Perhaps you should give specific examples of what isn't working and why you think it needs changed or fixed rather than making a generic post? This post is basically useless as it is written (I assume with AI).
since you asked
Specific examples:
This is not a generic complaint, and it is not a request for fewer features. These are major V3.0 problems that server administrators have had to diagnose, patch, work around, or build recovery systems for:
1. Turret death-state desync
Powered turrets can kill a player on the server while the client remains lying on the ground without receiving the death or respawn screen. We found that offline turret ownership makes the problem more likely. Fixing this required tracing the damage network packages, identifying stale owner data, writing a server-side watchdog, sanitizing the recovery packet, adding a second recovery stage, and performing repeated live tests.
2. Stuck use/action state
Players can become trapped in a broken use or countdown state. We had to add a /fixuse recovery command instead of having a reliable vanilla recovery path.
3. Stale lock states
Locks can remain in an invalid or stale state and require a dedicated server-side guard.
4. Sign-data problems
Sign data can become invalid or fail to synchronize correctly, requiring a separate sign-data repair patch.
5. Vending-machine restocking
Vending reset and restock ticks have required server-side patches so restocking continues to function correctly.
6. Backpack diagnostic spam
EntityBackpack repeatedly generates useless diagnostic log spam. We had to suppress it so actual server problems remain visible in the logs.
7. Trader chunk-entry problems
Entering trader chunks has required a dedicated corrective patch to prevent broken trader-area behavior.
8. Quest logout exploit
Quest state can be abused or mishandled through disconnecting, requiring a separate logout-exploit fix.
9. Drone recovery
Drones can become lost, inaccessible, or stuck in bad saved states, requiring administrator recovery commands and server-side repair tools.
10. Static-water failures
Water behavior has required an emergency corrective mod rather than functioning reliably through the normal game systems.
11. Death and respawn handling
We have had to maintain custom respawn handling and carefully test coexistence with the vanilla death system because death-state synchronization is not consistently reliable.
12. Scoreboard and gamestage updating
We have had to use safe caching and throttling to prevent unnecessary or unsafe scoreboard and gamestage updates.
These are not content additions, balance changes, or players demanding more sandbox toys. They are reliability fixes for basic dedicated-server systems: death, respawning, locks, signs, vending machines, backpacks, traders, quests, drones, water, and player interaction states.
The turret issue alone required decompiling the current server assemblies, tracing the client/server damage flow, writing multiple versions of a Harmony patch, conducting forensic audits, and arranging live testing with players. A player should not have to relog because a normal powered turret killed them, and a server administrator should not have to reverse-engineer the game’s network damage system to fix it.
I understand that experimental versions can contain bugs. That does not mean every bug should be dismissed with “do not use experimental.” Experimental builds still need useful bug reports, reproducible testing, validation, and fixes. Administrators reporting these problems are doing unpaid testing and engineering work that directly improves the game.
My request remains the same: stabilize the systems already present, document which setting has authority when values conflict, add useful validation and logging, and resolve the existing dedicated-server defects before continuously expanding the sandbox system.
