A Possible Model of a Kerbal Galaxy

[Danthalios]

I've seen a lot of threads where people ask the devs to implement interstellar travel - but the question is, how would interstellar travel actually be implemented in KSP?

There would have to be a galaxy for the stars to be in, and this model explains how I think the devs might implement such a galaxy (of course we won't be seeing the galaxy any time before KSP 1.0). I whipped it up on Paint, so I apologise if it's a little bit lacking in the quality side of affairs.

The bigger star on the bottom arm is the Kerbal system. It is surrounded by other stars, as are other large stars on the other arms of the galaxy, all of which are orbiting that big white thing in the middle - a supermassive black hole to study for scientific purposes (or just chuck your Kerbals into).

Each of the coloured dots may represent different stars: for example, a red dot could indicate a red (super)giant, a blue dot could indicate a blue (super)giant, a white dot a white dwarf, and a yellow dot a yellow star (like the Sun). The light blue dots close to the centre of the galaxy are supposed to be Neutron Stars - which could be explored or even mined to collect super-dense materials to research in career mode. The green rings around the galaxy are the orbits of Rogue Planets - i.e, planets without a parent star. These, if implemented, could be quite interesting, and along with neutron stars, could potentially be mined for new, advanced materials for research. Procedurally-generated wormholes could appear occasionally in the galaxy also - I'm thinking gateways across the galaxy to make pan-galactic travel easier, and possibly (with a sufficient tech level) even be modified to create portals to other stars.

Now, this is nowhere near official, and the devs may have something entirely different in mind to my model, but this is how I think a possible galaxy would work. Perhaps interstellar travel could be achieved via exploiting heliopause (the point where the SoI of one star ends and that of another begins), or as I mentioned above, wormholes - or something entirely different altogether.

Discuss!

[Diche Bach]

I wonder if an elliptical galaxy might not be easier to implement? One game I'd want to point out (to other players and the Devs) as being worth a serious look (just to see how such things have been successfully modeled in other games) would be Distant Worlds. Their fanbase are very devoted and I have to agree it is probably the best 4x space game out there at this point. To not learn from what this game does well would be folly.

[Rockhem]

Galaxies have millions of stars. That would be insanely taxing on the computer.

[innovine]

Galaxies have millions of stars. That would be insanely taxing on the computer.

I loved Frontier back in the Amiga days. That had a whole galaxy on a floppy disk (1.4mb)

[rkman]

For various reasons the number of stars in the KSP galaxy will probably have to be relatively small and the stars relatively close together.

But that need not be very unrealistic because in reality galaxies come in sizes ranging from a fraction the size of our own galaxy (dwarf galaxies, which have no distinctive shape), up to many times larger (giant elliptical). And even in our own galaxy are regions where the population density of stars is significantly higher than it is in the vicinity of the Sol system, up to several per cubic lightyear.

[itsme86]

Galaxies have millions of stars. That would be insanely taxing on the computer.

The game wouldn't have to do anything with stars that aren't currently "active".

[ixolate industries]

A cloud galaxy might be the best option as they don't contain nearly as many stars as an elliptical or spiral galaxy you could conceivably have a mini galaxy with just a few stars to begin with the possibility for modders to add more. This would also be much quicker to load than a full scale galaxy would.

[Diche Bach]

The largest number of stars per galaxy that Distant Worlds does (last time I played it) was 1400. You do notice some performance issues at that size, and obviously it isn't a replica of the Milky Way. Didn't know about these dwarf galaxies so I guess it isn't 'unrealistic.

But in any event, it is a very satisfying and interesting gaming experience that succeeds in conveying the vast distances. FTL is of course part of the game, though I don't remember the details.

ADDIT: Another idea that is worth considering by modders and devs: Initially do not worry about modeling the entire galaxy. Instead just model the immediate neighborhood of Kerbol. Assuming that the way interstellar travel is implemented is 'realistic,' then for the first phases of a campaigns interstellar exploration and expansion, it will only be the immediate neighborhood of Kerbol that will matter anyway. This map of the immediate neighborhood of Sol could serve as a good model in that it only contains the 55 stars within 14 light years.

Edited August 9, 2013 by Diche Bach

[Trainsfan]

The game wouldn't have to do anything with stars that aren't currently "active".

He's right. Think of how many stars Spore has. I'm guessing it doesn't have an advanced Physics engine like KSP, but KSP would necessarily render those stars. We aren't talking a legitimate galaxy like in real life, something smaller, more "Kerbal". I'd say 1,000 would be a good number, 1/3 of which have planets/asteroids. (still Wishful thinking, I know).

[daniel l]

Planetary habitability would not be possible in such a galaxy plus it would make the skybox obsolete

[MarkusA380]

Have you ever heared of Space Engine? It is the whole universe to be travelled through!

It works like the "Procedural Craters" on Mun, but it is procedural galaxys, stars, planets, moons, asteroids, landscapes usw...

a link: http://en.spaceengine.org/

It has a higher resulution for the planets than google earth...

[1of6Billion]

Yes! I LIKE what I read here.

I would like to see a galaxy full of user created star systems. With the current moddability, that would make an unlimited supply of interesting exploration goals.

