* re-add x and z checking for height
Now that we have better navmeshes, this will result in much better results and as such we can re-enable this check.
* Always run navmesh extraction
waste of time most of the time, but no other way to force update to the meshes easily.
* Navmeshes Version 2
- Add all missing zones
- Drastically improve several zones and their navmeshes, cleaning them up, making them more accurate and generally using more features of detour/recast.
* Update CMakeLists.txt
* update meshes
* Navmesh: Add pet cove navmesh
* Navmesh: Fix navmesh for fv
Fixes a possible nullptr access. This is the only call to GetPlayerData where we do not check the result for some reason, so this PR adds in the check and a resulting log line.
Code compiles, unsure how to reproduce the issue, however here is the crash dump I used to deduce this being the possible issue
```
Error: signal 11:
[00] CatchUnhandled(int)(+0x316) [0x561469100336]
[01] /lib/x86_64-linux-gnu/libc.so.6(+0x42520) [0x7f65e8e45520]
[02] /DarkflameServer/build/ChatServer(+0x32719) [0x5614690fa719]
[03] HandlePacket(Packet*)(+0x2a0) [0x5614690fcfb0]
[04] /DarkflameServer/build/ChatServer(main+0x92e) [0x5614690fb75e]
[05] /lib/x86_64-linux-gnu/libc.so.6(+0x29d90) [0x7f65e8e2cd90]
[06] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0x80) [0x7f65e8e2ce40]
[07] /DarkflameServer/build/ChatServer(_start+0x25) [0x5614690fc375]
```
* Components: Make ComponentType inline
Prevents the next commits ODR violation
* Components: Add new components
* Entity: Add headers
inline script component ComponentType
* Components: Flip constructor argument order
Entity comes first always
* Entity: Add generic AddComponent
Allows for much easier adding of components and is error proof by not allowing the user to add more than 1 of a specific component type to an Entity.
* Entity: Migrate all component constructors
Move all to the new variadic templates AddComponent function to reduce clutter and ways the component map is modified.
The new function makes no assumptions. Component is assumed to not exist and is checked for with operator[]. This will construct a null component which will then be newed if the component didnt exist, or it will just get the current component if it does already exist. No new component will be allocated or constructed if the component already exists and the already existing pointer is returned instead.
* Entity: Add placement new
For the case where the component may already exist, use a placement new to construct the component again, it would be constructed again, but would not need to go through the allocator.
* Entity: Add comments on likely new code
* Tests: Fix tests
* Update Entity.cpp
* Update SGCannon.cpp
* Entity: call destructor when re-constructing
* Update Entity.cpp
Update Entity.cpp
---------
Co-authored-by: Aaron Kimbrell <aronwk.aaron@gmail.com>
unsure how to reproduce the actual bug, however we can see that with the following crash dump
```
Entity::GetComponent(eReplicaComponentType) const(+0x4) [0x56095665e634]
BossSpiderQueenEnemyServer::OnDie(Entity*, Entity*)(+0x28d) [0x560956795d0d]
Entity::Kill(Entity*)(+0xf8) [0x5609566637a8]
ZoneAgProperty::BaseTimerDone(Entity*, std::string const&)(+0x89b) [0x56095683736b]
Entity::Update(float)(+0x2b6) [0x560956662676]
EntityManager::UpdateEntities(float)(+0x2e) [0x56095667305e]
```
that the actual crash issue starts at
```
Entity::Kill(Entity*)(+0xf8) [0x5609566637a8]
ZoneAgProperty::BaseTimerDone(Entity*, std::string const&)
```
BaseTimerDone calls Kill, and there is only 1 call to Kill in the function which calls Kill no arguments, meaning the killer is a nullptr. This propogates its way to the BossSpiderQueenEnemyServer::OnDie wherein we blindly check the killer pointer without verifying that the pointer is actually valid.
This patch simply checks that killer is valid before access to address the hole.
* Logger: Rename logger to Logger from dLogger
* Logger: Add compile time filename
Fix include issues
Add writers
Add macros
Add macro to force compilation
* Logger: Replace calls with macros
Allows for filename and line number to be logged
* Logger: Add comments
and remove extra define
Logger: Replace with unique_ptr
also flush console at exit. regular file writer should be flushed on file close.
