By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Removes continue in socket accept flow
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This pull request adds BEAMMP_PROVIDER_DISABLE_MP_SET variable
(true/false/1/0) to disable MP::Set for providers.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR adds a boolean parameter to the MP.SendChatMessage function that
allows logging the message or not (default true).
I took an example from the ``set_function("SendNotification" ...`` to
make the default value working.
It's my first time actually doing C++, I hope it's alright!
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Testing builds with -Wl,--export-dynamic
Closes#464
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
## Summary
Reject malformed vehicle reset and paint packets before slicing
structured payloads, and catch parser exceptions in the TCP client loop.
## Problem
Or:<pid>-<vid>:null and similarly malformed Op packets could reach
substr(npos) in ParseVehicle, throwing std::out_of_range. On TCP, that
exception could escape the client loop and kill the thread.
## Changes
- Guard Or reset packets when no { payload exists
- Guard Op paint packets when no [ payload exists
- Add a small helper and unit test for structured payload extraction
- Catch std::exception around GlobalParser in the TCP client loop and
disconnect the offending client
## Verification
- git diff --check
- Full build/tests not run in this shell because cmake was not installed
add/test Debian 13 build to github actions
Closes#464
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Decreases the likelyhood of lua stack corruption.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Update Lua version to make plugin development easier across platforms
and simplify Windows builds by using vcpkg.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
- Avoid redundant substr() calls in packet parsing hot-path
(TServer.cpp)
The previous code called substr(3) twice per packet, creating
unnecessary temporary strings. Now stores the result once.
- Replace .size() == 0 with .empty() for idiomatic C++
(TConsole.cpp, TLuaEngine.cpp)
When Acceptor.accept() returns an error (e.g., "Too many open files"),
the server was continuing to process an invalid socket, causing resource
leaks and potential infinite error loops.
Add continue statement to skip processing when accept() fails, allowing
the server to retry on the next iteration instead of crashing.
Fixes resource exhaustion DoS vulnerability where server would enter
error loop instead of handling gracefully.
<img width="1233" height="199" alt="image"
src="https://github.com/user-attachments/assets/bad8f559-6ef2-47ee-b1c1-3e6020cdfb77"
/>
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
When Acceptor.accept() returns an error (e.g., "Too many open files"),
the server was continuing to process an invalid socket, causing resource
leaks and potential infinite error loops.
Add continue statement to skip processing when accept() fails, allowing
the server to retry on the next iteration instead of crashing.
Fixes resource exhaustion DoS vulnerability where server would enter
error loop instead of handling gracefully.
Right now server only checks if chat message is empty but doesnt check
the max length. Client limits to 500 chars but if someone modifies the
client they can send huge messages. I added a check on server side to
reject messages longer than 500 characters, same as client limit. I used
translator because I don't know English well
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Right now server only checks if chat message is empty but doesnt check the max length. Client limits to 500 chars but if someone modifies the client they can send huge messages.
I added a check on server side to reject messages longer than 500 characters, same as client limit.
I used translator because I don't know English well
- Avoid redundant substr() calls in packet parsing hot-path (TServer.cpp)
The previous code called substr(3) twice per packet, creating
unnecessary temporary strings. Now stores the result once.
- Replace .size() == 0 with .empty() for idiomatic C++
(TConsole.cpp, TLuaEngine.cpp)
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR implements a new lua function and packet used for sends dialogs
to the client.
## Example:
https://github.com/user-attachments/assets/97bb5813-ea12-4b1d-a049-2f7ebf6b6da3
Example serverside code:
```lua
--MP.ConfirmationDialog(player_id: number, title: string, body: string, buttons: object, interaction_id: string, warning: boolean = false, reportToServer: boolean = true, reportToExtensions: boolean = true)
function onChatMessage(player_id, player_name, message)
MP.ConfirmationDialog(player_id, "Warning", "Watch your tone buddy!!",
{
{
label = "OK",
key = "dialogOK",
isCancel = true
}
}, "interactionID", true)
end
MP.RegisterEvent("onChatMessage", "onChatMessage")
function dialogOK(player_id, interaction_id)
MP.SendChatMessage(-1, MP.GetPlayerName(player_id) .. " clicked OK")
end
MP.RegisterEvent("dialogOK", "dialogOK")
```
### Details:
Each dialog can have multiple buttons, each button having it's own
callback event (`key`).
Each dialog can also have one button with `isCancel` being true,
settings this property to true causes the button's event to be called
when the users pressed `esc` to exit out of the dialog. If a dialog is
created without any button being the cancel button then the user will
only be able to exit the dialog by restarting the session or pressing
one of the buttons.
