Fix a non-working player check #725
Conversation
game/world/worldobjects.cpp
Outdated
| @@ -322,6 +322,7 @@ Npc* WorldObjects::addNpc(size_t npcInstance, const Vec3& pos) { | |||
| Npc* npc = new Npc(owner,npcInstance,pstr); | |||
| npc->setPosition (pos.x,pos.y,pos.z); | |||
| npc->updateTransform(); | |||
| npc->autoEquipWeapons(); | |||
There was a problem hiding this comment.
Another override for WorldObjects::addNpc(size_t npcInstance, std::string_view at) seem to be missing.
There was a problem hiding this comment.
I left it out because this function calls restPositionToTA where it's already done. You want to have it added for consistency with other addNpc function?
Additionally I found thatB_RefreshAtInsert is called as well and added it.
There was a problem hiding this comment.
You want to have it added for consistency with other addNpc function?
Ideally, I think, we need to make it robust and leave no room for inconsistency. After giving more thought to it: maybe it's even better to add ProcessPolicy as a parameter to npc constructor. This will make any explicit/implicit use of isPlayer correct.
thokkat
force-pushed
the
fix-weaponAutoEquip
branch
from
7cb8c7f to
2387ef2
Compare
fixes a non-working isPlayer check
thokkat
force-pushed
the
fix-weaponAutoEquip
branch
from
2387ef2 to
732807a
Compare
|
Merged, thanks! |
As described in #722 we don't know if
aiPolicy==Playerin npc constructor which means best weapon equip has to be done later.One of the
addNpcfunctions already does this indirectly by callingrestPositionToTA. The other one is never called for the player so weapon equip can be done here safely as well. This covers the case when marvin insert command is used which was the original motivation to do this in the constructor, see #653 (comment).Regarding the mentioned summons, I modified the wolf spell to summon the hero instead. Weapon was equipped while script had only a
createInvItemline.Fixes #722