Two words of warning about using messages from KTM to feed Threat-1.0 :
1) KTM sends merged pet and master threats, where Threat-1.0 keep them seperated.
2) KTM sends either global threat or threat against the main target, where Threat-1.0 keeps per-target amounts.
For thoses two reasons I think Threat-1.0 SHOULD NOT use KTM messages. Or at least should people have the option to disable it.
(BTW... knowing this, I won't use KTM again).
In answer:
1) It either sends pet or master threat depending which is higher.
2) Threat-1.0 handles that just fine. I made it handle the mastertarget and the global stuff for KTM.
1) It either sends pet or master threat depending which is higher.
I'm not sure that's better :/
BTW how do you handle it in Threat-1.0 ?
Read the code :p
Basically, if a mastertarget is set in KLTHM and data is received it's added to the data for that target in Threat-1.0. If no mastertarget is set it's added to the _GLOBAL_ data.
Basically, if a mastertarget is set in KLTHM and data is received it's added to the data for that target in Threat-1.0. If no mastertarget is set it's added to the _GLOBAL_ data.
Adirelle, while you're here, I think that your pet Growl calculation code is broken for less-than-max-level pets. My 25 hunter showed his pet generating ~900 threat with Growl last night, and I was pulling aggro off of it well before that 900 limit was reached.
That said, there's no good way to perfectly implement KTM threat in ThreatLib. To that extent, I've provided ThreatLib:IsUsingForeignThreatSource(partyMember), which you can use to tell if threat is natively coming from ThreatLib or not. In Omen, I flag non-native users with an asterisk to note that their threat may be inaccurate.
It shows the master, not the pet. No way around that.
Do you mean "It shows the master OR the pet, whichever is higher" ?
Quote from Antiarc »
Adirelle, while you're here, I think that your pet Growl calculation code is broken for less-than-max-level pets. My 25 hunter showed his pet generating ~900 threat with Growl last night, and I was pulling aggro off of it well before that 900 limit was reached.
My implementation of the aggro could be broken, yes. But for the calculation itself, it is the one you can found on wowwiki and in KTM. BTW, it seems that every threat-generating skills are scaled for pets.
So the question is, what is the Growl rank and your pet AP ?
...
Oh wait, I'm just seeing in KTM that Growl scales only for 60 and higher level pets... Gonna fix that.
Quote from Antiarc »
That said, there's no good way to perfectly implement KTM threat in ThreatLib. To that extent, I've provided ThreatLib:IsUsingForeignThreatSource(partyMember), which you can use to tell if threat is natively coming from ThreatLib or not. In Omen, I flag non-native users with an asterisk to note that their threat may be inaccurate.
That's a good workaround to the problem. (Gonna add a threat formatter to do so in Violation: p)
Hm, will Threat 1.0 send data in "KTM-Format" to other raidmembers?
Reason:
Even though Threat 1.0 perhaps is the better alternative, there will still be people who use only KTM - and it would be kinda unfair to cut them off.
Perhaps make this option toggleable (even though I doubt it coud have any significant impact on performance to send 2 addonmessages with your current threatlevel instead of 1...) or perhaps even better, let Threat 1.0 check if some others use KTM (addonmessages...), if yes, send your threat also "KTM-style" if not, don't...
Perhaps eventually, but Threat tracks threat on a per-target basis, and KTM expects a sum of all threat generated in the fight to that point; I'm still mulling over the validity of any data sent to KTM.
Perhaps eventually, but Threat tracks threat on a per-target basis, and KTM expects a sum of all threat generated in the fight to that point; I'm still mulling over the validity of any data sent to KTM.
I'd suggest implementing such to KTM communications as a seperate mod. That way people can decide for themselves whether they should or should not send their threat to KTM users.
I don't see the need. If people insist on running an inefficient and inadequate mod then isn't that their problem? I know my guild will be switching to ThreatLib once the values are all in :) .
I don't see the need. If people insist on running an inefficient and inadequate mod then isn't that their problem? I know my guild will be switching to ThreatLib once the values are all in :) .
Hm, as much as I know, KTM was the first threatmeter out there, or at least the best developed one...
Threat 1.0 uses many values for threat coefficients from KTM that otherwise would be very difficult to get in such a short time.
It would be also unfair/unlogical to get threatlevels from people who use the other _working_ threatmeter but be selfish and keep your values for your own people...
I see some kind of oRA2 <--> CT_Raidassist parallell here.
At least make it toggleable/LoD plugin to send optional KTM-style threatnumbers (no fixed target: GLOBAL-threat, if the target is fixed the threat for the target...)
i can see a com's system working provided that a master target is set, that i believe would work best.
If a master target is set, send and recive data for that target, and if none is set then don't do anything with KTM data
I don't see the need. If people insist on running an inefficient and inadequate mod then isn't that their problem? I know my guild will be switching to ThreatLib once the values are all in :) .