`interaction_id` will be sent as the event data with a button press
event, to track from which dialog the button press came. As when
multiple dialogs are opened they will stack and it will become difficult
to track what button on which dialog was pressed without having multiple
event handlers.
Waiting on https://github.com/BeamMP/BeamMP/pull/715 to be merged.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR adds a new console command (`nettest`) that sends a request to
the server check api in order to test connectivity via the server's
public ip (serverlist entry).
- [x] https://github.com/BeamMP/ServerCheck/pull/2
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Converts the vehicle stored client side from a raw string to parsed json
data. This allows us to more easily edit the vehicle state serverside,
which I've started using in this PR for updating the state after a paint
packet.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Adds `BEAMMP_PROVIDER_IP_ENV` for hosting panels, which allows the
server owner to configure which env var is read to get the ip interface
to bind to.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
# Mod database
This PR adds a local database of mods, which is used to cache mod hashes
and protection status.
## Mod hash caching
Mod hashes will now be cached based on last write date. This will speed
up server startup because only the mods with changes will have to be
hashed.
## Mod protection
You can now protect mods! This will allow you to host a server with
copyrighted content without actually hosting the copyrighted content.
Just run `protectmod <filename with .zip> <true/false>` in the console
to protect a mod. Users that join a server with protected mods will have
to obtain the file themselves and put it in their launcher's resources
folder. The launcher will inform the user about this if the file is
missing.
## Mod reloading
You can now reload client mods while the server is running by using
`reloadmods` in the console. Keep in mind that this is mainly intended
for development, therefore it will **not** force client to rejoin and
neither will is hot-reload mods on the client.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR fixes an issue where players would get personal events during
downloads, which would corrupt the download and block the user from
being able to properly join the server.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Reverts a PR that has been causing sol to crash.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR adds an option in the server config to bind to a custom IP.
- [x] Review config comment and make sure that it is not confused with
port forwarding
- [x] Validate IP addresses
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR makes it so that connections are denied if lua hasn't loaded
yet, and makes it so that lua waits for the server to load before
accessing then uninitialized memory.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Add ENV for missing settings.
Issue: https://github.com/BeamMP/BeamMP-Server/issues/414
Please let me know if there are any issues.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Force the use of ipv4 for backend requests because the launcher doesn't
support ipv6 yet
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR fixes the segfault caused by rapidjson on the first heartbeat,
and the memory leak that happens during mod hashing.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR converts all the http client code to use curl instead of
httplib. This will hopefully do something about the connectivity issues
with the backend and auth.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Logs the reponses from authentication and backend.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Closes#401
Replace deprecated `ip::address::from_string` with `ip::make_address` to
avoid build errors with Boost versions >= 1.87.0
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Just cleaning up some sentry related code, mentions, etc.
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
#398
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Closes#393 and #394
This PR cleans up and fixes the code in TNetwork::Authentication and
sends the kick reason to lua.
The event is now `function postPlayerAuth(Kicked: bool, Reason: string,
Name: string, Role: string, Guest: bool, Identifiers: table)`
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
Swap the sequence of the build instruction steps 2 and 3 so it actually
works
No AI involved, i used my own brain to mess it up and to fix it 😉
---
By creating this pull request, I understand that code that is AI
generated or otherwise automatically generated may be rejected without
further discussion.
I declare that I fully understand all code I pushed into this PR, and
wrote all this code myself and own the rights to this code.
This PR adds an "I" packet which returns the server information. This
can be used by external programs and the launcher to get information
about a server without having to connect it to. It can be toggled in the
config and in lua.
This can be used to check if UDP works :)
I've tested this a bit with snep ingame, even spamming this on localhost
with a large number of connections seems to not impact gameplay at all.
Adds `MP.GetPlayerRole(player_id)` to the Lua API to get a player's role
("USER", "EA", "MDEV", "STAFF", "ET") by their player id.
Currently you can only get someone's role in onPlayerAuth from the
parameters and in onVehicleSpawned and onVehicleEdited from the packet
data, but not in onPlayerJoin for example without storing it.
With this PR, returning 2 in onPlayerAuth will allow the player to join
without checking if the server is full. This makes it easier for plugin
developers to allow for example their staff to join without having to
change the max player count.
Adds `MP.Get(ConfigID)` to the lua api to get the current server
settings.
```lua
lua> print(MP.Get(MP.Settings.Name]))
[LUA] BeamMP Server
lua> MP.Set(MP.Settings.Name, 'Hello World')
[INFO] Set `Name` to Hello World
lua> print(MP.Get(MP.Settings.Name))
[LUA] Hello World
```
Closes#146
Adds `post*` events which are triggered after the respective `on*` event
has completed and the results have been sent.
