Will the planets actually move?

Or perhaps balance it by having such radical orbit changes cost so much delta-v that to fly off uncatchably-fast requires so much energy consumption or huge numbers of engines and leaves too little room left on the asteroid to put enough energy generation to make it worthwhile.

I would not want viable strategies to include stuff like asteroids escaping the system or moving giant planets significantly, but I would prefer if that balancing was accomplished not by artificial limits. If the engine supports it to begin with, it'd be nice to leave it in for the awesome and fun factor.

Another solution is to declare a body hurled into the abyss as lost with everything on it, after it exceeded some defined distance to the star.
You could use this to destroy said body ; it would take less energy (less thrusters) to do that than throw it into the star IIRC, but it would also take more time. And ramming it into something may take less energy and less time than both anyway.


Boring technical stuff :

Note that you don't need a full n-body simulation to have a good enough simulation. Games like Kerbal Space Program or Pioneer Space Sim, for example, give each body its own 'sphere of influence'. So if you are near the Earth, you are only affected by the Earth's gravity. If you close to the Moon, you enter its sphere of influence, and you are now only affected by the Moon's gravity.
Of course, the Moon itself is affected by the Earth's gravity, the Earth is affected by the Sun's gravity...
It is not perfect, but it is good enough for a rocket (explosion) simulator like KSP, so it should be good enough here.

A ship or an asteroid orbiting a planet is small enough so its influence can be neglected, as well as a moon around a gas giant or a planet around a star. For double planet/stars/gas giants (those are predicted to exist out there), you just do a simple 2-body simulation, with each object having its own sphere of influence, in addition to the bigger sphere of influence of the system itself. Those would allow for many orbits impossible IRL ; you could forbid them, at least in the system generation, but realism isn't the first objective here anyway.
The problem arises with triple+ systems, which do exist IRL, but they are rarer, so you may be fine ignoring them, or making them successions of 2-body systems (for example, a system composed of a star and another 2-stars system).

Oh, and in the system generation, you may want to add the Roche limit, just to have a nice (mostly) moon-less place where to put rings around planets. It wouldn't have to affect a player-moved body.


Also, physics-based systems would allow for nice game mods, like having a fully developed system attacked by fully developed planets/battle stations from outside the system.

This is no more a problem than giving a tank the command to drive across an ocean. It's an RTS game, you aren't flying flying the thing with a joystick. You are selecting a new orbit, not controlling the thrust vector manually.

And yes the planets will move.

Awesome.

That's a very good point. This isn't like Osmos, where you manually thrust and hope to move your sphere somewhere you want it to go, fighting physics every step of the way. In RTS games, there's such things as pathfinding, and the units fairly well manage their own motion once you give a command. I suppose in this case, an entire planet is a unit! :mrgreen:


I'm very curious how the interface will represent your orbital options. I think that's going to be fascinating. Perhaps where moving the mouse will give a preview of the new orbital path, and not allow the user to choose invalid orbits. The game would do all the heavy lifting to figure out what kind of thrust and direction is necessary to alter the orbit. Then I suppose you'd lock in your selection and the planet moving device / engine would appear as a buildable frame which you'd assign construction units to, or something.

Small catch: If the planets rotate, its going to cause the engine's position to be constantly changing, and thus make it impossible to apply consistent thrust to move it anywhere. The only answer to that is to have planets suddenly stop rotating when moved by an engine.. or not have them rotate to begin with... or to have another means of thrust... And if there's no rotation at all, there goes your day/night cycle and solars would only work on the day side, etc. I don't know to be honest. I wonder if they have a solution for that.


Great to hear that planets do indeed move though!!

Well one solution is to only fire the engines when they will add thrust in the correct vector, another is to have thrusters which fire in such a way to stop the body from rotating.

Yes those are useful solutions. Probably would have to be a combination of both. One to get the rotation aligned so that the engines even have a chance of hitting the vector, and the other to do the heavy moving. Still, it would likely not be sufficient because you'd only have the right vector for a few seconds out of each rotation cycle, rather than a constant thrust.

Maybe... they don't actually intend for "big" planets to be movable at all, but only the little ones like asteroids and such. Just like in the video. And maybe those don't rotate at all. After all, moving a big planet would be ridiculously difficult compared to a small asteroid. So you'd then chuck the smaller ones at the bigger ones to blow them up, but not likely chuck the bigger ones at each other?

Another solution is thrust vectoring of some kind.

Do we even know if planets rotate?

Jupiter is the fastest rotating planet in the solar system at 10 hours. So even in a game lasting a couple of hours the rotation wouldn't be very significant.

