pasdVn

Yes, thats the point. question is, if this is coorect for all 'summon wild' spells?

Rebased patches onto [9443] and merged branch with [9443]. Additionally I substituted the armor penetration by spell penetration (as there are no scaling auras and this seems to be valid for all pets, this is just a hardcoded part). Maybe in fact it would also be more clever to solve the hitchance/expertise like that, because also hit chance/expertise is for all pets...?

Rebased the patches onto [9443]. So pachtes should apply without errors now.

Include Lightguards suggestion and it will work really good :-) I did not expect that^^

No, that is another problem :-/ Thanks! I already discovered that (or in fact it was hunuza^^) but did not update yet. I'm gonna do that soon.Edit: Links updated.

Thanks for trying, but are you sure about that? I tried it one more time with spamming steady shot and also other shots in between. Could not find such a behavior :-/ Maybe you skilled 'Wild Quiver'... In combination with some client lags it will look like a machine gun^^ ...no idea. Keep me up to date if you found out something more. btw.: What do you mean by that?

Patches and first post updated.

Aha ... But enought of spamming now!

Hmm, this will become quite hacky, if this formula is really correct^^. But anyway: [9259] just implements the bonus of "Borrowed Time". This patch also did not implement the whole formula and was more a try to handle bonuses more generic. => Rejected Any update of the formula should be done in a new thread. Edit: It would be interesting to know, if e.g. Twin Disipline contains the family flag of PW:S...

[9259] The direct way, but maybe it is the best like that for now :-)

Hmm, ok. Maybe that makes more sense, as the value in DynamicObject::Create seems to be correct than. So seems that this fix is incorrect than. Edit: Tryed to remove this line, but it did not have any effect on the display size...

That is a wired target Problem :confused:. Target of this cast is the unit you want to attack - ofc, but both effects have only implicitTargetA SELF. So in the end there is an error in spell::checkCast because the triggered spell is a positive one with a harmful unit target. we have to change either something in checkcast (but the target system is very complex... :-/) or maybe better way to do it in spell::effecttriggerspell (send the caster as unit target for spell casts).

Hmm, ok, I'll take a look at the combat log, maybe that helps, thanks :-)

Hi everybody! I tried to fix the spell "Reckoning", and discovered the following Problem: When a normal weapon swing triggers a spell, that adds an extra attack (like Reckoning does) it will only be consumed in the next weapon swing (see Unit::AttackerStateUpdate). Additionally, when you have one unconsumed extraAttack left, the next normal weapon swing can't add a new one, because this is explicitly forbidden in Unit::HandleProcTriggerSpell (what makes sense, otherwise extra attacks could recoursively trigger each other). In the end you have the effect, that only every second hit will be a double hit with Reckoning (it has 4 charges... but you will need a very fast weapon to consume all in 8 seconds :-P) My sollution is, not to use this store variable "extraAttacks" in AttackerStateUpdate. If we need it more save. we could also use one, but after the proc calculation of the normal swing (or the meleespellcast). From 9c145f9b168aa081934a51a45746b5399d46f88e Mon Sep 17 00:00:00 2001 From: [email][email protected][/email] Date: Sat, 13 Feb 2010 22:36:24 +0100 Subject: fixed extra attacks handling It is possible now, that extra attacks, that have been added during a proc spell of a normal autoattack, becomes consumed at the same weapon swing. --- src/game/Unit.cpp | 10 +++++----- 1 files changed, 5 insertions(+), 5 deletions(-) diff --git a/src/game/Unit.cpp b/src/game/Unit.cpp index 6d2574c..74b7dda 100644 --- a/src/game/Unit.cpp +++ b/src/game/Unit.cpp @@ -2305,18 +2305,17 @@ void Unit::AttackerStateUpdate (Unit *pVictim, WeaponAttackType attType, bool ex else return; // ignore ranged case - uint32 extraAttacks = m_extraAttacks; - // melee attack spell casted at main hand attack only if (attType == BASE_ATTACK && m_currentSpells[CURRENT_MELEE_SPELL]) { m_currentSpells[CURRENT_MELEE_SPELL]->cast(); - // not recent extra attack only at any non extra attack (melee spell case) - if(!extra && extraAttacks) + // extra attack only at any non extra attack (melee spell case) + if(!extra && m_extraAttacks) { while(m_extraAttacks) { + // note: extra attacks can't proc spells adding new extra attacks AttackerStateUpdate(pVictim, BASE_ATTACK, true); if(m_extraAttacks > 0) --m_extraAttacks; @@ -2348,10 +2347,11 @@ void Unit::AttackerStateUpdate (Unit *pVictim, WeaponAttackType attType, bool ex ((Creature*)pVictim)->AI()->AttackedBy(this); // extra attack only at any non extra attack (normal case) - if(!extra && extraAttacks) + if(!extra && m_extraAttacks) { while(m_extraAttacks) { + // note: extra attacks can't proc spells adding new extra attacks AttackerStateUpdate(pVictim, BASE_ATTACK, true); if(m_extraAttacks > 0) --m_extraAttacks; -- 1.6.5.1.1367.gcd48 And the sql part for the talent-proc-aura (prox ex on block, on hit, on crit): /*Reckoning*/ DELETE FROM `spell_proc_event` WHERE entry IN (20177, 20179, 20181, 20180, 20182); INSERT INTO `spell_proc_event` (`entry`, `SchoolMask`, `SpellFamilyName`, `SpellFamilyMask0`, `SpellFamilyMask1`, `SpellFamilyMask2`, `procFlags`, `procEx`, `ppmRate`, `CustomChance`, `Cooldown`) VALUES (20177, 0, 0, 0, 0, 0, 0, 67, 0, 0, 0), (20179, 0, 0, 0, 0, 0, 0, 67, 0, 0, 0), (20181, 0, 0, 0, 0, 0, 0, 67, 0, 0, 0), (20180, 0, 0, 0, 0, 0, 0, 67, 0, 0, 0), (20182, 0, 0, 0, 0, 0, 0, 67, 0, 0, 0); Edit: Hmm, SQL not required. Just realized that this should not prox 'onBlock' :-/