They have the same arguments as the `on*` function, with the exception
that another argument is added in the beginning which contains whether
the `on*` variant was cancelled.
this change replaces the `_G` event handler lookup with a lua snippet
that fetches whatever the user has set, which could be a module's
function, a builtin global, or even a function defined in the
`RegisterEvent `call
Adds IPv6 support.
FreeBSD users beware: From this point forward, you will have to set
`sysctl net.inet6.ip6.v6only=0` so that the BeamMP-Server continues to
work. The server also prints this information.
When adding elements to a table, the .add() function does not behave as
expected in various cases, making it really shit and difficult to use.
Instead, we keep our own index and just add one by one. It works, and
it's easy to understand.
This may lead to indices which are nil, i.e. non-fully-sequential
tables, but I can't be asked to worry about it because that shouldn't be
an issue when we use .set() everywhere.
When adding elements to a table, the .add() function does not behave as
expected in various cases, making it really shit and difficult to use.
Instead, we keep our own index and just add one by one. It works, and
it's easy to understand.
This may lead to indices which are nil, i.e. non-fully-sequential
tables, but I can't be asked to worry about it because that shouldn't be
an issue when we use .set() everywhere.
Adds `Util.DebugExecutionTime()`, which returns a table of
`function_name: milliseconds`, in which each function's execution time
is averaged over all time.
You can call this function like so:
```lua
-- event to print the debug times
MP.RegisterEvent("printStuff", "printStuff")
-- prints the execution time of all event handlers
function printStuff()
print(Util.DebugExecutionTime())
end
-- run every 5 seconds (or 10, or 60, whatever makes sense for you
MP.CreateEventTimer("printStuff", 5000)
```
Pretty print function:
```lua
function printDebugExecutionTime()
local stats = Util.DebugExecutionTime()
local pretty = "DebugExecutionTime:\n"
local longest = 0
for name, t in pairs(stats) do
if #name > longest then
longest = #name
end
end
for name, t in pairs(stats) do
pretty = pretty .. string.format("%" .. longest + 1 .. "s: %12f +/- %12f (min: %12f, max: %12f) (called %d time(s))\n", name, t.mean, t.stdev, t.min, t.max, t.n)
end
print(pretty)
end
```
`Util.DebugExecutionTime()` returns a table, where each key is an event
handler function name, and each value is a table consisting of `mean`
(simple average), `stddev` (standard deviation aka mean of the
variance), `min` and `max`, all in milliseconds, as well as `n` as the
number of samples taken.
This PR adds a tags functionality to the BeamMP server config and
heartbeat.
The intended purpose of this is to enable better filtering on the server
list based on keywords such as gamemode or roleplay.
before, the handlers were not waited for, so the client was usually
destructed before lua got to the actual event handler call. Now, the
handler is called and waited on, and once all handlers are done, the
client is properly removed from the players internally, thus making
calls to GetPlayerName, GetPlayerIdentifiers, etc. return nil etc.
"PPS has no meaning anymore and is completely irrelevant. You should
ignore it, it is not an indicator of ANYTHING. If it’s high, that means
NOTHING. If it’s low, that means NOTHING. If it’s -, that means
NOTHING."
It's packets per second per player per vehicle, but is only sent every
30 seconds, its not averaged, and on the client-side, it shows a ping
icon next to it.
A client can open a new connection to the server and send a `P`, and
measure the time to the `P` response packet. The connection is then
closed. This was added ages ago, please use this instead for ping :)
- replaced all networking with boost::asio abstractions
- rewrote the SetStatus(), GetStatus() crap (now IsDisconnected() and
Disconnect() with reason)
- fixed bug related to ghost players / ghost cars
- handles packets as binary for the most part (vector<uint8_t> instead
of string) to fix various issues
Before, a lot of common errors went unnoticed, due to insufficient
compiler diagnostics. This commit fixes this by adding a lot of new
diagnostics, and fixing the issues found by this.
The event loop tries to run no faster than every 10ms. If it detects
that it goes faster, it would incorrectly calculate the difference, and
then wait (what I assume was) way too long or too short.
Either way, now it's fixed and it correctly works, even when introducing
new lua states.
onInit is now called on hot-reload, for the whole plugin. Arguably, this
is not expected behavior, since only one file is being reloaded, but
this is the easiest way to do it, and the entire hot-reload process is
only for development purposes. Open an issue if this breaks your stuff
:^)
While this doesn't make a difference with usual setups, it does if the
server path exists but is a symlink. In that case, the
create_directories call fails, and the server aborts.
this fixes that. :^)
@jimkoen
This was removed because, as useful and as much work as this was, we
can't reasonably take responsibility for this. Instead, a server like
this should *always* be localhost only, and if it's not, it should be
behind an nginx reverse proxy anyways. We're removing the config options
regarding this in one of the next commits.
