Live sent client_death=false on every Die, players included (19,428 in
the live captures, 452 of them player deaths); DLU sent true for
players. The client only uses the flag to match a death it already
predicted locally (LWODestroyableComponent::msgDie): with it set, a
Die that arrives while a post-death respawn is pending is dropped.
Check: die to an enemy, to falling and to a quickbuild trap; the death
animation, coin loss and respawn prompt all still appear, and other
players see you die.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Nothing in the server writes or reads a packet by hand any more, so the
helpers for doing so go:
- CBITSTREAM, CMSGHEADER, CINSTREAM, CINSTREAM_SKIP_HEADER, SEND_PACKET,
SEND_PACKET_BROADCAST and HEADER_SIZE leave dCommonVars.h;
- the free BitStreamUtils::WriteHeader (LUBitStream::WriteHeader writes the
same bytes) and the unused PacketUtils::SavePacket are deleted.
The last raw reads are replaced: WorldServer builds its input stream
directly, the master packet logs read the header with
LUBitStream::ReadHeader instead of peeking at packet->data[1] and [3], and
MessageInspector reads a sent game message's header with the new
NetGameMsg::ReadPacketHeader (the counterpart of WritePacket) instead of
memcmp/memcpy. packet->data[0] is still compared with RakNet's own
connection IDs.
The frozen oracles keep using the macros verbatim through the test-only
tests/dGameTests/LegacyPacketMacros.h; the HeaderSkip tests, which only
tested CINSTREAM_SKIP_HEADER, are removed.
docs/PacketArchitecture.md: "where we are" now describes the final state
and what still touches raw bytes (RakNet IDs, replica headers, behavior bit
streams), and a new section collects the known wire discrepancies found
during the conversion, with client addresses.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Added cooldown handling
* Made most of the logs hidden outside of debug mode
* removed weird submodule
* kill this phantom submodule
* updated to reflect reviewed feedback
* Added IsCooldownImmune() method to DestroyableComponent
* friggin typo
* Implemented non-pending changes and added cooldown immunity functions to DestroyableComponentTests
* add trailing linebreak
* another typo :(
* flipped cooldown test order (not leaving immune)
* Clean up comment and add DestroyableComponent test
* 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>
* Breakout rest of the enums from dcommonvars
so we don't have to deal with merge conflicts
ePlayerFlags is not a scoped enum, yet, due to it's complexity
* address feedback
* make player flag types consistent
* fix typo
* breakout the component types into a scoped enum
tested that things are the same as they were before
* fix missed rename
* fix brick-by-brick name to be crafting
because that's what it is
Implement GTest as a testing infrastructure.
Make windows output binaries to the build folder instead of the release type folder (potentially issue further down the line)
Add a simple unit test for DestroyableComponent