Yes, nearly^^ I think you have to change some more things (e.g in totem::unsummon) but in the end you are right: No big deal ;-)

How it should work: On official Servers you see icons of your summoned totems under your character label/mana bar. How it works: On Mangos this works just if you have a completely new shaman and you are casting any specific totem for the very first time. All further casts won't cause those icons to appear. I think this is caused, becuase we don't send the correct package on unsummoning totems. Currently this is done with SMSG_CANCEL_AUTO_REPEAT. Maybe this is the correct package type (deactivating any actionbutton at the client or so, as this is also done to stop auto repeating spells) but in this case the data is wrong somehow? Seems as if the client saves, weather a totem is correct unsummoned or not, and so you won't see the icon for a specific totem no more than one time!? Don't have access to an official account and no idea of packet sniffing, so I can't find a sollution here :-/

Maybe it is generally better to make totems abled to use more than one spell? Totem of Wrath for example also should cast two spells...

Question ist just, if you wand auto shoot starts instant, or if it needs 0.5s before fiering the first shoot, after activating. The rest does not need any update, imho (gcd client handled). It's just important, becuase we could remove the flag in this case autorepeatfirstcastflag (or how it is called), what makes everything more clear^^

OK, in detail: Hunter autoshot acts like that imho (on offi server): * it gets casted instantly when you start to cast it (without 0.5 sec 'casttime') * no ranged spell can suspend it, except channeling (ok.. just volley here :-/ ) * all other spell can suspend autoshot (e.g. scare beast), but won't delay it for the 0.5s 'casttime' (see first point) Question is, if the wand ('auto') shoot acts the same concerning the 'casttime' (that means does it start instantly or after the 0.5s casttime)? By the Way: The flag you found seems not to be the correct one :-/. I first thought it could be a melee attack timer reset flag, because it was changed at "Slam" with patch 3.0.2 but this also seems to be wrong :-(

Hmm, why two times? I think both spell's whole damage should do x% more damage, so the part in SpellDamageBonus would be enough, or am I wrong? PS: I would code one loop after the other, not interlace both in each other^^

Does anybody know, how "Shoot" (the wand shoot) currently reacts on official servers? Are the 0.5s castime removed, too?

Alternative implementation in core in the meantime. Please move to rejected!

I just realized, that pet's don't share the owners armor penetration, but the owners spell penetration. Can anyone confirm that?... I'm confused now. I could swear I read about it somewhere, or has this been changed already? Should really be just a DB issue now. The actual UDB has around 2400hp at level around 70 (I tink in older releases it was around 12k ).

Update: * fixed Ayane's crash :-) * make attack speed definable via creature_template (for all except hunter pets, ofc.), so shadowfiend should now be faster * added haste scaling for dk pet, I also checked the other stats, but they don't fit at all to my less values I gathered on an official server, so I leave it as it is for now * added the second water elemental^^ Damage school could be always set by db (also before this patch), so shadowfiend, fire elemental totem... should work (as long as you have the correct db values). Also updated for current HEAD (but no merge errors, so does not matter).

Yes, of course. Cooldown event will be send on pet death or unsummon (what comes earlier). Problem is, that the spell you are casting, is not the one that summons the pet, meaning, that the cooldownevent is not sended for the spell you originally castet, but a triggered one. In fact this should not be a problem, because those 3 spells have a shared cooldown (see here), but client does not act as expected (show the cooldown at the original casted spell as well). I solved this by additionally send a cooldown package for the cast-spell.