VladimirMangos

No, just need check how 1) talent points updates at quest loading 2) possible check in reset also need updated. [added] I not see problem... I will test.

Possible, but need way proper reset creature state in like case or way store and restore state. Ofc, this can be possible and just not implemented. But wrong show it as easy loop by creatures

Wast3d, your post only meaning: 1) you not understand that any relaese (in fact next after it clients switch) _always will break many many things. This not dependent from mangos team or any DB project team. Don;t like this problems not switch to new client support revision and has been happy with old mangos revision. 2) about what fix next: you not have rights point to devs or anyone what need fixed next. Devs fix/implement what they think more intresting for they in current moment. You can only fix some problem by self and suggest fix community. Aftre this fix or will be added to mangos sources (if it correctly implemented) when some devs find time review it or will not added (in other case) but community still will have possibility use it until problem will not solved in proper way. 3) about " i think all DB devs are pretty angry". Please avoid talk not from self position. I think DB devs can say own points by self. More nice part: Good show that you don't know context and have wrong understand state in development. This changes well researched and suggested to mangos by _DB_dev_ as proper way fix problem with wrong creature damage calulation.

Exist already comamnd sets .account / .character. Why not add subcommands in its instead ?

About emu farms... You can name mangos how you wish but if you will name it as emulator at forum yopu will banned at end. Closed.

it work fine. You need learn special skill for for this.

lol Closed.

Backported to 0.12 (it expected from 1.x.x versions)

I currerntly prepare pach to resolve it mixied use. (added: in [8077]) In short (thanks for tom for clarify this): MOVEMENTFLAG_ use in many packets, most visible in player move packets. Mostly player related. MONSTER_MOVE_* use _only_ in SMSG_MONSTER_MOVE packet, this is (moster moved flags) expected to be in all DB move_flag fields. Mostly creature related (and playes in case when player controled as creature in some cases)

Rewrited patch in [8076]. Base at most similarity to original thenecromancer's patch i make he main author of commit. But thanks all who take part in problem resolving

I test your patch and re-found why stop move added in this code. It prevent cheating with stay at griffon when at client path finished but server not stop taxi movment at own side. If this call not present then taxi user can fly to any point and stop taxi in this point instead expected point. As you already note, main problem with moster move call in StopMoving(), but it infact not have any real meaning in functionality. It just "nice" return player to point.

Hmm, but in suggested form it switch to run if start flight with walk mode... [added]nos4r2zod, With big chance I will appply your patch, but before need resolve really BIG probelm with use Unit::m_unit_movement_flags for store 2 totally different things: MovementFlags and MonsterMovmenetFlags, and restore fill this or new field by player movement flags received from packets. I reseatch ways of this problem fix currently. Not easy and affect DB/scripts...

I think only code...

It has been reverted only in 0.12 case where it wrong. In master it wrong 50/50 with old way. One cheating with it not possible, but possible another. At least with patch more cases work as expected.

in 0.12 different code that send aura already by another way.

In [8074]. Thank you

I test 0.12 branch and can't repeat crash...

item/corpse data field is another problem. And i agrree that some data in it also better recalculate at loading instead just use and maybe replace by normal field in same way as start do in this patch...

at just enter and leave?

One from main reasons (in final state): this will illiminate need to create character `data` convertion at client switches. Instead normal fields add/deleted will be used.

Time to close.

if writed [NNNN] then OFC it in master, _only_ master comkits have revisions in form [NNNN]

In [8068] with small cleanups. Thank you

Not help in fact: patch not follow mangos code style anyway: tabs, trailing white spaces...

In [8066]. Thank you