Depth in the thrust system: use some physics.

[BoerToon]

I find the lack in depth of the thrust system discouraging. It is just stack enough thrust blocks on your ship and you are fine. I propose a different system which is more physics based but does not need a lot of processing power.

The general idea

You can build multiple groups of thrusters at a possible orientation. That group only provides thrust in that specific direction. One must choose the size of each group and placement according to the center of mass (which is already implemented in the game).
​

The physics

Each group is reduced to a point, the center of mass of the thruster group. With this point you can calculate the momentum the thruster group forces on the ship. The thruster system becomes a balancing act: if your main thruster group is slightly off center, so moving forward would cause your ship to pitch down, the forward thruster group are needed to compensate this angular momentum. You lose efficiency. With that thruster placement, the forward engines also fire up slightly.

This system adds an extra design parameter: thruster palcement/efficiency. It also allows some interesting ship designs which for instance can only move forward really fast and rotate along it's axes to move "backwards".​

 

[Reviire]

This would severely limit designs, I can't see it being in Starmade. So many designs would become inviable because they'd be, even at best, slightly off the center of mass.

 

[BoerToon]

This would severely limit designs, I can't see it being in Starmade. So many designs would become inviable because they'd be, even at best, slightly off the center of mass.

To be slightly of center of mass is no issue, you would just need a thruster in the appropriate direction to counteract this. You would not need to fire this thruster manually, Starmade should fire this up automatically.

A way to implement this system is to reverse the calibration system. The calibration system can not be user selected, they way you set up your thrusters defines how your ship is calibrated. The efficiency of your thruster system is defined by how much counterthrust is needed in to gain certain thrust direction.

But the calibration system should be enhanced as rotating and movement should be split in a positive and negative direction.

 

[Criss]

This is likely a mechanic that we will never see in StarMade. The goal is to let creativity flourish. With this a player would need to ensure their build functioned properly instead of the way they wanted it to look. Thrusters have become a common decoration on many ships.

Whats more, I imagine this will put more pressure on the performance side of things. Instead of the entire ship acting as one entity with one thruster pool, we would have multiple, possibly thousands of thruster pools interacting with the structure.

 

[BoerToon]

Whats more, I imagine this will put more pressure on the performance side of things. Instead of the entire ship acting as one entity with one thruster pool, we would have multiple, possibly thousands of thruster pools interacting with the structure.

It will not give a lot of pressure, the calculations occur during calibration. Nothing changes to the way we fly our ships just the relation to maximum speed, required power and mass is redefined. You could calculate it as below in the following steps:

1. You just setup a complete linear space of movement, 3 strafe movements (strafe along x, y & z axis) and 3 rotation movements (rotate along axis x, y and z). You map each thruster group to this space, so you get for each thruster group 6 numbers (these are real numbers).
   
   So for N groups of thrusters you get N vectors of size 6 (lets name it thrust vector)
   
   
2. With this list of sets you calculate what the best combination of thruster groups used to gain the 12 basic movements (You limit the solution space to only positive numbers). The magnitude of these numbers have no meaning so you can now map them all between zero and one.
   
   So for each movement as in ( 1 0 0 0 0 0), ( -1 0 0 0 0 0), (0 1 0 0 0 0), (0 -1 0 0 0 0) you retrieve N numbers which represent how much between 0 an 100 % they need to fire up to gain that movement of the ship.
   
   
3. From these lists you can calculate how fast you can strafe and rotate along the axis and with the power required to do so. These values will be used during the game and they remain static.

To incorpate damage to the system, you reduce the thrust a certain group can produce ( % of damage -> % of thrust ) and reduce the max speed of a certain movement by using the thrust vector (of that group) and the thrust number of that group to that specific movement.

 

[Valck]

I'd love to see the game use the actual vectors of the engines as defined by their placement, and their magnitude determined by their respective group values.

I seem to remember prior discussions about something like this, and I doubt it's going to be part of the main game.See Chris' post...
However, all the necessary functionality is there, and maybe enough of it will be exposed if and when™ a modding API happens.

 

[Madman198237]

Why make it this complex? You'd spoil almost every asymmetric build, every build that has a thruster bank off center, builds that use thrusters for decoration, every Star Wars build ever (Raised bridges, and vertical asymmetry in general!), you'd DESTROY Star Trek things, just for a few examples.

This system is, as Criss said, creativity-limiting and absurd in a voxel-based nonrealistic videogame. Sorry. Maybe try KSP or something?

 

[BoerToon]

Assymetric builds are no issue, u just need to counter off center placements with appropriate small thrusters at certain placements.

It adds some depth to this system. The game now is just bulk placement for so many systems: power tanks, shield capacitors, shield rechargers, thrusters, ... all require no thought to placement. Even power reactors bother me. Straight lines or twisted cube is most efficient.

 

[Lukwan]

I don't see any specific advantage to this idea. Besides a general 'more depth' in thrusters...what does add to to the game? The disadvantages are huge so, on the face of it; no...just no (thanks). Circling the wagons around your idea does it no favors. If you are trying to win a debate you need to make valid points and address the arguments on the other side.

Starmade already has plenty of depth and needs streamlining for a faster learning curve more than it needs stuff like this.

 

[Valck]

Why make it this complex? You'd spoil almost every asymmetric build, every build that has a thruster bank off center, builds that use thrusters for decoration, every Star Wars build ever (Raised bridges, and vertical asymmetry in general!), you'd DESTROY Star Trek things, just for a few examples.

If engines would allow some degree of thrust vectoring, and the calibration system could take that into account, asymmetric engine placement would still work unless the design is severely unbalanced, "severely" being defined by the exact values.

This system is, as Criss said, creativity-limiting and absurd in a voxel-based nonrealistic videogame. Sorry. Maybe try KSP or something?

I don't think a different approach to thrust mechanics would be limiting creativity for part of the player base at least. Obviously some would like it, some wouldn't, thus it would best be added as a mod.
I do play KSP (which has an awesome mod API by the way), but that's like saying "Don't like that there are no food mechanics? Play Minecraft instead".

 

[Lecic]

This is the sort of thing that belongs in SE, not Starmade. No thanks.

 

[BoerToon]

It adds another step to ship design. You can make an esthetically pleasing ship but then you would need to make it fly.
Not every ship should have the inherent ability to fly superbe. It adds difficulty and diversity in pvp ships. It adds proper damage assesment to propulsion systems. Left back engine is blown up, my forward movement is slower (other engines need to compensate momentum from right engine, my propulsion is less efficient.)

 

[sayerulz]

No. People have already explained why. This is not a physics simulator. The point is to make stuff that looks cool.