There are two CallStrategies:
- BestEffort (default): Will try to get your event to trigger at the specified
interval, but will refuse to queue handlers if a handler takes too
long.
- Precise: Will enqueue event handlers at the exact interval specified.
Can lead to the queue filling up if the handler takes longer than the
interval.
CMake will now find packages in a modern way (include(Find*)), and will
also ensure that submodules are updated, unless told otherwise.
Also removed some apple-specific workarounds, we will need to look at
that again.
This fixes an issue where the log file is written to the original
directory, even if --working-directory=path was used. This can obviously
be pretty bad.
This fixes an issue where the log file is written to the original
directory, even if --working-directory=path was used. This can obviously
be pretty bad.
* Add preliminary work for HTTP health endpoint
* Http: Fix infinite loop bug in Tx509KeypairGenerator::generateKey()
* update commandline
* Add TLS Support class for use with http server
* Add preliminary HTTP Server; TLS still broken; fix in later commit
* Fix TLS handshake, due to server being unable to serve key/certfile in 'Http.h/Http.cpp'; Cause was httlib not being threadsafe due to being a blocking http library
* Run clang format
* Add option to configure http server port via ServerConfig
* TConfig: add HTTPServerPort to config parsing step
* Fix SSL Cert / Key path not auto generating when not existing
* Add health endpoint; Fix SSL Cert serial no. not refreshing when regenerating
* Switch arround status codes in /health route
* Run clang format
Co-authored-by: Lion Kortlepel <development@kortlepel.com>
Fix sentry url length print
remove quotes
github actions is cursed
add debug print
test action
Dont use curl on windows
I dont know why the windows build doesnt report to sentry, so ill try
this.
Change timeout to 20 minutes instead of 5
this is a hacky workaround anyways, so i really dont see why it should
only be 5. 5 is barely enough.
temporarily enable debug mode on sentry
CMake: Use breakpad on windows instead of crashpad
CMake: Sentry: use inproc backend
Since cmake refuses to set my variables, I will do it this way.
I am so tired of this github workflow garbage
Sentry: disable debug again, set
sentry_options_set_symbolize_stacktraces to true, fix memory leak
Sentry: hotfix: dont free options
somehow that causes it to crash, and i cannot be bothered to find out
why right now
Possible fix for sentry url not showing up in windows build
possible fix for sentry, again
add static_assert in attempt to fix issue with sentry
use target_compile_definitions instead of add_compile_definitions
this gets rid of a bunch of unclear cases which I mistakenly created
while refactoring for this rewrite. One example is having to call into
TTCPServer to do UDP sending in some cases.
* update workflows to accomodate new requirements
* add boost to vkpkg
* try without boost again
* use boost just boost
* might cache vcpkg or might not idk
* dont pull everything in boost
* clean up includes
* fix 1
* fix 2
* fix 3
* fix 4
* fix id 6
* remove libssl again
* move to boost 1.71
* add boost to windows cmake
* change from boost 1.71 to 1.70
* fix cmake again
* remove version because f cmake
* remove stuff again
* fix my mistakes
* fix linker args for unix
* openssl
* add openssl to vcpkg
* uhh change whats linked
* rename OpenSSL to OPENSSL????
**Is your feature request related to a problem? Please describe.**
A clear and concise description of what the problem is. For example: "I'm always frustrated when ...".
**Describe the solution you'd like**
A clear and concise description of what you want to happen. Also supply OS information if relevant, for example "*On Linux*, I would like to be able to...".
**Describe alternatives you've considered**
A clear and concise description of any alternative solutions or features you've considered.
**Additional context**
Add any other context about the feature request here.
Please replace this text <-> with your PR description and leave the below declarations intact.
---
By creating this pull request, I understand that code that is AI generated or otherwise automatically generated may be rejected without further discussion.
I declare that I fully understand all code I pushed into this PR, and wrote all this code myself and own the rights to this code.
- `BeamMP-Server.$DISTRO.$DISTROVERSION.$ARCH` for linux builds, for example `BeamMP-Server.debian.11.x86_64` for the Debian 11 build for x86_64. All require `liblua5.3` to be installed.
- `BeamMP-Server.exe` for the Windows build (x86_64). You need to install the [Visual C++ Redistributables](https://aka.ms/vs/17/release/vc_redist.x64.exe) to run this.
Unlike other parts of BeamMP, the BeamMP-Server does not have any dependency to the BeamNG.drive game.
To contribute *C++ code*, you'll need a MacOS, Linux or Windows PC, and intermediate to advanced knowledge of C++.
For reference, you should know be reasonably comfortable with the STL, the concept of RAII, templates, and generally know how to read & write post-C++17 code. To contribute anything else, you won't need most of this (though it'd be helpful to have some vocabulary about computers).
