Unvanquished Forum

I have been mistaken about the BP generation being reduced when RC/OM is moved. I didn't take into account the decay rate of the BP generation and the time it takes to relocate the base. However, not being able to transfer BP between power field is still there.

What do you mean by transfer? There are no longer any repeater build zones.

When you mark a building from RC or a repeater and try to build something else on another repeater's power zone, you cannot build even when you have BP from the marked building. Let's say you have 2 BPs and you mark a turret for 2+(8) BPs. You should have total of 10 BPs but you cannot build anything if you try to build on another repeater's power zone. If you try to move and mark all the buildings in the base, you can only build repeater's at the new location. You won't be able to utilize the BPs from the marked buildings back at the original base. If you move the RC, you lose all the BPs from the building powered by the RC.
Alien buildings don't have this problem.

Tim pointed out on IRC that RGS that are built too close to each other will not only work less efficient but will completely stop generation or even decrease BP. I think what we should be heading for is an approximation of this (in 3D space, ofc):

EDIT:
After going through the code and confirming that negative generation rates can't happen we found out that the decrease in BP might be related to human bots repairing buildings. Still the proposed approximation might be a good solution. :D

EDIT2:
On request a more formal description of my idea and the algorithm (which is just a rough approximation of it):

The ideal idea
I preclude that RGS have an area of effect (AOE) described by a sphere in 3D space. If a is an AOE of a RGS, |a| is the volume of a. The total mine rate should be proportional to the union of all AOE volumes, so that if n RGS don't intersect they would mine at a rate of n * base_rate but if they are in the very same spot they would mine at a total rate of 1 * base_rate. Now the idea is that if some of the RGS intersect, the total mine rate should be n * base_rate * (|A_union| / |A_max|), where A_union is the union of all a1,...,an and A_max is the union of all a1,...,an if they were rearranged to not intersect. This is what I try to illustrate with the image above: if the red, green and blue spheres are the AOE of three RGS, the total mine rate would be proportional to the part of the image that is not black. The red RGS would mine at full speed in the red area, at half speed in the yellow and purple areas and at a third of the normal rate in the white area. This asures that building another RGS in the range of a number of existing (friendly) RGS will never be worse than not building it.

The approximation
The exact implementation would be very computationally intense, so here's a rough approximation that should be easy to implement and light on performace. It is not accurate and may lead to a negative gain if a high number of RGS is built too close to each other (3 RGS in the same spot would mine at a total of 3/4 of the mine rate of a single RGS, 4 at 4/8, 5 at 5/16, etc). Maybe the algortihm can be improved to yield more accurate results when a high number of closeby RGS is involved.

For each pair of RGS r1,r2 whose distance is smaller than two times the RGS radius
{
Let r1.rate and r2.rate be the respective current mine rates of r1 and r2.
Calculate the (approximate) quota q in ]0,1[ of the intersection area of the two AOEs.
Set r1.rate to r1.rate * (1.0 * (1.0 - q) + 0.5 * q).
Set r2.rate to r2.rate * (1.0 * (1.0 - q) + 0.5 * q).
}

EDIT3:
I calculated the exact value of q which is a third grade polynom and doesn't need to be approximated ( q(r,d) = ((d/r)³ - 12*(d/r) + 16) / 16, r is the RGS radius and d is the distance of the two RGS). The patch is upstream and will be online on the development server soon.

Why not just prevent drills and leeches from being built within certain distance from each other?

I've also noticed the drop in BPs when drill was destroyed.

An artificial limit preventing drills/leeches built within a certain distance of each other makes no sense. The BP drops might be due to bots repairing stuff.

Ishq wrote:

An artificial limit preventing drills/leeches built within a certain distance of each other makes no sense. The BP drops might be due to bots repairing stuff.

Actually, it does. Atm, drill spam is an entirely profitable operation, and there is nothing to discourage it, especially to newbies who don't know about how it works.

Atm, drill spam is an entirely profitable operation

Except it is not. Although, I do agree I do need to discover a way to obviate the mine rate. Viech just gave me a nice patch to show the mine rate, which I will probably commit soon.

Ishq wrote:

Except it is not. Although, I do agree I do need to discover a way to obviate the mine rate. Viech just gave me a nice patch to show the mine rate, which I will probably commit soon.

Explain this.

Simple. Building like 3 drills close to each other gives each drill an output of like 0 resources per minute. It is therefore not profitable. (Also, for massive FPS boosts, in Options->Build Range Markers, disable Draw Intersection and Draw Frontline).