fixes#5810
probably fixes#4986#5810 was caused by the workaround for #4394, which broke in 1.20 for reasons I'm still unclear on.
As FLAG_TELEPORT does not work at all for non-player entities, and causes bugs with player entities, sending the teleport movement without the flag is the least buggy way to solve all of these issues. Having the client interpolate teleport movements is not ideal, but there doesn't seem to be a way to reliably prevent it without causing even more bugs, so this will have to do.
due to the extremely large performance cost of instantiating AsyncTasks, it's usually not worth bothering with async compression except for very large packets.
While this large overhead can be significantly reduced by using specialized threads, it's early days in the testing stages for such improvements, and for now, we still have this to deal with.
Since async compression is always used prior to player spawn, this change may slightly improve the performance of the pre-join stage of the game.
we rely on phpstan for validation of this internally, and plugins shouldn't be calling these methods anyway.
this significantly reduces the overhead of CompressBatchPromise.
moral of the story: do not trust that mojang things do what they say they do - the spectator ability layer always applies, regardless of whether the player is actually in spectator mode or not ...
When timings was disabled, internalStopTiming is not called, and timer depth is not decremented.
If timings is later reenabled, the next call to internalStartTiming will think the timer is already running, and won't generate any new records for the timer.
This has led to broken timings reports with missing Full Server Tick entries, amongst other things.
fixes#5722
I'm not very clear why this works. PM doesn't use real spectator mode yet (we're still using the faux spectator mode PM has had for years, because I haven't yet assessed how real spectator mode will affect stuff like block interactions), so this ability layer shouldn't have any effect.
thank you @Alemiz112
due to direct repeated usage of registerEvent() with closures, we've seen some libraries like muqsit/SimplePacketHandler generate very large timings reports, because a new timings handler gets created every time a plugin registers or unregisters a new packet handler callback.
This change fixes the problem by ensuring that any handlers derived from the same function, handling the same event class, will share the same timer.
this saves a considerable amount of memory.
we don't actually need this state array in PM4 anyway, since we don't support the client-side chunk cache yet.
when the time comes to support it, it'll be much more practical to cache binary states and copy bytes anyway, instead of doing it the current way, which is both slow and memory-intensive.
Measured footprint change: 9 MB -> 400 KB.
this allows saving about 4 MB of memory, because there are many blocks which have identical states, although they have different IDs.
this relies on a potentially risky assumption that the tags in knownStates won't be modified. If they are modified, the changes will influence all blockstates which share the tag.
However, I don't expect this to happen, and the 4 MB memory saving is substantial enough to be worth the risk.
closes#5724
this uses a (potentially bogus) assumption that the lowest mapped meta value associated with an ID is valid. I don't want to break this during a patch release, and this works for now.
In the future it would probably make more sense to bypass ItemTranslator entirely, and rely solely on the blockstate returned by RuntimeBlockMapping to fetch the correct ID. This is similar to how we serialize items for saving on disk in PM5.
this eliminates the performance issues described by #5719.
closes#5719
we may want to consider exposing a public API for this in the future, since it might be useful for plugins.
we can safely assume that:
- the inventory's max stack size won't change during the operation
- two items which stack together have the same max stack size
- the item's max stack size won't change during the operation
Since light population is required to make a chunk tickable, a chunk may not be tickable for some time if the async workers get backlogged.
The previous version of this system only cached the eligibility result if the result was a "yes", but we can also track it when it's a "no", rather than rechecking it every tick.
This change should improve performance in factions and similar gamemodes, which involve large maps with sparsely distributed players, where each player likely has an independent, non-overlapping ticking chunk circle.
We also ditch TickingChunkEntry in favour of multiple arrays to track the eligibility states. This allows us to avoid rechecking the (even cached) readiness of potentially thousands of chunks. If there are no ticking chunks to recheck, this reduces the cost of the selection process to zero.
