VladimirMangos

I think you submit patches for share with other ppl... with community. Yes, when patch added to offcial sources it solve many problems for author: now mangos team problems have related code up-to-date, fix found bugs, and etc. BUT... ppl can use patch and without adding to official sources. I for example (except cases when i write some big patch many days) don't know self what i will do next Dependent from last read threads or bug report or some point writed at some team discussion. I not very orginized man... I glad to read this. Maybe we all have different views how project must developed, but we all work at project development and improving.

Are you sure? spell_tables is just helper data for code work. _Often_ SQL data useless without code support. And in anycase it required testing and debug code for check correctness, or maybe outdated code. Wyk3d, your post break my strong personal possition ignore this and similar thread totally after past october. But i still think that discuss this just useless and harm for team and for readers, for project. Anyway this not change our possition and not chnage way how each team member select what he will do next in project. For me like cry-threads just harm my fun from mangos development and harm my project work then. I think something similar applied to all readers in osme rate.

In [7424]. Thank you

I can't agree with like statement. Yes, proved that algorithm can be writed using only if-then-else, loop-with-single-exist and block. But in case loops already show that in some cases not allowing to break in loop in many points make need use additional state flags for checks and make code lot less clean. You must understand when theaching programming like pedantic way learn programmer use structured code and avoid avil GOTO Truth as oftent in middle. Also note when structural programming pedantic rules writed programming languages mostly not have exceptions support and nothing break linear code execution controlled by if/loop.Filter write way of function write is more like how exception code work: test and catch exceptions, then execute if all ok normal defult code. Noone say that all code must be in writed in way if(a) { s1; return;} if(b) return; ... s2; But this useful when a,b is exception/secial cases that need proccessed in special way. when use if(a) s1; else s2; then expected from code that s1 and s2 are _both_ normal case; when used if(a) { s1; return } s2; then s1 expected as special case and this very helpful see this from code structure.

Totally wrong. 1) in if(a && b) only _a_ calculated if it have false result and _b_ not executed and save time for program work in if(a & b) _both a and b calculated that can be lot more slow 2) in C/C++ any non-0 values is true values and then f(4 && 2) is laways true BUT if(4 & 2) is always false.

Yes, this is filter way of function code organization: at first checks and special cases with fast return then normal way work code. I like also this way code organization: easy reable without 2 page in right block indents. And easy edited later as pointed already.

All this "not chained" ifs have code ends by "return" so no repeating checks in success case. In fail check case original and "else if" version do same checks... But i agree that for speed _maybe_ switch(m_spellInfo->Id) batter have first before ifs. BUT this also dependent how often different spells casted...

Then you dead.

as i comment in github it broke build at VC90, related to ACE include magic fail...

Instead code edit with new spell_area table you can find appropriate spell with sanctuary effect and just make it always appliable at enter to zone, or pacific effect (last can't be removed by player in most cases and then will impossible attack, or both...

Threads mostyl use common data. This is not so nice case if after some thread crash other will work. Batter full server crash and restart in comparison charactrs data and then character DB corruption and continue work. [added]Ofc, batter not crash and not have bugs totally. But question about what do batter _when_ crash has been.

In fcd438e46c467caa1fb771f48a121988abb1019c. Thank you

In [7390]. Thank you [added] NO!We can;t port to mangos-0.12 any commits with sql update that change DB structure, or in other case update from 2.4.3 to 3.0.9 will by nightmare for ppl.

Yes, you are right. We can have different message for this two cases. My comment has been not deep thinked

I quote correct, can be impossible extract old map version for show in error mesage.

This can be totally unrelated file for maps. Ppl crazy sometime. So maybe problematic extract old file version.

Maybe add in same time same options for realmd for completeness? Also i worry about demon(0,0) call. Possible #include require maybe? Please anyone who test patch post what platform used test.

I reopen thread but please always post only for "Getting into this list" another project site/forum (DB or scripting or both) or updates (hosting change/etc.

Agree, fixes can applied step by step without dependent if in general view something lookup as step back. General program correctness is sum of local correct code patchs. In defferent "just work by chance" way. Applied in [7375] and mangos-0.12. Thank you Note: in reallity this need fix in 3 functions instead single

Or batter reserch of current mangos code:

Not related to mangos core

All other DB problems as already write early. Thread closed for prevent silly repeating posting already uanswered questions.

In [7369]. Thank you

Patch in [7368]. Thank you Also fix for inventory DB data corruption and new option OffhandCheckAtTalentsReset for seelct when unequip applied. Default: at zone change, more expected but not standart: at talent reset itself.

Need alsoinclude offhand = INVTYPE_2HWEAPON case. And patch will complite. I will review it in evening if time let.