# Ways to Contribute
## Bug Reports
If you work with BeamMP-Server, either by simply using it, or even writing plugins for it, and you run into any issues, we definitely want to know about it. Please use [GitHub issues](https://github.com/BeamMP/BeamMP-Server/issues) and select the "Bug" template, read it, and fill it out accordingly.
## Bug Fixes
If you are interested in fixing bugs, check out the [GitHub issues](https://github.com/BeamMP/BeamMP-Server/issues). There, you can pick any issue that has nobody assigned to it. For example, some bugs which we definitely need some help with are marked with the "help wanted" tag.
Once you picked a bug, you need to reproduce it. Start by following the instructions in the bug report, and don't be afraid to ask for more information or clarification on the issue itself.
Refer to [getting started with the codebase](#getting-started-with-the-codebase) for more information on how to build the server. You can also ask on our [Discord server](https://discord.gg/beammp), or on IRC ([irc.libera.chat](https://web.libera.chat/), join `#beammp`).
## Features
If you want to add new features, please make an issue for it first or ask on our [Discord server](https://discord.gg/beammp), or on IRC ([irc.libera.chat](https://web.libera.chat/), join `#beammp`).
You need to make sure the feature isn't being worked on by someone else, and aligns with the vision we have for the server.
# Git Guidelines
**Read this carefully. Failing to follow these rules results in your changes not being accepted**. This applies for outside contributors, members of the BeamMP development team ("BeamMP Developers"), project owners, maintainers, frequent contributors, and literally everyone else. **It applies to everyone**.
## How to Commit
Commit messages **MUST** (mandatory):
- start with a **lower case action verb in present tense**, for example `add`, `fix`, `implement`, `refactor`, `remove`, `rename`. *Counter examples (these are bad): ~~`Fixed`, `fixing`, `added`, `removing`~~*.
- not have a first line much longer than 70 characters.
- explain briefly the changes made.
- reference the issue by number, if there is an issue the commit addresses, like so: `#<number>`. Example: `#123`.
If any of these are not followed, **your changes will not be accepted.**
Commit messages **SHOULD** (optional, "nice to have"):
- only address one "atomic" change.
- have an empty second line, and the subsequent lines explaining the changes in more detail (if more detail is available).
Commits may be squashed (via a Git "interactive rebase") in order to satisfy these rules, but history that is >1h old should not be rewritten if possible. Force pushes are ugly ;)
## Pulling, Merging
Do **NOT** pull with merge. This is the default git behavior for `git pull`, but creates ugly and unnecessary commit messages like `"merge origin/master into master"`. Instead, pull with rebase, for example via `git pull -r`. If you get conflicts, resolve them properly.
The only acceptable merge commits are those which actually merge functionally different branches into each other, for example for merging one feature branch into another.
## Workflow
### Making an issue and fixing it
1. Create an issue detailing the feature or bug.
2. Assign a milestone to the issue, or wait for a maintainer to add a milestone to your issue.
3. Fork the repository and base your work on the branch mentioned in the milestone attached to your issue (e.g. `v3.0.0 (develop)` -> `develop`).
4. Program your feature or bug fix.
5. Open a PR that references the issue by number in the format: `#12345`.
6. Someone will review your PR and merge it, or ask for changes.
### Fixing an existing issue
1. Fork the repository and base your work on the branch mentioned in the milestone attached to your issue (e.g. `v3.0.0 (develop)` -> `develop`).
2. Program your feature or bug fix.
3. Open a PR that references the issue by number in the format: `#12345`.
4. Someone will review your PR and merge it, or ask for changes.
## Branches
-`minor`: Minor releases, like `v1.2.3` -> `v1.3.0` or `v1.2.3` -> `v1.2.4`.
-`develop`: Major releases, like `v1.2.3` -> `v2.0.0`, and larger feature/minor releases.
## Unit tests & CI/CD
We use GitHub Actions, which runs our unit-tests. PR's which fail these tests, or even fail any of our actions (which run automatically), will not be merged and require further changes until they compile, link, and all tests pass properly. If you have issues with this, feel free to ask in our [Discord server](https://discord.gg/beammp), or on IRC ([irc.libera.chat](https://web.libera.chat/), join `#beammp`)
### What should I call by branch?
Keep branch names **unique**, **descriptive**, and **shorter than 30 characters**. Names must be in present-tense, such as `fix-xyz`, **not**~~`fixing-xyz`~~.
We generally use *feature branches*, so we keep one branch per feature or fix.
For example:
- You want to fix issue number #123? You could call the branch `fix-123`.
- You want to add a feature described in issue #456? You could call the branch `implement-456`.
