* This is also hack because some float with unknown exactly value used as 'size'
* Another not resolved problem: go size used in creature target autoselection, so for large 'virtual' go
creature can seelct strange points near in attempt select 'free' place near target.
* At every change to model or scale, the related data (bounding_radius/combat_reach) is now updated accordingly (note that player combat_reach are not changed like creature).
* UpdateModelData is called from within SetDisplayId while changes to scale has explicit call to UpdateModelData after new scale is set (mostly for aura scale)
* The updated values are calculated by (scale*bounding_radius)/(scale*combat_reach)
* Database values for bounding_radius/combat_reach are expected to be relative to scale like 1.0
Signed-off-by: NoFantasy <nofantasy@nf.no>
* Remove reduncent chekc and args for MapInstanced::CreateInstance(const uint32 mapId, Player * player)
* Rename for avoid name duplication raw InstanceMap creating to MapInstanced::CreateInstanceMap
* Move lookup code of player/group bounde instance save to Player::GetBoundInstanceSaveForSelfOrGroup
Data are stored in same table as for creature (like dbc models data), and provides the default values for playable races.
Currently the bounding and reach are applied only at creation and load. Note that these values are modified by scale. For player case, scale is always 1.0 as default.
For later, auras and spell effects that change scale and/or modelid must in addition make sure bounding_radius are updated accordingly to the new scale and/or model (combat_reach does not seem to be affected by such changes, and is always 1.5 for players).
Signed-off-by: NoFantasy <nofantasy@nf.no>
* Use single function for erase element from save list.
* Make lock bool checks class local (it need be replaced by normal lock object in future)
* Remove redundent now friend class declarations.
also isActiveObject check moved to WorldObject - all worldobjects are inactive by default, excluding players
this is a final part of cameras implementation
(based on commit efc9623)
Signed-off-by: VladimirMangos <vladimir@getmangos.com>
Also remove template-way code selection for object for Map::AddToActive/Map::RemoveFromActive.
This simplify function use from not Map code.
* Also fix triggered by related checks unread packet tail spam cases.
* Fix one case when expected death at fall controlled player
under ground has been possible avoided
* This allow prevent move home/to caster at apply
* Also replace AI and reset movegens at control time
* Also stop combat and clear threat/hostile list at lost control for prevent combat with freindly units.
Also do come constatification of quest status check functions
and fix one unsafe use std::map operator[].
Never never use std::map operator [] for _read_ access.
Good chance corrupt related structure state ;)
Not in found case but in general.
npc_vendor.ExtandedCost can be negative now that meaning:
price excluded default item BuyPrice and use only abs(ExtandedCost) items).
For example expected used for item 36908.
Only use TalkedToCreature when gossip does not have any following sub menu (and of course when the prepared npc options does not include any gossip)
Signed-off-by: NoFantasy <nofantasy@nf.no>
* Show skinning tooltip only after creature loot
(when creature can be explcitly skinned).
* Allow reopen skinning loot if still have not looted items.
* If player tap creature in group and leave then group will have access to creature loot if not disbanded
* If player tap creature and after join to group then creature loot will accesable only by player
* Also RewardPlayerAndGroupAtKill divided to simgle player and group reward versions used for group tap
and single player tap cases.
* Changes include (in fact as part cleanup) fix got NeedBeforeGreed locked item after expire roll timer
suggested by somedruid@mangos.lighthouseapp.com.
* Added cancel rolling at creatuer corpse remove.
* For NeedBeforeGreed and GroupLoot sued common code for prepare item roll.
Note: it not applied to crashes and shutdowns after mangosd/relamd startup completed.
You can continue without delay (as now) and fast exist, or wait <Enter>, or some secs
for continue. Last can be helpful for see for example not applied sql update revision
data and etc.
