Missile War

[Vomitus]

Hm... Maybe I don't know something but I think it must be much more simple than it seem to be.

I will take in attention only one barrier(Inner or Outer) to simplify the understanding of my reasoning.

Of course, SS is largely depend on a mechanism on which rockets movement and damage is built on. So, we can single out these runtime variable data on SS:

1. Max HP of barriers;
   
2. Regeneration rate;
   
3. Refresh timer(when it is destroyed);
   
4. Take damage from <source> (I dunno if this is handled in any special way but I think it shouldn't - any explosion from a rocket must deal damage to all barrier instances in a defined radius by a built-in mechanisms of a game and cause no additional lag in result);
   
5. Remove Shield trigger (condition of this trigger I must study a bit to understand it more precisely);

AFAIK there is no other sources of disturbance that may affect Shield State and therefore no other triggers to be involved in.

Analyzing:

1. This is determined by a number of Boosters(forget exact name) and Graded Boosters of each player in team. I also dunno exact formula for calculating this but this formula is not important. We can store this formula's result in a separate variable, which will be updated only when number of boosters is changed. And Max barrier HP will be calculated from this variable(or even make a variable that holds final number of Barrier Max HP and take that value from var directly without calculations). This will make all triggers that use barrier Max HP vale not to calculate each time all the formula. Calculation will happen only when number of boosters is changed in team what happens not that often.
   
2. This is the same as described above but with one small correction. The total factor that influences regeneration rate and max hp is the same and it depends on a number of boosters. So we can store two additional variables with numbers of Boosters and Graded Boosters(in case if we cannot use one var for storing "upgrade level" that could be used for both hp and regen - in case if graded booster gives different bonuses to hp and regen increase). And then calculate or store final vars that will represent final amount of hp and final regen rate(or regen multiplier).
   The idea here is that we wont make redundant calculations of the same formula for two different values.
   
3. This trigger is pinned to an event when hp of one of the barrier instances goes 0 or below. I don't see a problem here.
   
4. Also no triggers required.
   
5. See p.3.

So, nothing seems to be a problem if:

1. Event handling works as delegate calls, not as constant State checks(this I must research a bit to make sure). If we need to make constant checks - this really sucks and I don't generally know how could be this refactored.
   
2. Shield does not have self-linearizing feature. By this feature I mean self-equalization, when if one piece of barrier is damaged and other pieces are full hp, their hp will quickly flows to a damaged one so all the pieces in a shield will have equivalent hp amount. This process of equalization will take some time of course and it will be particularly constant, so it may cause significant lagging.

At this point I have no new conjectures so feel free to correct me.

Btw, where can I find source map that is not protected? I haven't tried to open it in editor yet(I have no WC3 by my hand right now) but I'll ask you first if it will e a problem.