I'm not sure if they rotate to be honest, but if there is a sun as a light source, and spherical planets, then either they rotate, or one half of the planet will always be night and the other always day, which would be disappointing, especially since solar farms with day/night cycle would be a cool dynamic.

As for the duration of the rotation cycle, well that's irrelevant when it comes to gaming. Look at Minecraft. The full day cycle is 20 minutes long. That's just right for a game.

There is the orbiting poll
viewtopic.php?f=61&t=35709

unless we just say "screw real physics" and tweak the game mechanics so everything (no matter how tight it's orbit around the sun or how distant in space it is) has a reasonable orbit and rotation for an average play session or multiplayer match. So say a planet like Mercury would have a 15 minute year and a 3 minute day. And an object like Pluto would have a 4 hour year and maybe a 1 hour day.

Oh yeah and make the planets really really close together compared to the real world. Space is just way too huge compared to how big stuff is to be fun to look at.

As far as I'm concerned, there will either need to be some fluff which says the commander experiences time at 100x normal, or there shouldn't be any orbital simulation calculations wasting my cpu time.

To our perception the motion of bodies orbiting a sun are unnoticeable. And if a game is to last only a couple of hours, only things like local asteroids and moons should move noticeably. And even then only because you are now 4 hours into the game and you recall that the moon used to be 60 degrees further away, not like you are actually noticing the moon move itself.

To be fair though, perhaps the commanders should experience time at such a pace. It would explain the rapid building and unit construction, and without it we should probably also scrap the moving asteroid thing, which was kinda the primary gimmick of the game.

Guys, clearly the game isn't going to be "realistic" in its depiction of orbital speed, or even scale. Real planets are ridiculously large compared to even the biggest planets you'd likely see in this game. Scale and time are both elements that need to be greatly distorted for a game to be even remotely fun. Its ok, really...

This isn't to say that the game can't have some realistic elements, such as physics and the like. They just won't be based on the actual physical constants we deal with in real life.

I haven't heard anything to suggest one way or the other if there will be rotation. I was just giving some examples because its interesting. The slowest rotation in the solar system is Venus at 243 earth days, which is longer than its orbit around the sun.

I would bet that the engine can support the solar system out to the Oort cloud at full scale (though I do not know what mechanics it would simulate). The only question is how powerful of a system you would need for that.

Sorry for bringing this game up once again (it is a great game, and is relevant often times in these conversations), but Sins of a Solar Empire doesn't have any kind of "orbit" system, other than a visual "spinning" effect on the planets (entirely visual/graphical, not gameplay related), and I've never felt that it cheated me in anyway as far as depth of gameplay. In other words, I don't think the planets need to move - it would only complicate things, really. However, if they can find a system of orbiting that really works then more power to them.

I think the real question is, "How fun would it be to play with absolutely real distances and orbit time?" And the answer is not very fun. Put everything close together (distance and time wise)

This is the one thing I hate about the orbiting discussion. It always dies in about three steps.

1. One good suggestion, "Hey wouldn't it be cool if planets rotated and had orbits, some neat game play could come out of that, and it's just kind of cool."

2. One bad idea gets intertwined with it, "Yeah! and Saturn could have a day night cycle of 400 real days, and orbit the sun every 30 real years!"

3. And finally a valid criticism of the bad idea which sinks the good part like a pair of cement shoes made out of pure failure, "No! That's dumb, just lock everything in place Sins Of A Solar Empire style."

Guys just adjust everything to what would actually be fun to play with during the amount of time it takes to play the game.

You forgot that last part where I basically said the same thing:

We both just vaguely said, "They might be able to make orbiting work somehow", which I guess isn't terribly helpful

Keep in mind that in Solar Empire, you don't actually land (or build or fight) on the planets. They are purely for decoration, where in PA, the planets are the place where everything happens.

I'll grant one thing about Solar Empire though - the gravity well in that game may very well be what Uber has in mind for the "Orbital Layer" of PA. Where the units run around and fight each other but don't actually orbit anything. If you had a unit orbiting (moving around) the planet in the orbital layer, as oppose to being able to come to a relative stop, that I think, would be insanely difficult gameplay, however cool it may be to see. Sad, yes, but again even I (who prefers realism in games far over arcade style) have to say that the complexities of real life orbital mechanics will need to be severely limited and simplified in order to create a fun game.

In PA, we've heard that the planets will move. This is great. In Solar Empire, they did not move, but the strategy of the game was heavily dependent on the moving of forces along a node based grid. Perhaps travel between planets will occur Solar Empire style, but with the added complexity of moving planets, and space lanes being more or less dynamic. Add to that the full scale of what we already know about terrestrial RTS combat, and you've got a full plate it would seem. Toss in the ability to send planets into collision courses, and you'll have people wetting themselves in excitement.

fkya! :lol: