errors, your database should be ready for DIRTSAND.
4) Configure dirtsand:
A sample dirtsand.ini has been provided in the root of the dirtsand sources. We can copy this to our install directory and then edit the fields we need. Specifically, you will need to adjust the server addresses and the RC4 keys. If you have dirtsand installed to somewhere other than /opt/dirtsand, you will also need to point the configuration to the right paths too.
$ sudo cp dirtsand.sample.ini /opt/dirtsand/dirtsand.ini
$ sudo chown dirtsand /opt/dirtsand/dirtsand.ini
$ <your-favorite-editor> dirtsand.ini
To generate the RC4 keys, you can simply run the command-line option
generate-keys
using dirtsand:
$ bin/dirtsand --generate-keys
You should now have a bunch of keys output on your terminal. The first
block (labeled Server keys) is the set you should paste into your
dirtsand.ini. Replace the dummy lines (with the '...' values) with the
output from the --generate-keys
command. The second set of keys can be
placed directly in the client's server.ini file (NOTE: This requires
either the H-uru fork of CWE or PlasmaClient -- the vanilla MOUL
client cannot load keys from a file, and you will have to enter the
keys as byte arrays directly into the executable.
5) Provide data for the client to use. This process can be performed mostly automatically using the UruManifest tool, which prepares a complete set of data files and generates the appropriate manifests. The instructions below describe how to set up the data files and manifests manually.
The most important data for the client are the auth server provided files -- specifically, the SDL and python.pak. External Plasma clients require these files in order to function, so you will need to provide them unless you are planning on making your server only work with Internal client builds.
The SDL and Python files can be found in the Scripts directory of the Plasma repository). The process of compiling the Python scripts and packing them into a python.pak is currently outside of the scope of this document.
To provide the files to the client, set up a directory for auth data (the default is /opt/dirtsand/authdata) and put the files in their respective subdirectories (SDL/*.sdl and Python/*.pak). Note that these files will need to be NTD encrypted with the key specified in your dirtsand.ini file (the default is the first 32 decimal digits of pi, without the decimal point). You can specify any 32 hex digit key, but make sure to update the files to match.
In order to let the client know what files are available, you will also need to provide the files Python_pak.list and SDL_sdl.list, which are simple manifest files of the format:
Path\filename.ext,size-in-bytes
So, for python.pak, you might have:
Python\python.pak,123456
Note the use of a backslash here, since this path is provided directly
to the client (which assumes a Windows path). Comments (starting with
the #
character) and whitespace will be ignored when parsing this file.
DIRTSAND also includes a full data server with its services. If you choose to use it, you will need to set up a data/ directory and the necessary manifests. At the very least, you will need the ThinExternal.mfs and External.mfs files, which describe the files necessary for the patcher and client (respectively) to run. You may also provide any number of age manifests, which will be requested by the client when it attempts to link to an age (e.g., Ahnonay.mfs).
The format of the manifest files is as follows:
remote_filename.ext,local_filename.ext.gz,decompressed_hash,compressed_hash,decompressed_size,compressed_size,flags
There is also a helper script provided in bin/dsData.sh which will gzip a file and generate a manifest line with the correct hashes and sizes for you automatically. Take care that the remote path expects to use a Windows path/filename, so it should use backslashes instead of forward ones, whereas the local filename should use Unix slashes.
Alternatively, you can specify a remote file server address in your dirtsand.ini, which will allow the files to be fetched from an external file server.
6) Run the server:
Assuming everything else went smoothly, you should now be able to start your server and connect to it! You'll have to create an account first, which can be done from the console:
$ bin/dirtsand dirtsand.ini
ds> addacct <username> <password>
If you can't connect for some reason, make sure you copied the keys and server addresses correctly into the client's server.ini, and check that your firewall is set to allow connections from port 14617.
If you want to leave the server running across different login sessions and you don't have an X or VNC server running, I recommend running dirtsand in a detachable GNU screen session.
Docker allows building and running isolated software programs (or "containers") in a platform-agnostic way. This allows users to bootstrap a new DIRTSAND testing server much more quickly and easily on Windows, Mac, or Linux.
1) Download and install Docker Desktop. If on Windows, you probably want to install WSL as directed by the installer. Your computer may need to restart several times during these installs.
2) Clone the dirtsand repo to a local directory.
3) From the local repo root directory, run ./dockersand build
.
dockersand.ps1
(for Windows) or the PROJECT_NAME
variable in dockersand.sh
(for *nix) to the desired base container name../dockersand
commands in Windows PowerShell may trigger
security exceptions. You can avoid these by using the Command Prompt or by
running Unblock-File dockersand.ps1
.4) Once the build is complete, run ./dockersand start
to start the new containers
for the database and the dirtsand server. Once this command has been run, you should
be able to see the containers in Docker Desktop. You can also view the docker
logs for the containers.
5) Retrieve the server.ini
file from the build/etc folder. You will need this file
for connecting to the server with a game client.
6) Use ./dockersand attach
to attach to the dirtsand server and run commands as
desired. For example, use addacct <username> <password>
to create the user
account you want to use to connect locally via the client.
If you have bugs to report, or you wish to submit code to improve DIRTSAND, you can visit the github project page.
To get the H-uru fork of the client, which has the best support for DIRTSAND servers, visit the H-uru/Plasma project.