- You want to add a feature or fix a bug that has no issue? You should probably make an issue for it first! Or, if you're not ready for that yet, you could call it by the feature name or bug description, for example for a bug that makes cars disappear: `fix-disappearing-cars`.
## Pull Requests, Code Review
Once you are ready to show what you did, and get feedback on it, you open a Pull-Request on GitHub. Please make sure to pick the right branches, and a descriptive title. Mention any related issues with `#<issue number>`, for example `#123`.
Make sure to explain what the PR does, what it fixes, and what needs to still be done (if anything).
A BeamMP-Developer must review your code in detail, and leave a review. If this takes too long, feel free to @ a maintainer/developer, or leave another comment on the PR. It helps to say something like "Ready for review", for example.
# Getting Started with the Codebase
1. Look at current Pull-Requests, look at the git branches, or ask in our [Discord server](https://discord.gg/beammp), or on IRC ([irc.libera.chat](https://web.libera.chat/), join `#beammp`), in order to find out which branch you should work on to minimize conflict. See [this section on branches](#branches) for more info.
2. Fork the repository (with that branch) on GitHub. GitHub, by default, gives you only the `master` branch when forking, so make sure you fork with other branches enabled when you want to work on a branch that isn't master (it's a checkbox when you fork).
3. Clone the fork to your local machine.
4. Check out the branch you want to work on (see 1.).
5. Run `git submodule update --init --recursive`.
6. Make a new branch for your feature or fix from the branch you are on. You can do this via `git checkout -b <branch name>`. See [this section on branches](#branches) for more info on branch naming.
7. Install all dependencies. Those are usually listed in the `README.md` in the branch you're in, or, more reliably, in one of the files in `.github/workflows` (if you can read `yaml`).
8. Run CMake to configure the project. You can find tutorials on this online. You will want to tell CMake to build with `CMAKE_BUILD_TYPE=Debug`, for example by passing it to CMake via the commandline switch `-DCMAKE_BUILD_TYPE=Debug`. An example invocation on Linux with GNU make would be
`cmake . -DCMAKE_BUILD_TYPE=Debug` .
9. Build the `BeamMP-Server` target to build the BeamMP-Server, or the `BeamMP-Server-tests` target to build the unit-tests (does not include a working server). In the example from 8. (on Linux), you could build with `make BeamMP-Server`, `make -j BeamMP-Server` or `cmake --build . --parallel --target BeamMP-Server` . Or, on Windows, (in Visual Studio), you would just press some big green "run" or "debug" button.
10. When making changes, refer to [this section on how to commit properly](#how-to-commit). Not following those guidelines will result in your changes being rejected, so please take a look.
11. Make sure to add Unit-tests with `doctest` if you build new stuff. You can find examples all over the latest version of the codebase (search for `TEST_CASE`).
# Code Guidelines
## Formatting
1. Use `clang-format` to format your code before committig. A `.clang-format` file is provided in the root of the repository.
2. All identifiers, type names, function names, etc. should be `PascalCase`. Type names may also have the `T` prefix, although this is not enforced (for example `TNetwork`).
## Modular code
Write code that is modular and testable. Generally, if you can write a good unit-test for it, it's modular. If you can't, it's not.
Don't overdo it though - sometimes its okay to just write code, do the job, be done with it. You'll get feedback on this in the code review for your PR.
[](https://github.com/BeamMP/BeamMP-Server/actions?query=workflow%3A%22CMake+Windows+Build%22)
[](https://github.com/BeamMP/BeamMP-Server/actions?query=workflow%3A%22CMake+Linux+Build%22)
This is the server for the multiplayer mod **[BeamMP](https://beammp.com/)** for the game [BeamNG.drive](https://www.beamng.com/).
The server is the point through which all clients communicate. You can write Lua mods for the server, there are detailed instructions on the [BeamMP Wiki](https://wiki.beammp.com).
**For Linux, you __need__ the runtime dependencies, which are listed below under [Runtime Dependencies](#runtime-dependencies)**
## Support + Contact
Feel free to ask any questions via the following channels:
- **Discord**: [click for invite](https://discord.gg/beammp)
- **BeamMP Forum**: [BeamMP Forum Support](https://forum.beammp.com/c/support/33)
## Minimum Requirements
These values are guesstimated and are subject to change with each release.
* RAM: 30-100 MiB usable (not counting OS overhead)
* CPU: >1GHz, preferably multicore
* OS: Windows, Linux (theoretically any POSIX)
* GPU: None
* HDD: 10 MiB + Mods/Plugins
* Bandwidth: 5-10 Mb/s upload
## Contributing
TLDR; [Issues](https://github.com/BeamMP/BeamMP-Server/issues) with the "help wanted" or "good first issue" label or with nobody assigned.
