Fix for issue #1768 where the bot master was not getting reset.
I also cleaned up leave group action to focus up function scope.
I moved the resets from #612 to the actual leaving function.
Disclosure: LLMs were NOT used in authoring this PR.
Test cases to consider for testers.
1. Master disbands group with random bots. Bots should go about their
business.
2. If you can dual box, create a raid where two real player both invite
random bots. One player leaves group, their bots should also leave.
(edge case, if a random bot that is supposed to leave the group becomes
leader they may disband the whole group.
Fixes a thread safety issue where multiple bots dying in battlegrounds
simultaneously would corrupt the shared static unordered_map, causing
segmentation faults.
Changes:
- Remove: static m_botReleaseTimes map from AutoReleaseSpiritAction
- Add: bgReleaseAttemptTime member to PlayerbotAI (per-bot storage)
- Update: All references to use per-bot storage instead of static map
Why this fixes the crash:
- Each PlayerbotAI instance is accessed by only one map update thread
- No cross-thread access to shared data structures
- No mutex/locking required - thread-safe by design
- Automatic cleanup when bot is destroyed
Thread-safe solution: Per-bot state eliminates race conditions without
performance overhead.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Revert "[Large server fix] #1537 Serialize playerBots/botLoading with a mutex and use snapshot-based loops to fix concurrency crashes (#1540)"
This reverts commit 3fff58df1a.
* Revert "[Fix] teleport to invalid map or invalid coordinates (x , y , z 200000, o ) given when teleporting player (g UI d full type player low , name , map , x , y , z , o ) (#1538)"
This reverts commit ca2e2ef0db.
* Revert "Fix: prevent MoveSplineInitArgs::Validate velocity asserts (velocity > 0.01f) for bots, pets, and charmed units (#1534)"
This reverts commit 4e3ac609bd.
* Revert "[Fix issue #1527] : startup crash in tank target selection — add TOCTOU & null-safety guards (#1532)"
This reverts commit c6b0424c29.
* Revert "[Fix issue #1528] Close small window where the “in a BG/arena” state can change between the check (InBattleground() / InArena()) and grabbing the pointer (GetBattleground()), which leads to a null dereference. (#1530)"
This reverts commit 2e0a161623.
* Revert "Harden playerbot logout & packet dispatch; add null-safety in chat hooks and RPG checks (#1529)"
This reverts commit e4ea8e2694.
* Revert "Dont wait to travel when in combat. (#1524)"
This reverts commit ddfa919154.
* Revert "nullptr fix (#1523)"
This reverts commit 380312ffd2.
* Revert "Playerbots/LFG: fix false not eligible & dungeon 0/type 0, add clear diagnostics (#1521)"
This reverts commit 872e417613.
* Revert "nullptr exception (#1520)"
This reverts commit 3d28a81508.
* Revert "Removed bot freezing at startup and system message, not relevant anymore (#1519)"
This reverts commit bcd6f5bc06.
Bots will now add level- and spec-appropriate oils and stones when maintaining and, with respect to randombots, leveling. All bots (other than those with class-specific temporary weapon enchants) will apply oils and stones to their weapons. General clean-ups to associated code were made.
* feat(playerbots): staggered taxi take-off for bots
Adds four new configurable settings to playerbots.conf:
- AiPlayerbot.BotTaxiDelayMinMs: Min random delay before the 1st follower bot clicks the flight-master
- AiPlayerbot.BotTaxiDelayMaxMs: Upper bound for the overall taxi delay window – larger spreads big raids
- AiPlayerbot.BotTaxiGapMs: Fixed gap added per group-slot so bots never take off together
- AiPlayerbot.BotTaxiGapJitterMs: Extra small randomness added to each gap so launches don’t look robotic
These options allow server owners to fine-tune how bots queue up and take off from flight masters, making their behavior appear more natural.
Closes#1017 : Bots use Flight master nearly the same time.
* fixed build errors
Was missing a header and variable declarations.
* New rpg startup speed up and refactor
* New rpg do quest
* Fix invalid height in quest poi
* Add quest accept and reward limitation
* New rpg quest improvement
* Organize quest log, reward quests and fix grind target
* Quest dropped statistic and remove redundant code
* Decrease grind relevance lower than loot
* Fix new rpg drop quest
* Go to reward quest instead of innkeeper when quest completed
* Fix incorrect logic in do quest reward
* Fix reset quests in factory
* Fix crash on grind target value
Co-authored-by: SaW <swerkhoven@outlook.com>
* Fix a minor error in DoCompletedQuest
* Let bots get rid of impossible quests faster
* Increase loot fluency (especially for caster)
* Remove seasonal quests from auto accept
* Enhance quest accept condition check
* Add questgiver check (limit acceptation of quest 7946)
* Questgiver check and localization
* Near npc fix
* Fix quest item report
* Add lowPriorityQuest set for quests can not be done
* Improve gameobjects loot
* Do complete quest
* FIx move far to teleport check
* Accept or reward quest from game objects
* Fix possible crash in rpg game objects
* Fix ChooseNpcOrGameObjectToInteract crash
---------
Co-authored-by: SaW <swerkhoven@outlook.com>
Resolves#947
Equip logic was failing as projectiles were never returning ITEM_USAGE_EQUIP in ItemUsageValue.cpp, added two cases where equip is returned:
If no ammo is currently set
If new ammo has higher DPS than old/currently equipped ammo
While testing this using "b [itemlink]" and "b vendor" to purchase arrows I noticed some issues with BuyAction.cpp and have resolved them:
Bots will now perform the "equip upgrades" action for any bought item that has an equip usage
When using "b vendor" to buy all useful items from vendors within interaction distance, it now sorts the list of available items by calculated item score and buys the highest scoring item (if it is higher than the currently equipped item) for each slot. It should not buy multiple items for the same slot anymore, saving gold/emblems/etc.
"b vendor" will now only attempt to buy 1 of each item. Consumable and projectile item types can be bought up to 10 times per execution as long as it is still useful to buy the item in each iteration of the for loop. All items were following this behaviour previously and since the equip command was only given after the for loop it would buy 10 of an item before triggering it wasn't useful to buy more.
And finally, resolved issues where a bot runs out of ammo mid-fight:
Re-enabled combat and non-combat "no ammo" strategies to perform "equip upgrades" action.
Modified GenericTriggers.cpp; AmmoCountTrigger::IsActive to return true when the bot has ammo but it is not equipped yet.
* Update PlayerbotAI.h
* Refactored a number of functions in PlayerbotAI.cpp
* Update PlayerbotAI.cpp
* Update PlayerbotAI.cpp - update for commit done
Take
568592f188
into account.
* Missing check for aurEff
* Update PlayerbotAI.cpp
nvm...
* Update PlayerbotAI.cpp
GetAura
* Update PlayerbotAI.cpp
Simplified/Optimized sPlayerbotAIConfig->dynamicReactDelay logic for in-combat.
* Update PlayerbotAI.cpp
Dubass fix
* Update PlayerbotAI.cpp
Fix bots leaving dungeon group,. again.
* Update PlayerbotAI.cpp
* Update PlayerbotAI.cpp - order correction
...Required for proper pet behavior.
* Update PlayerbotAI.cpp - UpdateAIGroupMembership()
Final refactor of helper function as all now works as required.
* Update PlayerbotAI.cpp
FindItemInInventory
* Update PlayerbotAI.h
Added helper functions, correct public -> private
