17,530 Commits over 1,614 Days - 0.45cph!
deer has a rough ragdoll and test left and right run anims.
Added MaxStatConsideration which takes an array of stats and returns the max value as score
Stat implements IComparable
Added a Description field to Condition and Consideration Settings
Renamed MaxStatConsideration to GroupedConsideration, now takes a list of Considerations and returns the highest score, should/could probably be handled throgh some grouping system
-You should finally be able to create Conditions with a template, like you do with classes. It will create $Name$Condition.cs and $Name$ConditionSettings.cs
-And more templates for Action and Consideration
/Nature/Grass folder stuff.
Added Decision Log debugger window, allows for stepping through DM ticks for the selected Unit's Agent
Fixed dict lookup error, improved consideration logging
Fixed some issues with DecisionLog
Deleted some old un-used actions with missing types
DecisionLog twaeks
- Return null if best decision has a score of 0.
- If we traverse the goal hierarchy, but failed to find a result down one branch, look down another sibling branch, or go up one level and continue looking there. Just make sure we return the new branch result, like we were NOT doing!!!
SmartObject's CanSubscribe, when there is a required interaction id in place, we should return whether the incoming interaction id is the one required or not.
Removed unsubscribe hack in SmartObject.
Added a proper unsubscribe when a BehaviourChain Completes (need to verify that it's enough to just unsubscribe the current behaviour plan).
Was storing the current SmartSubscription on the InteractionPlan, which is owned by the SmartObject... so this is kind of wrong
Moved caching of SmartSubscription over to Agent, and exposed that in BehaviourChain, that we can properly manage unsubscribing from a SmartObject when a BehaviourChain is ended abruptly.
-conditions and considerations made system serializable in templates, else you could not assign them to a goal.
Agent.DecisionLogs uses FixedArray
Metas, apparently
deleting this so I dont re grab it when updating workspace
Added EntityConditionFilterDrawer
Added a summary in BuildingStates to un-confuse the confused
Units now get the MaxSpeed stat base value from SpeciesSettings on creation
Added auto grass spawn refresh toggle to world debug tab
Some stat/SO serialization tinkering
-AlwaysWin consideration
-CanFindUbuiltBuilding Condition
-yam
-Fixed Self-Physio goal always winning
Agent.DecisionLog clears when it reaches capcity (1000)
improved deer
added basic material colours to the bear
More \nature\grass test files.
Fixed CreateActivityActionSettingsEditor not inheriting from base
Added a lot of animation param actions back into Interaction ActionChains
Made human gathering use 3 states in mechanim (in/loop/out) cc frankieh
Fixed fruit bush missing collider/lod
Renaming some Actions
Fixed DecisionLog resetting when editor is paused
We weren't rejecting on comparison with traversal memory quite the right way. Fixed now.
GoalSettingsEditor will force DecisionPlan.ScoreForDecisionMaking to false if the Goal has no plans or children in data
Optimized the check for whether we've traversed the entire decision space, so that we don't check unless our traversal memory is at least as big as the decision space.
Added a big more Decision Debug Log output for agents, that we can clearly read where we find Best Goal, Best Goal Plan and Best Goal Plan Variant.
EntityConditionSettings.DecisionLayer == Any
Added StoreTarget Action
EntityCondition/BuildingFilterSet improvement
Made DebugBestGoalPlanVariant not derp out as easily when target are not a unit.
Lots of data changes
Added MachineCanProcessItem Condition
Re-wrote how target filtering works in UnitDecisionMaker (cc ptrefall please take a look)
Asset save
Added Wander Ability
Aded Goal : Idle under Goal : Self - Social
Cliffside topo fix
Added AbilityManager
Moved TOD xmls, data save
Idle ability, Wander ability
Agents now have access to all abilties from AbilityManager, no longer stored in AgentSettings
Added a lowest score threshold to DSE, hardcoded to 100'000 for now, to prevent derp decisions of really low score.