Logger: Remove constexpr on variable
* Logger: Simplify code
* Update Logger.cpp
Adds a missing edge case on the client where you may have the is_smashable flag set, but dont actually have a destroyable component. In this case you should create only a destroyable component and add it to the component and serialize it last after all other data. We also needed to add the set_faction var reading and setting as this is how Oliver Sudden actually gets their faction since they have no actual information in the database, it is all stored in ldf keys.
Tested that Oliver Sudden no longer logs Unable to unserialize logs when serialized or constructed.
Fixes#1222
addresses an issue where the death behavior of a destructible component was not being respected and enemies with destroyable components that had special death animations were not able to play the animation on death. This pr adds in the hardcoded constant the client uses for the same metric of 12 seconds.
Tested that claiming Nimbus Rock and completing the property guards mission allows him to vacuum away and then network the destruction packet 12 seconds later.
Fix gravity not being adjusted when colliding with a phantom physics object
Tested that while on moonbase, the players gravity is no longer reset to 1 when they change their cheat info.
Fix a race condition that would overwrite the port of zone 0 with the auth server port or a chat server port which would cause a domino effect of desynced server ports
Tested that a sleep(5) in auth still allows a player to connect and login without issues
* Move away from constructor queries
Fix up other large tables to have proper backup lookups
Revert "idk im just dumb ig"
This reverts commit 5d5be5df53b8959b42b291613d7db749a65a3585.
idk im just dumb ig
* Fix slow components registry lookup
* add define for cdclient cache all
* Huge loot namespace rework
- Remove all excess memory usage
- do not cache components registry
- cache loot matrices on startup of the destroyable component
- convert loot singleton class to a namespace
- rework loot cdclient tables to operate closer to how someone would actually use them (basically doing the previous LootGenerator::LootGenerator caching but in those tables)
- Memory usage reduced by 10%+ across the board
* cache rebuild matrix
* Database: move reading to own function
Also change name of cache to PascalCase
* Database: Move common function rading
* Make serialize actually virtual
yep
* Abstract to PhysicsComponent
Move shared functionality of all physics related classes to a base class.
Tested that there were no failed to unserialize errors when in main gameplay in Gnarled Forest or in a race.
Tested that 2 players were able to see each other in the above scenarios just fine as well.
* Update PhantomPhysicsComponent.cpp
* Add SimplePhysicsTest
* Add construction test
* Update SimplePhysicsComponentTests.cpp
* remove flags and fix override
* Update VendorComponent.h
* Re-write AOE behavior for new filter targets
Update Tacarc to use new filter targets
Added dev commands for skill and attack debugging
* Get all entities by detroyable
rather than controllable physics
Since destroyables are what can be hit
* Re-work filter targets to be 100% live accurate
reduce memory usage by only using one vector and removing invalid entries
get entities in the proximity rather than all entities with des comps in the instance, as was done in live
* remove debuging longs and remove oopsie
* address feedback
* make log more useful
* make filter more flat
* Add some more checks to filter targets
add pvp checks to isenemy
* fix typing
* Add filter target to TacArc and update filter target
* fix double declaration
* Some debugging logs
* Update TacArc reading
* make log clearer
* logs
* Update TacArcBehavior.cpp
* banana
* fix max targets
* remove extreanous parenthesesuuesdsds
* make behavior slot use a real type
---------
Co-authored-by: David Markowitz <EmosewaMC@gmail.com>
This reverts commit d7e16ab589697fd1a0a270a02c622b8fa752638f.
Fixes an issue where the incorrect mission is marked as being offerable and this overwrote the original mission offered by wisp lee and forced him to offer a different, unacceptable mission.
Tested that completing the mission twice from a new daily state and a repeatable state both completed the chain correctly. Unsure what the original bug was, but it does not appear to be present.
On top of that, there is no pre-requisite for mission 1883 anywhere in the cdclient, so at best, this check was always false, but the correct behavior is exact equivalence for the mission state.
* Fixed Scrapped Racecar Stuck Issue
Changed RacingControlComponent to boot players back to the hub world when trying to race after dismantling a vehicle.
* Modified OnPlayerLoaded to fix Vehicle Inventory issue
* Change 3
---------
Co-authored-by: David Markowitz <EmosewaMC@gmail.com>
Currently on line 116 of AssetManager.cpp, we have the following
`*data = (char*)malloc(*len);`
however on line 45 of AssetManager.h we have `delete m_Base;` which is not the same allocator as we used to allocate the memory.
This PR matches the malloc and free to be the correct calls.
