pasdVn

Hey guys, just a short question: Is it possible, to make git log, git diff etc. make tabstops show as 4 space (instead of 8, what is the default setting, as I think). I did not find any option until now :-( I try to use git with another project, with an IDE/editor, that does not support the conversion from tabstop to real spaces, as the vc++ IDE does. pasdVn

Won't work :-(. This requires some deeper changes in proc system, because frost trap applies a persistent area aura and has no direct unit target. Edit: But the cooldown part is correct ofc imo!

The original spell is just an aura, that is castet on the caster itself. This aura periodically triggers the "real" spells (dmg/healing, whatever). This original target is not stored in the aura anywhere. That is why we have to get it from somewhere else (this dummy aura in this case). I anyway have to update the patch, but don't have musch time currently, sorry :-(

Hmm, would be really a nice catch if this bit is correct. I usually do such researches quite unprofessional: Simply make a loop over all spells in the code anywhere you like (e.g. a spelleffect) and print out all spellids that have this flag.

As I wrote I excluded this Idea for now. So yes, this spell does not work correct, but not because this patch is incorrect, but because thie feature is simply not yet implemented ;-)

Thanks a lot for your feedback :-) I'll do my best to follow your advices.

Hmm, I did not notice such problems, but I'll have a look on that. That is what that patch is about!! (did you read first post!? :-P) Maybe that could be the next step. Will at least save some cpu time...

Yes, I also noticed that this system is not 100% offi like, when I had the possibility to check that. But I did not really get a logic out of that. The pet is definitive faster than it's owner whe it's out of combat and to far away. Seemed to me, that this speed could be even faster than the max possible pet movement speed and that this is also the case, when the pet is in combat (maybe maxspeed does not matter if the pet follows it's master!?). Don't know. I have too less information about that :-/

Fixed bug, when pet was not on maxspeed, while using the TMG in chase mode and rebased branch onto rev [8606].

Added "Improved Stormstrike" and "Concussive Barrage". Updatet Entrapment provflags (just snake- and frosttrap now).

Hmm, I can try that. But include it into this patch...!? Maybe one more word to the mana per int scaling: Really strange thing :-/ I did not manage to find a simple connection between both. Maybe anybody has more detailed information on that? Edit: I changed the stamina scaling of the warlock pet scaling aura to 70%. Links in first post are updatet.

Thank you for the patch! I'm currently working on the pet stats, too. Did you made any more detailed researches on the mana/intellect ratio? I have just some datas from lower level warlocks, but seems to be a non-linear connection - at least at those low levels.

Yes probably that, too. But anyway he is right. They don't get the normal 10hp per stamina and 15 mana from intellect, as normal (what was already tried to fix in the the thread reeshack postet). I did not notice this problem up to now :-( Hunter pets e.g. have their pet passive auras (e.g. http://www.wowhead.com/?spell=56634), what makes them in the end gain more hp from 1 stamina (this is also shown in the clinet - at least at the 3.2.0 version). Maybe there is a similar mechanic for wl pets. I try to find a clean sollution.

(OT): Maybe I'm wrong, but I thing the guardian_ai has nothing to do with guardian pets, but just with the normal guards (npc's).

