mirror of
https://bitbucket.org/theswgsource/swg-main-1.2.git
synced 2026-01-15 22:04:54 -05:00
README.md edited online with Bitbucket
This commit is contained in:
SWG Source
170
README.md
170
README.md
@@ -1,41 +1,165 @@
|
||||
# Star Wars Galaxies Source Code (C++) Build Tools
|
||||
# SWGSource V1.2 Build Instructions
|
||||
|
||||
This is the main server code. Please feel free to contribute merge/pull requests, or fork the code to create variations. Please keep the license and credit information intact.
|
||||
## Credit
|
||||
Credit the StellaBellum team (DarthArgus, Cekis, SilentGuardian, Ness, SwgNoobs, DevCodex) for making their repositories open. All source is forked from those repositories and progressed from that point.
|
||||
|
||||
The majority of the work is thanks to Devcodex, with more fixes and optimizations provided by DarthArgus.
|
||||
## What Do You Need To Do To Get A Server Running?
|
||||
|
||||
# src Repo:
|
||||
# Main Branches
|
||||
* master - stable, no debug flags
|
||||
* testing - bleeding edge, works in progress, probably stable
|
||||
Most of what is contained in this section assumes you're running a Linux base VM - specifically Debian. It is assumed your home directory is /home/swg.
|
||||
|
||||
# Works in progress
|
||||
* testing-64 - fully 64 bit version thatbuilds but doesn't run, some typedefs and things are wrong
|
||||
Download the private main repo (replace the <git_user_name> part with your GIT user name):
|
||||
|
||||
# Building
|
||||
cd /home/swg
|
||||
git clone https://<git_user_name>@bitbucket.com/theswgsource/swg-main-1.2.git swg-main
|
||||
|
||||
Only use the Debug and Release targets unless you want to work on 64 bit. For local testing, and non-live builds set MODE=Release or MODE=debug in build_linux.sh.
|
||||
When the GIT repository has been cloned successfully, open the swg-main directory:
|
||||
|
||||
For production, user facing builds, set MODE=MINSIZEREL for profile built, heavily optimized versions of the binaries.
|
||||
cd /home/swg/swg-main
|
||||
|
||||
Create a .setup file such that the build script knows where to start:
|
||||
|
||||
## Profiling and Using Profiles
|
||||
touch .setup
|
||||
|
||||
To generate new profiles, build SWG with MODE=RELWITHDEBINFO.
|
||||
## Building
|
||||
The process is nearly fully automated by executing the build_linux.sh script located in swg-main. The script will ask you a series of questions with each
|
||||
question corresponding to a different step in the build process. Each step is defined below. To complete building, kick off the build script and follow
|
||||
these instructions as detailed here to understand each step.
|
||||
|
||||
Add export LLVM_PROFILE_FILE="output-%p.profraw" to your startServer.sh file.
|
||||
### Starting the Build Process
|
||||
Follow the instructions of the build script. The binary building phase will take roughly 1 hour for each CPU core assigned to your VM. Do not skip any steps.
|
||||
Execute the following script to start building (see detail on each step below this):
|
||||
|
||||
WHILE THE SERVER IS RUNNING do a ps -a to get the pid's of each SWG executable. And take note of which ones are which.
|
||||
./build_linux.sh
|
||||
|
||||
!! Ignore the install dependencies PART !!
|
||||
|
||||
After you cleanly exit (shutdown) the server, and ctrl+c the LoginServer, move each output-pid.profraw to a folder named for it's process.
|
||||
### GIT Phase
|
||||
Upon executing the build script, it will ask you for your GIT username. This is needed in order to get the source code correctly. You can obviously supply any
|
||||
name you wish here.
