Global threat cap?

[Pagezero]

Here's a question, after you reached the cap did the tank swaps work, I assume they did otherwise you would said something. The reason I ask is because I thought taunt only kept after the 6 sec if your threat has increased by some amount from the point the taunt occurred. If this is case and the taunts were still working then the server's threat number must be a larger integer and is still be growing and the limit is in the client or the protocol.

[Muffin Man]

Think of it this way. Once the tank hits the threat cap, you no longer need to continue to generate threat at all, let alone faster than anyone. Threat becomes completely irrelevant at that point. In which case your priorities would shift completely to survivability and damage only (depending on the context) and that affects your ability selection.

IE if this doesn't get fixed, all Warriors will spec Impending Victory.

[Martie]

I haven't been raiding for a few months now, but there's some things I'm wondering about.

Do fights last ten minutes all the time now? If most fights are shorter, the cap won't be reached that easily.
Do dps do 40-50k dps now? If dps can't catch up with the tank anyway, this all doesn't matter that much.

[jere]

Here's a question, after you reached the cap did the tank swaps work, I assume they did otherwise you would said something. The reason I ask is because I thought taunt only kept after the 6 sec if your threat has increased by some amount from the point the taunt occurred.

This would be incorrect. Taunt does 3 things:
1. Makes you the current threat holder (permanent)
2. Makes your threat equal to the previous threat holder (permanent)
3. Forces the target to fixate on you for a short period.

Challenging Shout only does the 3rd part, but Taunt does all 3. And by permanent, I mean that the effects don't revert back after time elapses. You don't lose threat unless someone else has already passed 110% (or 130% if at range from the mob).

You can literally taunt a mob off of someone and if neither of you do anything else, that mob will stick to you from that point on.

[Dragaan]

Taunts still worked, and the last 2 minutes of the fight you can see the tank's names swapping positions on omen between #1/2 back and forth for some reason.

[Kojiyama]

Do fights last ten minutes all the time now? If most fights are shorter, the cap won't be reached that easily.
Do dps do 40-50k dps now? If dps can't catch up with the tank anyway, this all doesn't matter that much.

Even in current gear levels, tanks seem to be approaching cappable levels of damage on 7-8 minute fights quite easily.

[feralminded]

Even in current gear levels, tanks seem to be approaching cappable levels of damage on 7-8 minute fights quite easily.

Yeah with full vengeance I am always in the 40-50k TPS range when standing still.

[Olog]

The threat cap is very easily reached on Halfus because of the damage taken debuffs the drakes do. Especially earlier when the debuffs were applied as soon as you released a drake. But still I regularly get capped, just a bit later in the fight. I remember seeing myself and our other Halfus tank being capped on our very first kills there when we were mostly in blue gear. And yes taunts still work normally but you can't pull aggro any other way. In the end it doesn't really affect Halfus all that much but it might be a limiting factor for future fights where damage increasing buffs play some kind of a role.

[damnpatch]

It's likely I'm overlooking something obvious, and about to embarrass myself, but since threat is just a number, couldn't they just make that number smaller? As in, all threat values are 10% (or even 1%) of what they are currently, and keep the same precision, and problem solved? For a while at least. Sure, it's nice seeing big huge numbers, but we'd adjust to seeing smaller ones too.

~Patch

[Quinafoi]

Which method is actually easier actually depends on how they programmed it. It may actually be easier to convert the datatype to a 64-bit interger value instead of adjusting the precision of the number. Given the scope of how threat is used, I find it relatively unlikely this value would ever be written to disk. This data only exists in runtime (or more specifically combat). Given standards which exist for in the design of object-oriented programming languages, it could actually be considerably easier simply changing the parameters and argument value datatypes than adjusting how that value is actually used in the logic. It may be simple to say reduce the precision by two places, however in practice you must keep in mind that how you do this in programming is considerably more complicated than simply dividing by say 100. In order to adjust the precision, you have to adjust the precision of all calculations. Because the program can't know the larger number you wish to divide by 100. From a programmers perspective I would actually think it would be easier to change the datatype than to change the precision simply because it is easy to identify all the touch points where the variable is declared, but it is difficult to change all the points where the variable is used (there is simply more of them, and you have adjust logic instead of simply changing type). Since this data isn't saved to disk (the database) you don't have to worry about changing database data structures. And all math done to a 32-bit interger type still applies to a 64-bit interger type because any value which could be represented in a 32-bit interger can always be represented in a 64-bit interger.

From a programmer perspective when dealing with a value not stored in the database but only in runtime memory, it is much easier for me to change the word "Integer" into the word "Long" in the declarations than it is to dive into each function and change how it is used. And this would also be the prefered solution anyway long term.

I've had to do this type of change in programs many times myself, and in object oriented programming it tends to be the case that changing the datatype is always the easier approach. It may not be the case for them depending on how they programmed it, but my guess is it would be from my years of experience as a software developer.