To contribute, look at the active [issues](https://github.com/BeamMP/BeamMP-Server/issues). Any issues that have the "help wanted" label or don't have anyone assigned are good tasks to take on. You can either contribute by programming or by testing and adding more info and ideas.
Fork this repository, make a new branch for your feature, implement your feature or fix, and then create a pull-request here. Even incomplete features and fixes can be pull-requested.
If you need support with understanding the codebase, please write us in the Discord. You'll need to be proficient in modern C++.
## About Building from Source
We only allow building unmodified (original) source code for public use. `master` is considered **unstable** and we will not provide technical support if such a build doesn't work, so always build from a tag. You can checkout a tag with `git checkout tags/TAGNAME`, where `TAGNAME` is the tag, for example `v3.4.1`. See [the tags](https://github.com/BeamMP/BeamMP-Server/tags) for possible versions/tags, as well as [the releases](https://github.com/BeamMP/BeamMP-Server/releases) to check which version is marked as a release/prerelease. We recommend using the [latest release](https://github.com/BeamMP/BeamMP-Server/releases/latest).
## Supported Operating Systems
The code itself supports (latest stable) Linux, Windows and FreeBSD. In terms of actual build support, for now we usually only distribute Windows binaries and Linux. For any other distro or OS, you just have to find the same libraries listed in [Runtime Dependencies](#runtime-dependencies) further down the page, and it should build fine.
Recommended compilers: MSVC, GCC, CLANG.
You can find precompiled binaries under [Releases](https://github.com/BeamMP/BeamMP-Server/releases/).
## Build Instructions
On Linux, you need some dependencies to **build** the server (on Windows, you don't):
```
liblua5.3-dev curl zip unzip tar cmake make git g++
```
You can install these with your distribution's package manager. You will need sudo or need root for ONLY this step.
The names of each package may change depending on your platform.
If you are building for ARM (like aarch64), you need to run `export VCPKG_FORCE_SYSTEM_BINARIES=1` before the following commands.
You can build on **Windows, Linux** or other platforms by following these steps:
1. Check out the repository with git: `git clone --recursive https://github.com/BeamMP/BeamMP-Server`.
2. Go into the directory `cd BeamMP-Server`.
3. Specify the server version to build by checking out a tag: `git checkout tags/v3.4.1` - [Possible versions/tags](https://github.com/BeamMP/BeamMP-Server/tags)
4. Run CMake `cmake -S . -B bin -DCMAKE_BUILD_TYPE=Release` - this can take a few minutes and may take a lot of disk space and bandwidth.
5. Build via `cmake --build bin --parallel --config Release -t BeamMP-Server`.
6. Your executable can be found in `bin/`.
When you make changes to the code, you only have to run step 4 again.
### Building for FreeBSD
Building is only supported for major release branches of FreeBSD that are currently not EOL. The build process is the same as on Linux, although build dependencies can be universally installed from ports via pkg:
```
pkg install git cmake-core zip bash devel/ninja devel/pkgconf lua53
```
After installing the necessary build dependencies, follow the Linux build instructions beginning from step 3. Beware that running the initial cmake command will compile vcpkg from source, as vcpkg has no native FreeBSD port - this may take some time.
On systems with a single logical CPU core, `make` may fail to build the server when using the `--parallel` option when calling CMake. If you see error messages related to make, simply omit the `--parallel` from the command: `cmake --build bin --config Release -t BeamMP-Server`.
### Runtime Dependencies
These are needed to *run* the server.
Debian, Ubuntu and friends: `liblua5.3-0`
Other Linux distros: `liblua` of *some kind*.
Windows: No libraries.
## Support
The BeamMP project is supported by community donations via our [Patreon](https://www.patreon.com/BeamMP). This brings perks such as Patreon-only channels on our Discord, early access to new updates, and more server keys.
message(SEND_ERROR"git submodule update --init --recursive failed with ${GIT_SUBMOD_RESULT}, please checkout submodules. This may result in missing dependencies.")
endif()
else()
message(SEND_ERROR"git required for checking out submodules, but not found. Submodules will not be checked out - this may result in missing dependencies.")
endif()
endif()
if(Git_FOUND)
else()
message(STATUS"Git not found - the version will not include a git hash.")
message(SEND_ERROR"Checking out vcpkg in source tree failed with ${GIT_SUBMOD_RESULT}.")
endif()
else()
message(FATAL_ERROR"Could not find git or vcpkg.cmake. Please either, install git and re-run cmake (or run `git submodule update --init --recursive`), or install vcpkg and add `-DCMAKE_TOOLCHAIN_FILE=<path-to-vcpkg>/scripts/buildsystems/vcpkg.cmake` to your cmake invocation. Please try again after making those changes.")
beammp_errorf("Assertion failed in '{}:{}': {}.", __func__, _line, #cond); \
} \
} while (false)
#define beammp_assert_not_reachable() \
do { \
beammp_errorf("Assertion failed in '{}:{}': Unreachable code reached. This may result in a crash or undefined state of the program.", __func__, _line); \
// Copyright (C) 2024 BeamMP Ltd., BeamMP team and contributors.
