Schmoozerd

Db-script system Fresh version: http://paste2.org/p/1963058 (need to add #include "GridNotifiers.h" #include "Cell.h" into ScriptMgr.cpp for non-pch compile, thx to neatelves for testing) This patch expands all commands to be able to use the "buddy" (that is search npc/go by entry in range) concept. Therefore I replaced two datalong columns with entry, radius columns. Reasons: - Code Simplification, because to get source and target for a command could be generalized - For DB-people the format of the script tables will be easier, because all commands get much more similar - Many commands will become more powerfull Features: * Except outdated flag commands, all commands support the buddy concept * Unified and improved a lot of debug and error output * Engine moved to ScriptMgr.cpp, in Map only the code calling is still located Please test! For testing be aware to run the sql files in sql/update which start with 77777_ as revisions Please focus on the convert changes! Note when testing: This will change content and structure of the _scripts tables, hence you should test on a clean world-database, which you can reset easily.

WorldStateUI.dbc has information related to many worldstates and how they would be displayed in the client. there are texts including %NUMBERw where the number is the worldstate that is displayed within this text.

Fixed in 11930. Sorry for not giving further credit, I didn't look for suggested patches

in 11929, thank you and sorry for the delay

hmm, the patch is not entirely complete - in this form lot of junk will remain within mangos sources But deleting is usually rather simple refer to the variable m_mCacheNpcTextIdMap which must be removed along with others Edit: Just talked with crackm from UDB. He wishes the table to last a little longer until we drop it. So my suggestion would be summer this year.

I cannot see why this patch actually works .. The whole Player loading related to alive state is a bit unclear to me atm.. Also I think this very rare case should be handled by Player::LoadFromDB as well (the plr == NULL check is kind of ugly to check for this case, there could be some complicated other cases)

In general you don't need to be aware of bad intented users. Players are usually not lurking around on mangos forums for exploits. I assume there are other places for hackz and cheats and bad stuff, where they might be lurking around, but most likely not here.

which means I have no git utilty here - and doing the changes by hand on github I didn't want to. Will push "soon"

Suggestion: You posted already lots of detailed information: Focus also on how to get them That would be likely 1) Basic information (wowwwiki, wowhead, videos own experience) 2) Available informatioon - SpellWorks 3) Interpretation of spell.dbc data with current mangos 4) Detailed information if available (combat logs and similar logs from real life servers)

You can find many cool stuff inside silverice's github repo.

most likely not related to SD2 freezes related to commands could be related to problems with some db connection issues. If your custom patches did not affect these areas of code, then maybe you have a problem with your mysql server.

There is one rule related to optimation: Analyze/ measure performance before trying to improve performance, and perform big consumers if possible.

well, default behaviour (returning to fly) should likely be handled as part of the aura-unapply event in core. Also I think this core change also requires the additional rooting. EventAI will be required for normal spells and other behaviour - but most likely not related to this not yet complete patch

probably caused by https://github.com/mangos/mangos/commit/de5521c24b000f4546aae6939f0babd0a6c50035 I think this should be good idea to just add a check of "enemy != owner" for such cases.

if his bone minions have fixed spawn positions (not how you coded the patch) you can do both spells in database

you must import all files for mangos db with rev-number greater than 11785 until 11885

Forward: I wrote a PM to UnkleNuke to ask him if he would agree to be mod here. If he will, he should PM luda/MCP for permissions. Also we should likely discuss commit access and permissions for faramir118?

http://getmangos.eu/ <- second line there

A little bump here. I am told - and can agree - that the current uint64 magic numbers are rather hard to read. But I also agree with vladimir that using "custom" enums to hide the actual masks make things very complicated, and also the template interface is not really easy to understand when trying to research. actually I think the version that would be best would look something like HasMask<WARLOCK>(CF_WARLOCK_INFERNO | CF_WARLOCK_FIRERING); .. enum WarlockClassMasks uint64 { CF_WARLOCK_INFERNO = 0x0000010000000000, CF_WARLOCK_FIRERING = 0x0010000000000000, } However as far as I know such constructs are not possible with compiler < C++11 But it might be possible to use (or something very similar) static const uint64 CF_WARLOCK_FIRERING = 0x0010000000000000; ... Any input about this?

Well, this is just some basic test if the content of creature[_template]_addon.auras or EventAI, Action CAST have reasonable spells. http://paste2.org/p/1898514 Especially this is a good starter to prevent crashes introduced in recent GetAngle ASSERT I think all ppl interested in good DB things should use this .) (Will probably add some similar logging on master in a while)

Here is an updated version. I am just too lazy to do a proper test. Would appreciate if somebody would test this. (both summoned-type cases if possible) diff --git a/src/game/WaypointMovementGenerator.cpp b/src/game/WaypointMovementGenerator.cpp index fb3012a..23e5ffa 100644 --- a/src/game/WaypointMovementGenerator.cpp +++ b/src/game/WaypointMovementGenerator.cpp @@ -41,6 +41,7 @@ alter table creature_movement add `wpguid` int(11) default '0'; #include "ScriptMgr.h" #include "movement/MoveSplineInit.h" #include "movement/MoveSpline.h" +#include "TemporarySummon.h" #include <cassert> @@ -224,10 +225,24 @@ bool WaypointMovementGenerator<Creature>::Update(Creature &creature, const uint3 return true; } -void WaypointMovementGenerator<Creature>::MovementInform(Creature &creature) +void WaypointMovementGenerator<Creature>::MovementInform(Creature& creature) { if (creature.AI()) creature.AI()->MovementInform(WAYPOINT_MOTION_TYPE, i_currentNode); + + if (creature.IsTemporarySummon()) + { + TemporarySummon* pSummon = (TemporarySummon*)(&creature); + if (pSummon->GetSummonerGuid().IsCreatureOrVehicle()) + if (Creature* pSummoner = creature.GetMap()->GetCreature(pSummon->GetSummonerGuid())) + if (pSummoner->AI()) + pSummoner->AI()->SummonedMovementInform(&creature, WAYPOINT_MOTION_TYPE, i_currentNode); + } + else if (Unit* pOwner = creature.GetCharmerOrOwner()) + { + if (pOwner->GetTypeId() == TYPEID_UNIT && ((Creature*)pOwner)->AI()) + ((Creature*)pOwner)->AI()->SummonedMovementInform(&creature, WAYPOINT_MOTION_TYPE, i_currentNode); + } } bool WaypointMovementGenerator<Creature>::GetResetPosition(Creature&, float& x, float& y, float& z) (fixed copy and paste typo, thx to reamer for pointing)

Actually, you are free to use VCExpress as long as you want without fee. You only need to register within the 30days period

some alternate version to assert that user is valid in 11922. Thank you

@Jethro What do you mean with M0 or G0? mmaps extraction is rather long, expect 8 hours/ numCpu (numCpu <= 4) MMap Extraction Extracting mmaps takes very long, especially the continents take > 1 hour If you start on windows in Git-Bash, you won't see changes now and then, just be patient. Also you can look into the log-files for changes You can skip a few (or all) mmaps (continents are the biggest by far), and just add them whenever you want to

First crash-fix added in 11920. Thank you. But I agree that there is much more to be checked and improved as well.