Oniryck

Thanks Vladimir, I'll try to find the time to test it before tomorrow, check if I find any problem before release

Really really nice job. Big thanks to you guys! *bow*

Damn you're fast ! congratulation once again! First : enjoy your vacations Then, thanks again for looking into it. If you need data from a 2.4.3 TBC client no problem. Being a git noob I don't really know what cherry-pick is and how to do it, I'll find out. I mean I suppose it's about applying parts of commit to another branch but should it be done manually or does git handle it? I'll try to learn more about that, maybe I'll be more helpful after that. Edit: Okay I got the cherry-picking thing. Useful stuff ! If you need any help for this, let us know. I'll try my best with my modest knowledge.

vmaps v3 need to be ported first. Finding accurate documentation of old file and struct layouts is a challenge. If it would get us more testers, I guess I can start poking around in old MPQs for mangos zero/one. I'm sure it would bring more testers. I think a dev of the oregon core (I forgot his name hope he'll excuse me) already converted the vmaps v3 to be compatible with formerly mangos-0.12. I don't have any more info about that, but if it can help... I'd be willing to help but this side of mangos is not really known to me! Thank you to consider the question in any case.

I know I'm bringing the subject back from the abyss, but do you think this amazing project will be ported to mangos-one one day ? We'd really need something like that too And congratulation for the fabulous work, in any case!

Any news about porting this (and why not mmap redux, that is quite linked to this) to mangos 0.12? I guess it's not in the priority list, but if you have any idea about how many time it will take to do it... I'd be glad to help even if I'm not very fluent in that part of mangos. But I think it would be great for mangos 0.12 to have this too.

Hi, after commit [s0361], a small typo in code prevent players to unlearn their pets talent. Here's the correction : diff --git a/src/game/PetHandler.cpp b/src/game/PetHandler.cpp index 8d59705..d1887a8 100644 --- a/src/game/PetHandler.cpp +++ b/src/game/PetHandler.cpp @@ -569,7 +569,7 @@ void WorldSession::HandlePetUnlearnOpcode(WorldPacket& recvPacket) return; } - if (pet->getPetType() != HUNTER_PET || pet->m_spells.size()) + if (pet->getPetType() != HUNTER_PET || pet->m_spells.size() <= 1) return; CharmInfo *charmInfo = pet->GetCharmInfo();

Okay, Vladimir, thanks for this clarification

Merci pour ton travail Milk13, je te dirais si je constate d'autres sorts à ajouter à la liste. Thank you for you work, Milk13, I'll tell you if I found more data to add to the table.

I guess the only way is to manage all the spells manually For the example in this thread: INSERT INTO `mangos`.`spell_bonus_data`(`entry`,`direct_bonus`,`dot_bonus`,`ap_bonus`,`comments`) VALUES ( '10461','0','0','0','Shaman - Healing Stream Totem Rank 5'), ( '10460','0','0','0','Shaman - Healing Stream Totem Rank 4'), ( '6372','0','0','0','Shaman - Healing Stream Totem Rank 3'), ( '6371','0','0','0','Shaman - Healing Stream Totem Rank 2'), ( '5672','0','0','0','Shaman - Healing Stream Totem Rank 1'); At least I guess ! Milk13 will do this better than me.

Yeah, many coeff seems to be overevaluated. I modified them to find approximately the same score as before the update. Maybe you should do the same ?

Okay, for this, here my personal hack/fix. Works fine for this specific spell.

Yes after many search and a nice report from one of my players, I got the bug that way : In stratholme, a hunter using eye of the beast (controlling his pet) makes it enter a Ziggourat in the instance : crash. Then, everytime he tries to reconnect, the server crashes (that could explain why you got repeated crashes, one of your player should have this kind of issue and tried to reconnect regularely too). After deleted his pet from the DB he could reconnect and the crashes stopped. So it's definitely about the pet and around the code you pointed.

Patman128, I meant there is still chance some characters remains on the same position so the map is requested once again (the server load the map that had the problem once again) so if it has a problem, it's normal some crash come again (Dunno if I'm clear, my english is pretty lame)

Kinda normal, the map is still loaded if you restart the server right after the crash so it still try to unload itself and the creature in it, so it do the same crash for everyone creature in it posing problem

Hmm I'm still investigating any problem related to the CleanupsBeforeDelete() function for a Creature type object (since THAT seems to be the problem) but no result so far, except that the Creature object doesn't seem to have it's own CleanupsBeforeDelete function but uses directly the Unit:: one (at least I suppose so), unlike WorldObject and Player objects (but I don't see how this is a problem : the code there didn't change lately AND Creature object is a Unit, so it's normal it uses the Unit::CleanupsBeforeDelete function) I'm not sure I'm realy clear here Anyway, that crash happened to me once, but these last couple of days, no more crashes of that kind. I got a RemoveAura one, thought...

Do all your crashes happen approxymatively at the same hour every day ?

I was on s0036 before my merge. I kickly took a look but for now wasn't able to find anything that could be source of the problem, but I lack time lately to run a deep search through this.

I confirm that I got the same crash today after updating to last rev. (s0096)

Thanks to you Vladimir I've enhanced my understanding in this part of mangos. I think I've found the problem (even if I'm not entirely sure but it looks like it) : a creature summoned by another creature had a MovementType = 1 (random). And, as you said, it seems that shouldn't happen. So I put the MovementType back to 0, I'll see if I have the same crash again. Maybe if it happens again with another creature I should write a fix directly in the core to consider MovementType of a creature summoned by another creature automatically to 0 in Spell::EffectSummonPet

okay, thanks for your advice, I'll work on that!

Yeah that's what I saw too. I tried to call SetSummonPoint in EffectSummonPet. Dunno if this will help, what do you think ? I tried hoping this would solve anything, 'cause it's the only possible issue I could found so far.

Hi, ( rev. = s0036 custom patches = some, but I can't see one related to this ) From some time, I've got this crash on my server : http://pastebin.com/Yyz7Bjvy (I have several of these log always the same) After doing my homework, it seems like the RandomMovementGenerator<Creature>::_setRandomLocation(Creature &creature) function ask for the creature.GetRespawnCoord that return false values because the creature implied is a summoned one (so not referenced in the mCreatureDataMap). If I'm correct, I guess there would be two way of fixing this : either create an exception in the GetCreatureDataPair(uint32 guid) function to handle summoned creature, or do it in the RandomMovementGenerator<Creature>::_setRandomLocation(Creature &creature) function. But Maybe I'm totally wrong So I'm waiting for experts tip for this one.

read the command and use your config file. And type ".gm visible on" for the npc to be able to see and attack you...

Could find the time to search through the entire commit, but I thought too this was just a small thing : the comment by FragFrog in the commit says it all ^^ One again, thank you for your work Vladimir !