//
// BeamMP Ltd. can be contacted by electronic mail via contact@beammp.com.
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published
// by the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
#pragma once
#include"Profiling.h"
#include"TNetwork.h"
#include"TServer.h"
#include<any>
#include<chrono>
#include<condition_variable>
#include<filesystem>
#include<initializer_list>
#include<list>
#include<lua.hpp>
#include<memory>
#include<mutex>
#include<nlohmann/json.hpp>
#include<queue>
#include<random>
#include<set>
#include<toml.hpp>
#include<unordered_map>
#include<vector>
#define SOL_ALL_SAFETIES_ON 1
#define SOL_USER_C_ASSERT SOL_ON
#define SOL_C_ASSERT(...) \
beammp_lua_errorf("SOL2 assertion failure: Assertion `{}` failed in {}:{}. This *should* be a fatal error, but BeamMP Server overrides it to not be fatal. This may cause the Lua Engine to crash, or cause other issues.", #__VA_ARGS__, __FILE__, __LINE__)
#include<sol/sol.hpp>
structJsonString{
std::stringvalue;
};
// value used to keep nils in a table or array, across serialization boundaries like
// JsonEncode, so that the nil stays at the same index and isn't treated like a special
// value (e.g. one that can be ignored or discarded).
// const std::regex time_regex(R"((\d+\.{0,1}\d*)(min|ms|us|ns|[dhs]))"); //i.e one of: "25ns, 6us, 256ms, 2s, 13min, 69h, 356d" will get matched (only available in newer C++ versions)
conststd::regextime_regex(R"((\d+\.{0,1}\d*)(min|[dhs]))");// i.e one of: "2.01s, 13min, 69h, 356.69d" will get matched
std::smatchmatch;
floattime_value;
if(!std::regex_search(time_str,match,time_regex))
returnstd::chrono::nanoseconds(0);
time_value=stof(match.str(1));
if(match.str(2)=="d"){
returnstd::chrono::seconds((uint64_t)(time_value*86400));// 86400 seconds in a day
}elseif(match.str(2)=="h"){
returnstd::chrono::seconds((uint64_t)(time_value*3600));// 3600 seconds in an hour
conststd::stringDefaultUpdateMsg="NEW VERSION IS OUT! Please update to the new version ({}) of the BeamMP-Server! Download it here: https://beammp.com/! For a guide on how to update, visit: https://docs.beammp.com/server/server-maintenance/#updating-the-server";
// Copyright (C) 2024 BeamMP Ltd., BeamMP team and contributors.
//
// BeamMP Ltd. can be contacted by electronic mail via contact@beammp.com.
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published
// by the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
#include"Env.h"
#include<optional>
std::optional<std::string>Env::Get(Env::Keykey){
autoStrKey=ToString(key);
autoValue=std::getenv(StrKey.data());
if(!Value||std::string_view(Value).empty()){
returnstd::nullopt;
}
returnValue;
}
std::string_viewEnv::ToString(Env::Keykey){
switch(key){
caseKey::PROVIDER_UPDATE_MESSAGE:
return"BEAMMP_PROVIDER_UPDATE_MESSAGE";
break;
caseKey::PROVIDER_DISABLE_CONFIG:
return"BEAMMP_PROVIDER_DISABLE_CONFIG";
break;
caseKey::PROVIDER_DISABLE_MP_SET:
return"BEAMMP_PROVIDER_DISABLE_MP_SET";
break;
caseKey::PROVIDER_PORT_ENV:
return"BEAMMP_PROVIDER_PORT_ENV";
break;
caseKey::PROVIDER_IP_ENV:
return"BEAMMP_PROVIDER_IP_ENV";
break;
}
return"";
}
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Tixx
SaltySnail
SaltySnail
alex
TechStreet
0R3Z
boubouleuh
wadyankaw
Kipstz
pedrotski
Mathijs
Lion
0R3Z
O1LER
Deer McDurr
Lucca Jiménez Könings
sla-ppy
redracer
Maximilian Rehms
Maximilian Rehms
Neptnium
alex.ita
Lambdax
Bennett
Starystars67
Luuk van Oijen
rgnter
Simon
rgnter
Qest333
ㄗㄠˋ ㄑㄧˊ
20dka
Mackenzie
Rouven Himmelstein
awesome_milou
Lucca Jiménez Könings
Lucca Jiménez Könings
Anonymous-275
Anonymous-275
root
jojos38