Leaving players will have PEER_ID_INVALID for database saving purposes.
This patch clarifies that, and properly protects the batch send function.
Bug introduced by 5f5ea132.
Reduces memory usage on the server, especially with many user and/or large viewing distances.
Currently disabled on the client due to known data races on a block's data.
Large structures which are generated in on_generated callbacks
independently by Lua cannot influence decoration placement. This
change enables such a callback to assume responsibility for generating
decorations itself, presumably after structures are placed, by
disabling decorations in mg_flags and executing
core.generate_decorations.
---------
Co-authored-by: Po Lu <luangruo@yahoo.com>
These changes were initially made to improve performance. However,
on modern hardware, these changes turned out to make no difference.
This commit unifies the calculations in 'draw' and 'getDimension' and
adds comments to make it more understandable.
A few functions tried to dereference a ServerEnvironment nullptr by
calling 'getEnv()'. This change makes use of a macro where possible.
This also cleans up incorrect macro uses, with no functional difference.
This change prevents issues arising from partial generation of MapChunks, which are liable to be regenerated completely when ungenerated MapBlocks within are encountered.
Co-authored-by: Po Lu <luangruo@yahoo.com>
Co-authored-by: sfan5 <sfan5@live.de>
Gimbal lock is a situation where the pitch (the middle angle) of the Tait-Bryan angles (usually called Euler angles incorrectly) is 90 degrees. If the angles specify a rotation close to gimbal lock, the precision requirements increase significantly, beyond what a single-precision float can provide, and at exactly gimbal lock, there's a loss of information. The test didn't take this into account. Fix this by decreasing the expected precision when close to gimbal lock.
The increased error rate on ARM Macs is probably caused by lesser precision in trigonometric functions. IEC-559 does not specify any semantics for those, and while Intel typically has a precision < 1 ulp for trigonometric functions with angles < 2*pi, it's likely that ARM's precision is a bit worse.