|
||||
|
||||
Then, proceed to combine them into usable profiles for the compiler:
|
||||
What is your GIT username (so we can get the code correctly):
|
||||
|
||||
#### Cloning the Repositories
|
||||
The script will first ask if you would like to retrieve any GIT folders necessary (it knows which ones it needs to get). It will attempt to get the pre-configured
|
||||
repositories that are outlined in the build script by cloning them. It will simply do a pull if the repositories already exist to bring them up to date.
|
||||
This currently includes "src" and "dsrc":
|
||||
|
||||
llvm-profdata merge -output=code.profdata output-*.profraw
|
||||
Do you want to pull/update git? (y/n)
|
||||
|
||||
If you answer "y" to the question, it will attempt to retrieve the latest version set of the code. If you hit enter or choose "n" it will skip to the next step.
|
||||
|
||||
Finally, then replace the profdata files with the updated versions, within the src/ tree.
|
||||
### Compile Phase
|
||||
After getting the latest version of the code, the build script will then attempt to build the "src" and "dsrc" folders.
|
||||
|
||||
See http://clang.llvm.org/docs/UsersManual.html#profiling-with-instrumentation for more information.
|
||||
#### Compile Mode
|
||||
The system first needs to know how you will be compiling the code. You have a choice of either DEBUG (d) mode or RELEASE (r) mode. If you do not enter "d" or "debug"
|
||||
for the mode then it is assumed you wish to build a RELEASE build.
|
||||
|
||||
# Buy Darth A Caffinated Beverage
|
||||
Is this for DEBUG mode or RELEASE mode? (d/r):
|
||||
|
||||
bitcoin:16e1QRRmnBmod3RLtdDMa5muKBGRXE3Kmh
|
||||
#### Compiling SRC
|
||||
The first compile step is to build the server C++ code. This is the source found in the "src" folder. Once compiled, you should have compiled code in the "build"
|
||||
directory of swg-main:
|
||||
|
||||
Do you want to recompile the server code (C++) now? (y/n)
|
||||
|
||||
#### Building the Server Configuration Files
|
||||
The next step will attempt to build the configuration files. This step captures data needed for the server to execute properly. The following information is captured in this step:
|
||||
|
||||
Do you want to build the config environment now? (y/n)
|
||||
Enter your IP address (LAN for port forwarding or internal, outside IP for DMZ)
|
||||
|
||||
Enter the IP address of the host machine that will be the server. This is typically your VM IP Address.
|
||||
|
||||
Enter the DSN for the database connection
|
||||
|
||||
Enter the name of the database. This is typically just "//127.0.0.1/swg" without quotes. It uses the 127.0.0.1 localhost loopback instead of the VM IP on purpose.
|
||||
|
||||
Enter the database username
|
||||
|
||||
Enter the user name for the database connection. This is also typically just "swg" without quotes.
|
||||
|
||||
Enter the database password
|
||||
|
||||
Enter the password for the database connection. This is also typically just "swg" without quotes.
|
||||
|
||||
Enter a name for the galaxy cluster
|
||||
|
||||
This will be the cluster name of the galaxy. Use something creative (or not), this is the name of the galaxy you're going to choose when creating a new character.
|
||||
|
||||
#### Compiling DSRC
|
||||
The next step will be to build the Java code that resides in the dsrc folder. This code is where 99% of the game content and AI resides. This step takes a long time
|
||||
so go watch a movie or something while it builds. Note, it will give you the option of using "multi" or "safe" building. Multi takes advantage of multi-threaded
|
||||
compiling (typically faster). However, it was found that using it SOMETIMES caused issues, but these issues are no likely around any longer. Needless to say,
|
||||
if you encounter issues using MULTI, try using Safe instead which does not use multi-threaded compiling:
|
||||
|
||||
Do you want to recompile the scripts (.java)? (y/n)
|
||||
|
||||
#### Compiling the mIFF files
|
||||
This step will compile all *.mif files into *.iff binary files.
