One-dimentional shape exploit solution

[gd80]

Ok, everybody knows that new power system is exploitable through building dumbbell- and spaghetti-shaped ships. That happens because you just need to place stabilizer at some distance from reactor and voila - it works, what causes one dimentional shape to be exploited. What about to make stabilizers efficiency be dependent on their distance from other stabilizer blocks? i.e.:
S
S R☐☐☐☐ S
S
~equals
SR☐☐☐☐SS
S
~equals
S☐☐R☐☐S
S S
~equals
SS☐☐R☐☐SS
~equals
S
☐
S ☐ R ☐ S
☐
S
Where R - reactor blocks, S - stabilizer blocks, ☐ - non-power/empty blocks

IDEA:
The idea, as I mentioned above, is to make big reactor require sertain amount of stability to be operational BUT amount of stability gain from each stabilizer block should be dependent on distance from current stabilizer to all others. Formula example for stabilizer block #1:

#1StabilizerEfficiencyCoefficient =
(AbsoluteDistanceFrom#1To#2byX+AbsoluteDistanceFrom#1To#2byY+AbsoluteDistanceFrom#1To#2byZ)
+
(AbsoluteDistanceFrom#1To#3byX+AbsoluteDistanceFrom#1To#3byY+AbsoluteDistanceFrom#1To#3byZ)
+
(AbsoluteDistanceFrom#1To#4byX+AbsoluteDistanceFrom#1To#4byY+AbsoluteDistanceFrom#1To#4byZ) +
...

I bevieve, this, because of + instead of * everywhere will equalize 1-dimentional ships with 2-dimentional and 3-dimentional ones from power point of view.

Of course, 1-dimentional ones will still gain advantage in combat because they have less blocks to be aimed to while met "face to face". There is solution for this problem also - make ship spin momentum gain depend on size of the ship:
spaghetti ship that is long by X should be very slow to spin around Y and Z.

Maybe it's too late to make such a proposals, but better late then dumbbell, right?

 

[NeonSturm]

What if the first one is one far away and all others are close together?

And it is not good to enumerate through all blocks of type x and ask for xyz distance. It can be made with groups of them.
But even then, you shouldn't have too many if you get each xy pair to compare.

 

[gd80]

What if the first one is one far away and all others are close together?

Shouldn't be a problem, because we have sum instead of multiplication. Anyway, this formula is just an example.

 

[NeonSturm]

sqrt( sum(dx², dy², dz²) )

And you need to check 2 against 3+4 and 3 against 4 too.
/-------------------------4--\
|
1-------------------------2---1 alternatively
|
\-------------------------3---/

 

[Valiant70]

This is essentially the same thing that others have suggested here and here, and possibly elsewhere too. It's an obvious solution that can be spiced up as much or as little as Schema wants to. If stabilizers remain one-dimensional, they'll probably be turned off (set to zero distance) on most servers. The power system is functional without them.

Personally, I like the idea of changing stabilizers into heatsinks/radiators for a heat/cooling system. At the very least, I think they should be connected by conduits to lower the effectiveness of island builds.

 

[MacThule]

It would certainly be a step in the right direction... I'm guessing it could still be gamed to optimize low-profile, linear ships.

The main issue is that one way or another if you want to terminate linear building, you must involve more relationships than a single [maximum distance] figure between only two element clusters. Using only two elements (reactor - stabilizer) literally creates a two-point dynamic for controlling shape. The shortest distance between two points is a straight line, can only ever be a straight line. The oversimplified geometry of the reactor-stabilizer relationship is why the meta has gone linear.


The OP suggestion inspires me to speculate about what would happen if you required multiple, smaller stabilizer groups (instead of allowing a single stabilizer group for each reactor), each having a [minimum distance] type relationship with each other and also a [minimum AND maximum distance] type relationship with the reactor.

I believe it would almost certainly force a spherical expression of reactor-related components.

EDIT: Because by involving a maximum distance relationship you couldn't just string each stabilizer cluster further and further back along the same line.