Hey guys, I had the possibility to make quite a big research on offi servers about pet stats and scaling. Originally my aim was to find analytic expressions who the pet's stats scale with their level (and also owner), but scraped that idea soon, as every pet is different, and there are just too many irregularities. What is left of this work are two additional values in pet_levelstats (see patch discription). Anyway I found a lot of stuff to improve (in fact too much... :mellow:). Biggest change of this patch is the handling of scaling stats from owner attributes. Very important point is at first, that we have to decide between dynamic scaling pets (e.g. hunter- , warlock-, death knight pets) and pets that apply bonuses only on summon (temporary (controlled and non controlled) summons, guardian pets). For the dynamic scaling there are special scaling auras (http://www.wowhead.com/?search=pet+scaling#abilities). The patch will implement the use of those scaling auras. Because patch became quite big I tried to devide as much as possible to be clear: 1st Main Patch: http://gist.github.com/273788 * added mindmg and maxdmg field to pet_levelstats, the currently used formula in the code is quite inexact, especially for some kinds of guardian pets (e.g. the snake trap snakes, that do ways to high dmg with that formular, but also hunter and wl pets) * cleanup in Pet::InitStatsForLevel, the static bonuses from owner will be added there as unit mods (you won't have to acces them any more, so you don't need to store it somewhere else), also implemented some more bonuses for temp. summons and guardian pets (shadow fiend...) and scaling mods (glyph of feral spirit) * implemented basepoint calculation for dynamic scaling auras, and off-like update interval system what additionally prevents to update those auras too often (maybe this is too special now, with the aura-modifier-change-system suggested by thenecromancer..) * updatet scaling coefficints of hunter pets basing on my researches, witch give really exact results now. Note that the bonus values you get to see in the players char window are incorrect (e.g. if you mouse-over your armor value: increases your pet's armomr by...). Unfortunatly I could not carry enough data to verify the wl-pet scaling, but they seem to fit quit good. * also added dk pet scaling factors, basing on research * added scaling of hitchance and expertise, formular based on my understandig of the patchnotes with a quite simple system: pet gains all hit chance modifier of master (e.g. if master has 3% increased hit chance at all pet's hit chance will also increase by 3%) and a correspondig dodge/parry chance reduction (3% increased hit chance of master -> 3% less chance of an opponent to dodge/parry an attack of your pet)... hmm, I just see that is in fact an 3.2.0 feature ... but auras already exist... Maybe one more note to the pet bonus damage: Those auras exclude the physical spell school, but on official servers also physical spells will benefit from this damage. I think they did this, beacuse otherwise also the normal autohit damage would benefit, what is not wanted. As bonus damage calculation for physical-non-weapon-damage-based-spells is currently not supported by mangos we don't have to concern about that (probably best would be to simply exclude schoolmask check for all pet scaling auras). 2nd, just the sql part: http://gist.github.com/273793 I know that mangos does not provide any values of pet_levelstats. Anyway I won't to post them here (maybe any udb likes them). In fact there are not too huge differeces to the already exsisting values in udb (or at least no importatant stats). Most important is the damage here, of course. I will provide some more values soon (also wl pets and ghoul), but I'm to lazy to type now^^ 3rd: http://gist.github.com/273797 Just a temporary sollution for the following problem: If you recast -for example- a stamina increasing aura on a target, the targets current health will be set to the maximum health without any of both auras applied. Quite annoying when updating, because your pet will always loose health when updating the stamina aura. 4th: http://gist.github.com/273799 Add the hunter pet's happiness damage mod to spell bonus damage calculation (this is also scaling for me ) 5th: http://gist.github.com/290507 Implementation of spell (new, updated at 24.02.2010!!) penetration sharing of pets. 6th: (already in mangos/master) Make creatures also benefit of SPELL_AURA_MOD_EXPERTISE. This is just to make the expertise aura of pets (1st patch) have an effect :-) 7th: http://gist.github.com/273801 The custom scaling of hp/mana per sta/int. Original patch by corfen. last: http://gist.github.com/435611 A fix for scaling problem concerning the pet talent "Wild Hunt" (new, updated 12.06.2010) All pastes(gists) rebased on [10051]. Sorry for huge patch, but I don't want to spam the forum with that much singel threads, as all belong together -somehow-. To get all use: git fetch git://github.com/pasdVn petstats (merged with [10051]) I will update this branch (not rebase) this branch from time to time while the gist-pastes will be always rebased. Try it! A bit of feedback would be nice :-) pasdVn

Nono, I am not the starter of this thread. It was splitted and my comment became -whyever- the first post of this new thread. Patch is written by qsa.The starting post is #9.

Yes it is. I rechecked the autorepeating spell mechanics at an official server at 3.2.0. I think I'm going to write a patch to improve autohot performance when mangos is supporting this client version. This old patch is just a kind of temporary hack.

Ahhhh, allright. Now I understand your ideas :-). I really did not want to stress the discussion about those negligibilities... but good that this is clarified now

Hmm, very strange. You mean it is possible that an creature can have two ai's at same time (stack!?). Anyway besides that: As I understand the current code, the AI is selected by CreatureInfo->ScriptID. Imagine any creature that uses e.g. eventAI. The same creature (or at least a unit with the same creature template) can be summoned by a hunter. Your provided code will use the ai provided in CreatureInfo (eventAI in this example), but not the PetAI. This is a normal case and has nothing to do with the responsibility of the script writer. You can not simply claim, that beasts do not have any ai, because they can also exist as pets^^ I know that this is not the main idea of the patch Just want to clarify that because I started to comment on this. Also possible that I am totally wrong...

Looks quite promising from first view, thank you :-) One hint: * CreatureAI* selectAI(Creature *creature) * { * - // Allow scripting AI for normal creatures and not controlled pets (guardians and mini-pets) * - if ((!creature->isPet() || !((Pet*)creature)->isControlled()) && !creature->isCharmed()) * - if(CreatureAI* scriptedAI = Script->GetAI(creature)) * - return scriptedAI; * + // Try to get scripting AI first - go on only if there is none assigned * + if(CreatureAI* scriptedAI = Script->GetAI(creature)) * + return scriptedAI; * * CreatureAIRegistry &ai_registry(CreatureAIRepository::Instance()); No good idea. Imagine creatures with scripted ai, that can also exist as hunter pets. Edit: This thread has been splitted. Post #9 is the real starting post.

Hmm, are you sure about that? I did not work before at a similar problem, but I think this is done no where else in the code and is not needed (have a lok at send spell go, or send spell start, etc.). I tried the patch and did not have any problems. Edit: You took the wrong method (WorldObject::SendMessageToSet). It is redefined for Players -> Polimorphism (Player::SendMessageToSet), what is doning nothing more than you mentioned here.

Ahh, I see :-/ Thank you for fixing an reviewing!

No that is not correct. There are only multiple family flag fields, because they need more than just 32bit (there are 3 fields, the first two are united in a 64 bit value in the core, while the 3rd one is 32bit). Just imgine that you have 96 bits and want to find out if any of those bits fits^^

Return false if none of the bitmasks fit. Meaning, if any bit of the current triggering spell fits to the bitmask of the aura, the if will be false and you will go to the code after the if (the return true). So code should be ok like that ;-)

I noticed that I should also care about the category cooldowns... and some more things...:mellow: Finally patch grew quite big, but pet spell casting is more fun now updated first post with description and new patch