|
||||
|
||||
Do you want to build the mIFF files (.mif)? (y/n)
|
||||
|
||||
#### Compiling the Datatable files
|
||||
This step will compile all the *.tab files into *.iff binary files. Again, speed constraints aside, you have the choice of using MULTI or SAFE for compiling (see DSRC Compiling above):
|
||||
|
||||
Do you want to build the datatables (.tab)? (y/n)
|
||||
|
||||
#### Compiling Template Files
|
||||
This step will compile all the *.tpf files into *.iff binary files. You have a choice of Multi or Safe here as well:
|
||||
|
||||
Do you want to build the template files (.tpf)? (y/n)
|
||||
|
||||
If you have built the TPF files in this step (i.e. you didn't skip this step) then the build script will also attempt to recreate the Object Template and Quest CRC
|
||||
tables and subsequently will attempt to push those changes to the database since this will also be required. A GREAT feature to have when creating new
|
||||
template files or changing existing ones.
|
||||
|
||||
### Database Phase
|
||||
#### Building Object Template and Quest CRC Files
|
||||
This step will compile the object template and quest CRC files. These files translate the long name of these files (including file path) into a very short code
|
||||
that allows the server to identify them without the danger of long text being transferred over the internet in packets. Basically a optimization that SOE implemented:
|
||||
|
||||
Do you want to build the Object Template or Quest CRC files? (y/n)
|
||||
|
||||
Building these files will also trigger the build script to then populate the database with the CRC's that were generated. If you are doing this build script
|
||||
in pieces (i.e. you're selectively building), this is a GREAT way to re-import new or changed TPF file changes. In order to re-import CRC's into the database,
|
||||
if you haven't already entered it above, it will ask you for the database information here.
|
||||
|
||||
#### Importing the Database SQL
|
||||
This final step will attempt to create the database schema from the SQL files that are in the src directory structure. It will create all database tables, functions,
|
||||
views and stored procedures the game needs to use. You can run this step without fear of ruining anything that is currently in place. Keep in mind that if you do
|
||||
attempt to import the database over the top of an existing swg database, it will simply throw errors after a very short attempt, but will not cause issue with
|
||||
your current data. Basically, you can't mess this part up:
|
||||
|
||||
Import database? (y/n)
|
||||
|
||||
If you haven't previously entered the necessary database information above, it will ask you for that information here.
|
||||
|
||||
## Clientdata and Appearance Files
|
||||
The Clientdata Repo is copied from another location (i.e. not on swgsource at the moment) its in /home/swg/clientdata. You MUST have this in place and have the
|
||||
symbolic link created (see below) before the server can execute successfully. Keep this in mind when you get errors while trying to start the server:
|
||||
|
||||
* mkdir ~/swg-main/data/sku.0/sys.client/compiled
|
||||
* ln -s ~/clientdata ~/swg-main/data/sku.0/sys.client/compiled/clientdata
|
||||
|
||||
## Manual Modifications
|
||||
You need to make the following modifications manually to make sure the clientdata directory can be found and read by the server at runtime:
|
||||
|
||||
Add (or replace because this is all you should have for [SharedFile]) the following lines to exe/shared/servercommon.cfg:
|
||||
|
||||
[SharedFile]
|
||||
searchPath2=../../data/sku.0/sys.shared/compiled/game
|
||||
searchPath2=../../data/sku.0/sys.server/compiled/game
|
||||
searchPath1=../../data/sku.0/sys.shared/built/game
|
||||
searchPath1=../../data/sku.0/sys.server/built/game
|
||||
searchPath0=../../data/sku.0/sys.client/compiled/clientdata
|
||||
|
||||
|
||||
## First Start
|
||||
To start the server after building, execute the following script:
|
||||
|
||||
./startServer.sh
|
||||
|
||||
Point your login.cfg in your game folder (on you client machine, NOT the VM) to the IP address of the Virtual Machine such that you can connect successfully.
|
||||
|
||||
AND YOUR DONE!
|
||||
Reference in New Issue
Block a user