Newtonian Aurora - Galactic Map Question

[Dutchling]

I think the flattened 3d map is the best/easiest and still realistic way to do a real star game.
Instead of starting in the Sol system in a real star game, you could give people to option to start in a planet in a random system (like someone a few posts back said). This would make it possible to still use real stars, without reducing the replay-ability. You might see a bunny-NPR race living on Earth if you manage to find the Sol system :).

[PTTG]

"Naturally, these higher psuedo-dimensions, while similar to Einsteian or Cartesian space, have a few distinctions.  Most obviously, while three dimensional volumetric objects can exist in them, real-space three-dimensional maps, when projected into "hyper-space" appear to be two-dimensional.  The fact that they allow faster than light travel is merely a practical application of this fascinating theory. "

So yeah, you could simply rearrange a real space map onto a 2D area with some kind of algorithm based on the real distance to the star; a star that is above the orbital or galactic plane but close might end up being adjacent to a star that's below the galactic plane but just as close to Sol. . .  but that's just the "mysteries of Hyperspace" for you.

Now, that all said, a random galaxy map would be interesting and allow for a lot of control over the generator.  I'd say that on balance, I'd rather have the random galaxy with controls, but there's no reason why having a premade "real" galaxy map wouldn't work- or even having a galaxy construction set for other premade galaxies.

[Antagonist]

Star Control 3 to me had a very nice way of doing a 3D map, basically rotateable 3D points.   This does however require the ability to display 3d points, often suggesting GL or DX rendering (though not required with some math. )

Alternatively I like Brian's 'lanes' idea, though it would basically be predefined connections between stars, since on the fly recalculations might create some gameplay confusion, such as two stars further away on xy being closer than another star.

But on a pure gameplay I would suggest jRides' suggestion of flattening and basically just throwing away the Z.   Yes, you lose some 'realism', but not that much and it is MUCH easier to handle.

I don't like a 3D tactical map however since all the interesting stuff in a star system is in a single plane anyway.   Adding a third dimension might allow spacecraft some extra tactical freedom, but it would all be in vacuum and not around stuff.   Theres no reason to go up or down except to pursue ships who are up or down.

[Zed 6]

If you have to use 3D, then a flattened version. I'd vote for real stars and then random stars. Just my opinion.

[UnLimiTeD]

Just randomly take out a third of the real stars and add a very small amount of randoms, and it'll not always be the same.
Though I much prefer a random start anyways, what do I care for real starts then?
Earth is boring, I already live on it.

[waresky]

2300AD or Megatraveller Star map are the best for RPG and strategical effort..more addictive for me.

My 2 cents

[Sirnik]

IMHO, a semi-random approach might prove useful.  I do totally agree that if a player chooses a non-Solar start the generation of system must be random.  However, for a Solar start it might be feasible to use the following method:

Select several stars (say, two or three dozens at most), that are either the closet to Sol or the most prominent upon the sky and place them at their appropriate places, projecting them on the galactic plane.  Then divide the whole 360 circle in sectors roughly corresponding to constellations, again projected on the galactic plane.  Then randomly place systems giving them names according to sector - just like the innumerable Glieses at present.  Maybe the placing should not be random, but rather have a fixed distribution - set a direction for the Galactic core and star density growing in that direction, while the opposite happens in the other.

[Aldaris]

Here's a solution. Use the real positions of stars, and then flatten them onto a randomly aligned flat plane, with some maths to keep them from cluttering, and voila. Real stars, real-ish distances, and different each time.

[Father Tim]

Random map, random start please.  I have no objection if you want to build a 'real galaxy' map for those that want to play on it, but I'm bored with the Sol System and would probably never use it as a starting location.  I'd be SMing an empire on to a chosen body somewhere else on the map.

I'd be renaming virtually every system as soon as I explored it - I can't stand having to deal with Wolf 359 and Wolf 352 and Wolf 456 and Wolf 7 something something or remembering if I wanted Sigma Draconis Rubens or Sigma Draconis Luteus, or Tau Draconis Minimus, etc.

I see nothing wrong with a 2D map - it's what almost every game uses, and since the galaxies I'm familiar with are essentially 2D spirals, I find it an adequate approximation.

[Garfunkel]

Flatten the 3d map to a 2d. Should not be too much of a problem. I'd prefer real stars over random stars, though it is true that for replayability, some randomness is good.

[bean]

My vote is for both in 3D, though I know that's not exactly on the table.  I'd prefer real stars, though, all in all.

[HKZ8]

Realistic 3D map, as closely to the link as possible.
hxxp: kisd. de/~krystian/starmap/

Combat could be 2d, even if full realism would ultimately mean 3 Dimensions.

[UnLimiTeD]

And an equal amount of extra required processors and programming years.

[HaliRyan]

If you wanted to keep the 3-dimensional distances you could still use a flattened 2d map or a Mercator projection. Just preserve the 'vertical' position of each system in the background and the math calculating travel times will work fine. Then the easiest way to intuitively display the relative vertical positions (that I can think of, others may have better ideas) in a 2d mode would be a simple color coding. Something like red shift for 'higher' systems and blue shift for 'lower', so traveling red to red or blue to blue with the same horizontal distance would take less time red to blue.

[Steve Walmsley (OP)]

If you wanted to keep the 3-dimensional distances you could still use a flattened 2d map or a Mercator projection. Just preserve the 'vertical' position of each system in the background and the math calculating travel times will work fine. Then the easiest way to intuitively display the relative vertical positions (that I can think of, others may have better ideas) in a 2d mode would be a simple color coding. Something like red shift for 'higher' systems and blue shift for 'lower', so traveling red to red or blue to blue with the same horizontal distance would take less time red to blue.

It is an interesting idea and I have considered something similar. It wouldn't be too bad if local space was a relatively thin disk but it is spherical so it could get crowded and hard to read the map.

One other option I am considering is to flatten real stars to 2D by changing the x,y coordinates of real stars so that without the z coordinates they are still the same distance from Earth. The ratio of x to y would be the same. This would preserve the distances and general direction from Earth but could significantly change the distances from known stars to each other. As the vast majority of players would probably be familiar with the star names, a smaller proportion would be generally aware of their distances from Earth and a very small proportion would be aware of their distance from each other, this might achieve the illusion of a real stars map without affecting the suspension of disbelief for most players.

Steve