VladimirMangos

As i remember only thing that stop apply it early has been unfinished discussion about effectiveness by memory use suggested structures. Note: in fact patch also expected speedup work code because lookup skill data is very often operation.

update spam part must be fixed in [8890]

In [8858]. Thank you for patch publishing at this forum, specially useful in result current problems with getmangos.ru. Only please provided original author of patch like QAston in this case as pointed in one from posts. In short: i will not post in patch thread (and read it in most cases) if not sure that have time for patch review or have fresh data in memory for useful comments in it without code search and other reseach. Other devs maybe do in same way. If you think that i must check all patches and make quick after publishing useful notes or add it to core for each suggested at forum patch, then (if i will so crazy agree with it) i will not add any code in mangos until next year and use all time just for read and reseach all published patch threads and comment in its... because if i finally add some patch before review all other patches then other patch author will waving why other patches added but his waiting so long If return to patch in added form. It has one performance problem: it send non-empty threat list always each 1 sec including case when threat list not updated from prev. send time. So make useless spam in network.

This is not mangos problem if someone call this function with NULL. And in mangos case this low chance happen or must fixed properlty instaed adding random place checks

I for example read thread first time as i remember. This just impossible read all (if you plan read with proper review problems and patches). Ofc, some time useful (and sometime strange) patches not have long dev remarks. But you must understand 3 point 1) is mangos dev review your patch or not, community memebers still can use it. So not added to mangos core patch not make it useless 2) base at time limit often patch long time wait review. But this not meaning that devs not interested to patch or related problems, just his time not reached. 3) You will not right if think that i can look at pacth and say in _all_ cases is patch correct or not correct, and is this best way fix or not. For me need dependent from patch look code, understand if i forgot how this part work, look possible related problem if accept suggested way for fixing and etc. This is not read and answer process. And i sure community members in many case know more about related functionality, at least from how it must work from player view point.

In [8857]. Thank you

In [8856]. Thank you

In [8854] with changes. Thank you

As i understand comment this ttoaly wrong...Why you prevent heal not at dispel remove or stack? why have this all checks in general..

As i understand from discussion all referecned cases is guardians in fact and then already work with current code anyway. I see only single less stricted check maybe: allow scriting for charmed/controlled pets with creatures owner. They anyway work as guardians mostly.

In [8852]. Thank you

Alt. patch (similar in fact) applied in [8851]

In [8851]. Thank you

Only _proper_ data collection and calculation by each level base at armory data will accepted. Like fast fix will not accepted.

In [8847]. Thank you. I change sql for fix 1) used spell family mask 2) also spell description point to J-spells, not J-effects this important, because spells always targeted to enemy, but effects can be targeted to enemy and to caster also. in sql mask used fact that J-spells is _ranged_ spells by spell damage mode.

then high part must go to spellmask1 (20335, 0x00, 10, 0x00800000, 0x00000000, 0x00000008, 0x00000010, 0x00000000, 0.000000, 100.000000, 0);

I use thenecromancer code version as more correct and laise data for sql update. In like form patch added to [8841]. Thanks to both authors

at windows this is _not_ solution. TBB speedup mangos work at windows alot. In different from more limited affect at Unix/Linux (dependent from OS)

then maybe need config option for set mail limit (1..255) with default 50?

for express version already pointed many times: you need express edition with integrated SP1 for build mangos.

Well known wiki suggest for player accessable spell with this target only 90 degree cone. I not understand why creatures casted spells must be different? And then not undestand why target need 2 cases... instead just fix angle to correct.

qsa suggest in most time ready for adding patches or with smale changes, so maybe... but recently has been founded that some patches just not base at proved data angles in wowwiki is cleary reported different for some cone spell targets pasdVn suggest and have added many patches but as i remember i often need rewrite its before add

best way configure git always use Unix line ends at setup of git client and at commit own local chnages make sure use unix-line ends patches (and then use editores that preserve unix-lines while edit and save.

Build porblem confirmed. But suggested fix wrong. 0.12 packet not have realCount field and then data.put<uint32>(0, realCount); and related incorrect.

Check added in [8773]. Thank you I use continue, because client expect receive real mail amount also.