An option for a randomely generated solar system

[LostElement]

somewhat related, why do peole keep saying duna has no oxygen, I thought it did.

Since Duna is supposed to be mars, I suspect that Duna has a CO2 rich atmosphere

[TimMartland]

I would love this if it could be done, but I'm not the most confident and I quite like the Kerbol system as it is. Having a randomly generated system would take half the fun out of the shared missions, the "ooh, have you seen such-and-such place" and being able to go see what your friends saw in that crazy Mun crater.

If it was for interstellar travel, with random places to xplore, it would be so mind blowing. Plus, it would need insane work and might not be possible

[Poryy]

I would love this if it could be done, but I'm not the most confident and I quite like the Kerbol system as it is. Having a randomly generated system would take half the fun out of the shared missions, the "ooh, have you seen such-and-such place" and being able to go see what your friends saw in that crazy Mun crater.

If it was for interstellar travel, with random places to xplore, it would be so mind blowing. Plus, it would need insane work and might not be possible

This is an OPTION. You can still get the standard system by not checking the box. Plus, the mun and kerbin are always the same anyway.

[sumghai]

I still prefer that the option only applies to other planetary systems - the home Kerbol system should be the same no matter what.

[SuperWeegee4000]

I've started on a dice-roll chart. So far I've completed the gas giant category:

Size:

1: 1x Jool

2: 1.1x Jool

3: 1.2x Jool

4: 1.3x Jool

5: 1.4x Jool

6: Roll again

[second Roll]

1-2: 1.5x Jool

3-4: 1.6x Jool

5-6: 1.7x Jool

Gravity:

1: Eve

2: 0.9x Jool

3: 1x Jool

4: 1.1x Jool

5: 1.2x Jool

6: Roll again

[second Roll]

1: 1.3x Jool

2: 1.4x Jool

3: 1.5x Jool

4: 1.6x Jool

5: 1.7x Jool

6: 0.8x Jool

Color:

1: Green

2: Tan

3: Orange

4: Blue

5: White

6: Roller's Choice

Rings:

1-3: Yes

4-6: No

Moons:

1: 1

2: 2

3: 3

4: 4

5: 5

6: Roll Again

[second Roll]

1-3: 6

4-6: 7

Example:

Size: 1.4x Jool

Gravity: 1.2x Jool

Color: Orange

Rings: No

Moons: 1

Edited October 1, 2013 by SuperWeegee4000
Added example.

[Scoundrel]

A few things:

Firstly, there are a few threads in the mod board where NovaSilisko talks about the difficulty and effort required to make a single planet, let alone a randomly generated one, which might put the difficulty level of this suggestion into perspective. The closest the devs have ever come to talking about implementing something like this was as DLC, which is pretty far down the road. If you can find the old livestreams and IRC logs you can see where the devs stood on that point.

Secondly - and keep in mind that this is just my humble opinion - the effort required to create randomly generated systems could be better put towards improving and expanding actual gameplay elements. If there were hundreds of hours of non-grinding gameplay to be had on each moon and planet in the system, a randomly generated system would be moot as it could be the specific gameplay features that could be randomly generated, which would be just as good as random planets.

Anyhow, just my 2¢.

[rkman]

A few things:

Firstly, there are a few threads in the mod board where NovaSilisko talks about the difficulty and effort required to make a single planet, let alone a randomly generated one,

Procedural, not random. Current planets and moons are procedural, with some manual adjustments.

Procedural saves a lot of time instead of being harder to do, as you seem to think.

The closest the devs have ever come to talking about implementing something like this was as DLC

Squad has said they want us to have lots of places to explore and the Kerbal universe/galaxy will eventually contain many procedural solar systems. They have said it will be added later, not that it will be DLC.

[Dooz]

Well we do have the random craters will this apply to names and the procedural terrain random systems if this was implement it might take up lot more disk space and push computers to there limits

But still i like the idea

[rkman]

Well we do have the random craters will this apply to names

Procedural, not random. It already applies to Kerbal names.

and the procedural terrain random systems if this was implement

It already is implemented for terrain. (procedural, not random)

it might take up lot more disk space

To the contrary: a few formulas and seed numbers needed for procedural content generation takes up far less disk space than models and textures of 100's of hand crafted planets and moons.

and push computers to there limits

No, procedural content generation does not push computers to their limits. See http://en.spaceengine.org/

[Firenexus13]

I like the moon part, but gas giant moons should share the planetary generation, not the moon generation. this will allow stuff like volcanic moons.

other than that, everything is LOGICAL AND SHOULD BE DONE.

[Poryy]

I like the moon part, but gas giant moons should share the planetary generation, not the moon generation. this will allow stuff like volcanic moons.

other than that, everything is LOGICAL AND SHOULD BE DONE.

You just want Mustafar, don't you...

[SuperWeegee4000]

You just want Mustafar, don't you...

He clearly wants Jeb to have the higher ground.

[Poryy]

He clearly wants Jeb to have the higher ground.

I don't get it.

[SuperWeegee4000]

I don't get it.

Watch Episode III again.

[VaporTrail]

Here's a thought toward making procedurally generated systems:

Make the procedure mimic current theories of system formation...

Start with semirandom system mass... (this is going to be an absolutely huge number in terms of tons, but small when rendered in solar masses). Put some variable amount in the system primary (upwards of 99%, depending on spectral class of the system in question). The primary's mass will dictate the Primary SOI. The leftover mass is what is available for planet formation.

Generate the ecliptic. Basically take a radial from the primary's Roche limit, to the edge of it's SoI. Rings that would form Roche tidal breakup about a star actually get blown outward into the system. Generate a semi-random number (bounded high/low) of semi-random points on that radius and call them planetary orbits.

Plot a weighted bell curve over the radius, and assign planetary masses based on where the planetary orbits are positioned relative to the bell curve. Calculate the SoI's of the planets based on the mass assignments. If two SoI's overlap, then either the two affected bodies become one (high probability) one gets gravity assisted into a high inclination orbit (moderate probability and requires a recheck of SoI overlaps...) or become a dual planet system orbiting a common center of mass (low probability). There exists the possibility for asteroid belt formation here, but that's gonna get complicated in a hurry.

Now you've got the mass for each planetary system (Kerbin, Mun, and Minmus would be an example of a planetary system) and their orbital radii. You do a similar operation to the primary system generation, but on the planetary scale for each planetary system... SoI's of moons overlapping result in single moons, or paired moons orbiting about a common center of mass, or possibly even having a moon ejected from the planetary system to become a rogue moonlet. Dual planet systems are unlikely to have moons. Won't say it's impossible... but I'd like to see the orbits on that.

Once you've got the orbits for each body in a planetary system set, and the orbits for the planetary systems themselves set, then you just have to generate the eccentricity of the orbits. The amount of play between SoI's are your limits for eccentricity and elliptical orbits. As long as the SoI's don't get too close, the eccentricity will be fine. Now, inclinations... are any of the moons, planetary systems, etc highly inclined? Semi-random assignment with a heavily weighted probability toward non-inclined orbits would work here. Now just some random starting points along the orbits (random number between 0.00 and 359.99) and the system is set at Y0 D0 H0 M0 S0.

From there determining atmosphere, surfacing and texturing the objects is pretty much all that's left... and part of that is determining the liquid H20 zone of the system primary.

[shadowsutekh]

The idea for procedural systems has already been put forth multiple times, and is in fact on the WNTS list (what not to suggest) under the new stars, planets, asteroids. I can't seem to find the postings on it, but I know it was discussed as a future possibility (hence being on the WNTS list).

Thanks for being active in the suggestion subforum!