And if your guild forces you to use that inefficient and inadequate mod because they're too set in their ways? A way to report data from Threat to KTM is pretty essencial.
But it gets sent with your playerName not the petName
Right, and that's the problem. I can hear from here people telling me to "FD NOW !" when I just did, because it's my pet that is very threatening... :/
(BTW, that remembers me when we did Onyxia with KTM and I was BM-specced.)
Quote from FlareCDE »
And if your guild forces you to use that inefficient and inadequate mod because they're too set in their ways? A way to report data from Threat to KTM is pretty essencial.
Wouldn't be to hard to get a KTM-sender working though I wouldn't put it in the ThreatLib core.
I don't see the need. If people insist on running an inefficient and inadequate mod then isn't that their problem? I know my guild will be switching to ThreatLib once the values are all in :) .
And if your guild forces you to use that inefficient and inadequate mod because they're too set in their ways? A way to report data from Threat to KTM is pretty essencial.
It also is better that even if your guild doesn't force you to, there's always a few people that view new developments on these things with some distrust and simply refuse to update because what they have 'works well enough' even if it's not as good.
I know a lot of the people I know still use CTRA, but this isn't an issue as oRA and CTRA can work together.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
-
View User Profile
-
View Posts
-
Send Message
Curse Premium1) KTM sends merged pet and master threats, where Threat-1.0 keep them seperated.
2) KTM sends either global threat or threat against the main target, where Threat-1.0 keeps per-target amounts.
For thoses two reasons I think Threat-1.0 SHOULD NOT use KTM messages. Or at least should people have the option to disable it.
(BTW... knowing this, I won't use KTM again).
In answer:
1) It either sends pet or master threat depending which is higher.
2) Threat-1.0 handles that just fine. I made it handle the mastertarget and the global stuff for KTM.
-Ammo
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumI'm not sure that's better :/
BTW how do you handle it in Threat-1.0 ?
Read the code :p
Basically, if a mastertarget is set in KLTHM and data is received it's added to the data for that target in Threat-1.0. If no mastertarget is set it's added to the _GLOBAL_ data.
-Ammo
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumI meant pet/master.
-Ammo
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumThat said, there's no good way to perfectly implement KTM threat in ThreatLib. To that extent, I've provided ThreatLib:IsUsingForeignThreatSource(partyMember), which you can use to tell if threat is natively coming from ThreatLib or not. In Omen, I flag non-native users with an asterisk to note that their threat may be inaccurate.
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumDo you mean "It shows the master OR the pet, whichever is higher" ?
My implementation of the aggro could be broken, yes. But for the calculation itself, it is the one you can found on wowwiki and in KTM. BTW, it seems that every threat-generating skills are scaled for pets.
So the question is, what is the Growl rank and your pet AP ?
...
Oh wait, I'm just seeing in KTM that Growl scales only for 60 and higher level pets... Gonna fix that.
That's a good workaround to the problem. (Gonna add a threat formatter to do so in Violation: p)
no, it sends the threat as master threat afaik, even though its secretly the pet, but I could be mistaken.
-Ammo
Reason:
Even though Threat 1.0 perhaps is the better alternative, there will still be people who use only KTM - and it would be kinda unfair to cut them off.
Perhaps make this option toggleable (even though I doubt it coud have any significant impact on performance to send 2 addonmessages with your current threatlevel instead of 1...) or perhaps even better, let Threat 1.0 check if some others use KTM (addonmessages...), if yes, send your threat also "KTM-style" if not, don't...
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumHm, as much as I know, KTM was the first threatmeter out there, or at least the best developed one...
Threat 1.0 uses many values for threat coefficients from KTM that otherwise would be very difficult to get in such a short time.
It would be also unfair/unlogical to get threatlevels from people who use the other _working_ threatmeter but be selfish and keep your values for your own people...
I see some kind of oRA2 <--> CT_Raidassist parallell here.
At least make it toggleable/LoD plugin to send optional KTM-style threatnumbers (no fixed target: GLOBAL-threat, if the target is fixed the threat for the target...)
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumIf a master target is set, send and recive data for that target, and if none is set then don't do anything with KTM data
-edited for sanity.
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumAccording to this piece of code extract from KTM_Net.lua, KTM sends max(playerThreat, petThreat), period : http://ace.pastey.net/55529
But it gets sent with your playerName not the petName
-Ammo
-
View User Profile
-
View Posts
-
Send Message
Curse PremiumRight, and that's the problem. I can hear from here people telling me to "FD NOW !" when I just did, because it's my pet that is very threatening... :/
(BTW, that remembers me when we did Onyxia with KTM and I was BM-specced.)
Wouldn't be to hard to get a KTM-sender working though I wouldn't put it in the ThreatLib core.
I know a lot of the people I know still use CTRA, but this isn't an issue as oRA and CTRA can work together.