[One] Problem in HomeMovementGenerator.cpp ?

[Oniryck]

Hello,

I don't know if this problem is only for me, but I had many crashes since the modification in this file in recent commit, because all creature don't necessarily have AI. So maybe it's useless but I post my patch to fix this problem in case anybody have the same issue :

my patch:

diff --git a/src/game/HomeMovementGenerator.cpp b/src/game/HomeMovementGenerator.cpp
index b53f2c3..38ea935 100644
--- a/src/game/HomeMovementGenerator.cpp
+++ b/src/game/HomeMovementGenerator.cpp
@@ -99,8 +99,9 @@ void HomeMovementGenerator<Creature>::Finalize(Creature& owner)

        if (owner.GetTemporaryFactionFlags() & TEMPFACTION_RESTORE_REACH_HOME)
            owner.ClearTemporaryFaction();

        owner.LoadCreatureAddon(true);
-        owner.AI()->JustReachedHome();
+        if(owner.AI())
+            owner.AI()->JustReachedHome();
    }
}

[Schmoozerd]

a HomeMMGen must be initialized onto an NPC which has if (owner.hasUnitState(UNIT_STAT_NOT_MOVE))

then i_travel_timer might not be initialized, and hence could be (compiler dependend) 0

which would trigger the call in Finalize for owner.AI()->JustReachedHome

But I don't understand how this can happen (because after a creature is created it is expected that it has AI initialized.)

also I don't see why a HomeMMGen would be pushed in a place where the npc has UNIT_STAT_NOT_MOVE, and Finalized where no AI exists..

[Oniryck]

After some research, it seemed to happen with one of my custom scripted mob that uses "SimpleAI* ai = new SimpleAI(pCreature);" as its script. So it IS a personal problem, sorry for inconvenience . But I didn't have that problem before, so if my patch can help someone that use that kind of script too, I shared it.