[Trux]

Planetary habitability would not be possible in such a galaxy plus it would make the skybox obsolete

The skybox was never meant to be a completed thing and needs replacing at some point down the line anyway

[TimMartland]

I think a cluster/cloud galaxy of a few stars (a few dozen tops in vanilla game) is a realistic aim if the devs were to do this. The KSP team is incredibly imaginative and are very able, but to complete an entire galaxy (even a small one) would be a difficult feat for anyone. I think that a large part of any interplanetary gameplay would be community-based, E.G lots of moddable planets people could create if they wanted extravagant, complex worlds, or even a system where high-quality systems and worlds created by the community could be refined and added to the game if they were good enough to reach the standards of planets already implemented

[Diche Bach]

I think a cluster/cloud galaxy of a few stars (a few dozen tops in vanilla game) is a realistic aim if the devs were to do this. The KSP team is incredibly imaginative and are very able, but to complete an entire galaxy (even a small one) would be a difficult feat for anyone. I think that a large part of any interplanetary gameplay would be community-based, E.G lots of moddable planets people could create if they wanted extravagant, complex worlds, or even a system where high-quality systems and worlds created by the community could be refined and added to the game if they were good enough to reach the standards of planets already implemented

The system used in Distant Worlds is to randomly generate galaxies. User gets to specify a bunch of parameter settings, then the galaxy generates (complete with lots of planets and moons, black holes, as well as alien homeworlds from whence they expand).

Obviously hand making a galaxy of 1400 star systems would be prohibitive for even a very large and filthy rich studio. But defining the rules for randomly generating such galaxies that are fun to play in and include as much or more fine grained detail as does KSP has already been done. The game runs great on my machine. If the Distant World makers can do it, I'm sure Squad can do it; although whether it would work with their current solar system centric engine I don't know.

Just to be clear; I'm not like some Distant Worlds fanboy. I bought the original game and played it a good bit, but set it aside for various reasons. I have not followed up with any of the several addons or expansions. But it is a good game, and I bet a lot of you guys would like it a lot. Plus, Squad should hear about it if they don't know about it. Always smart to learn from others. I would guess that even if you don't actually reverse engineer their code, just seeing what other spacey games can do is bound be edifying.

Have you ever heared of Space Engine? It is the whole universe to be travelled through!

It works like the "Procedural Craters" on Mun, but it is procedural galaxys, stars, planets, moons, asteroids, landscapes usw...

a link: http://en.spaceengine.org/

It has a higher resulution for the planets than google earth...

Whaaa?? FREWARE? How? Why?

The video with the Enya song is amazing, and I'd be very keen to check this out. But that little emblem about "100% safe and clean" actually worries me given it is "Free."

Whats the catch? Why would anyone give away something that looks that awesome?

Edited August 12, 2013 by Diche Bach

[Seshins]

there is no catch, it is actually amazing

[Rockhem]

even dwarf galaxies have at the minimum thousands of stars.

[Diche Bach]

even dwarf galaxies have at the minimum thousands of stars.

Actually I thought it was billions as opposed to hundreds of billions in the full-sized ones.

[markonaut]

First post here, how about modelling around an open cluster such as the Pleiades? It could be kept realistic then and not so computer taxing? No fun with black holes though!

[golegogo]

one idea i had was having 30 galaxies and having only 10 having planets. and to save on memory loading each as you enter its sphere of influence. also allowing people to make mods with star systems would add alot of systems fairly fast

[itsme86]

You don't even have to guess. There are certain sites on the Internet that let you search for vast amounts of data. One of these is http://www.google.com. The very first sentence from the resultant very first link that comes up when you search for "dwarf galaxy" is:

A dwarf galaxy is a small galaxy composed of up to several billion stars, a small number compared to our own Milky Way Galaxy's 200–400 billion stars.

[Ittiz]

I loved Frontier back in the Amiga days. That had a whole galaxy on a floppy disk (1.4mb)

You took the words right out of my keyboard.

[Guernica]

Here's my idea on implementing a galaxy-esque environment into KSP:

I like the idea of procedurally generated star systems. The names of the stars could come from a pre-defined list (not that hard, I don't know how to program, and I've made callable lists in Python). The names of the planets around those stars could follow the name of the star such that around a star named Rigel, planets would be named Rigel a, Rigel b, so on and so forth, until you landed on said planet, plant your flag, and are given the option to rename it (or something like that).

The next big road block seems to be the amount of memory tons of star systems loaded into KSP would take. Perhaps it would be possible to code such that once one exits a given star's SOI, only the star itself would be rendered, perhaps with a list of the planets in a right-click drop down menu.

Basically there would be maps at varying levels of detail. One map view, like the current one, would show everything within a given star's SOI. The other map would be a Galactic view, showing only the stars themselves, with a clickable/hover menu detailing the planets around those stars.

As for the actual design of the galaxy, I like the idea of keeping with the current trend of modelling after our own Solar system, but with it's own character. So a Spiral galaxy, even if it were only 100 stars and 20 light years across would be more than sufficient to keep even the most ardent player occupied for a very long time. Not to mention, if this tiered map view were possible to the galactic level, there is nothing to say someone couldn't then mod a higher order map, creating nearby galaxies.

If I knew how to program/mod, I would start on this project right now.

Edited August 13, 2013 by Guernica
Grammar Errors

[markonaut]

Thinking about it, if my maths are right, at 100000x time warp, it would take around 6 days to travel 20 light years. New time warp settings would be required.

[ZenithRising]

100000x time warp would not be even close to enough to get to another star.