Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[SirHoneyBadger]

I'm just going to bump this to mention constellations:

Groups of stars might be formed together into the shapes of the gods, great heroes, great monsters, etc.

Where a single star might have only a very limited effect, a constellation might have a *lot* of astrological "weight".

Constellations might be formed (hey, I never noticed before, but that group of stars looks like a...) during the life of a Fortress, or even during the life of an adventurer, possibly as a goal (hey, look up there! isn't...is that Steve?!?)--in that a great adventurer might be "set among the stars" by the gods:

Constellations themselves might have souls, sentience, or even physical forms they could occasionally take. Beings could be "born under the sign of", etc.

The waxing/waning of constellations might have an impact on the game. Different species/cultures might have different basic constellations (dwarfs might have the Hammer, Pick, Axe, Anval-while humans might have traditional constellations, like the Big Dipper, the Bear, the Scales, the Hunter, etc. and elves might have the Tree, the Bush, the Shrub, the Pointy-eared Obnoxious Bastard... Goblins might have the Skull and the Phallus (to represent the cycle of death and rebirth), while kobolds could have the Beetle, and other insects, and the Lion (a representation, to them, of death). Water-breathing (amphibious creatures too) might have the Kraken. Even HFS might have their own set of constellations. The Dragon might be universal.

Constellations could be shared among species/cultures, with somewhat different interpretations. Goblins might also have the Kraken, but to them it represents strength in diversity.

The stars would stay the same-interpretations would differ. 

Also, astronomical architecture: Constructions might be built in such a way that they bear astronomical import. Massive "calendar" style megoliths could be built, for magical purposes, or just to help your farmers predict the coming season, and when to plant.

Towers etc. could be built specifically as astronomical observatories. Temples might be built "along divine lines", by aligning them with the sky--which could mean building a temple first, in DF mode, and then adventuring in the name of that temple, until Steve McUrist gets set into the sky, thus fulfilling the prophecy.

Also, other planets might very well exist in DF, but I think Mercury through Pluto are already taken. 

DF planets could be named after DF gods, though, while the sun (suns?) and moon(s--13!) could actually embody divine power, or the "elder gods" themselves, in physical form. Stars could additionally represent lesser gods, or perhaps "divine powers" that don't concern themselves with your particular world.

By the way, how many major gods do we actually have in DF so far? I'm familar with Armok, Tsarnok, Tsathoggua...er Toady the Great,

http://en.wikipedia.org/wiki/Tsathoggua

and probably ThreeToe, in some form or another, but what other gods have actually been established as canon? As in, "powers that exist in every world, despite the RNG".

[corvvs]

Sorry to bring the topic back a few pages, but this is the first time I noticed it.

There's one reason for modeling orbital mechanics vis-a-vis astrology that I haven't seen mentioned - retrograde movement. That's something that's probably more difficult to model with epicycles than it is with (some degree of) orbital mechanics. One reason why the heliocentric model won out - according to general relativity they're both accurate, but accounting for retrograde movement with geocentric epicycles is way more complicated.

It would take somebody who knows more about astrology than the cheesy column in the newspaper says (i.e. not me) to say what the heck a retrograde planet *does* to its sphere of alignment, but I know I've heard the phrase "Mercury is in retrograde the next six days" or whatever a few times.

[Granite26]

Retrograde Linky

SirHoneyBadger:  Nice stuff...  It's important to note that Constelations move at the same rate as the stars.  They are fixed celestial objects, just bigger (take up more space).  Also of note are the disapearing constellations of Krynn.

Corvvs, I agree.  There's a lot of neat things happening with astronomy, and the easiest way to model the complexity of pattern is outright emulation.

[Neonivek]

What could I also add?

There could also be the Hidden Stars which are only visable via a Telescope (which exists at this time... but good luck affording it).

Only astrologers with the equipment will be able to predict events by the hidden stars.

[Vattic]

I have only skim read this topic so far and so i might have missed this already being suggested but I'll throw it into the mix anyway in case it hasn't.

It has already been mentioned that Powers might get points to buy certain months and stars etc, I think it would be interesting to make it so Gods gain extra points through history through the power and number of their followers. It would be interesting to see a God loose followers to another God and have that God defeat the former with the extra power, this might fit better for deamons of course.

[Tigershark13]

Your all forgetting the most important and feared sphere!

The sphere...of the fish!

[SirHoneyBadger]

Well, one culture's god is another culture's demon.

[Felblood]

wilsonns brought the idea of astrology back to the front page, so in keeping with the official guidelines, I'm bringing this old monster back from the deeps.

Arise thread! ARISE!

Does anyone here have good enough chops to build a CPU cheap astronomy model (it has to be able to run for the duration of worldgen) that could keep track of the movements of several planets across a backdrop of stars? We never did find anyone who could explain what needs to be done, last time this thread was awake.

Beyond that. Astrology. What do you think of it? What should it do?

[zwei]

Does anyone here have good enough chops to build a CPU cheap astronomy model (it has to be able to run for the duration of worldgen) that could keep track of the movements of several planets across a backdrop of stars? We never did find anyone who could explain what needs to be done, last time this thread was awake.

Beyond that. Astrology. What do you think of it? What should it do?

If you disregard gravity, then astronomy model is trivial to compute: you have elipse (circle even), movement speed and time.

You do not even have to simulate anything, you can get position of every celestial body at any future. present or past moment by simple formula.

[Felblood]

Are they simple enough to compute several thousand times a second, so that every world-gen event can use them both before(so they can effect things) and after (so they can be effected)?

If so, drop it in a spoiler tag, so that people who don't get woozy at the sight of trig functions can have a look at it.

The way I see it, we need three functions.

One for boring cyclic stuff, like the un and moon and to a lesser degree the stars. The moon crosses the sky every few hours in a basically sinosudial motion. The stars are effectively a single field that moves across the sky every night and moves up and down as the seasons and latitudes change.

This stuff up here is trivial enough I could do it, if you didn't mind me cutting out more of the trig than you're likely comfortable with. It's a waste of CPU cycles anyway, and we only really need to know what times of year a star is visable.

Then you have planets. Planets are just crazy business, folks. You've got to track the apparent location of a thing that is going around another thing (the sun)  that is effectively going around you at a rate of once a day. Then you need to tell the apparent rate of movement, and the direction. I know enough about the problem to tell you that it involves running integral calculus on trig functions (consistently failing to do exactly that is what cost me my college career--that and Rainbow Six).

If I had a pound of graph paper, and restricted all the planets to a 2D plane of movement, I might be able to produce a workable formula for that, but the sight of it would probably make an astronomer cry. I'd like to see this in the hands of a more competent mathematician who could finesse out some of the trig calculations with reasonable approximations, or an astronomy buff, who could tell us what approximations are reasonable. The real reason I asked is that if nobody better qualified is available, I'd be willing to take a crack at it.

Finally, there's comets and other stuff that has to have an elliptical orbit if it's going in.
I have neither the knowledge nor the skills to even comment on this problem.

Now, we can fall back on the claim that DF heavenly bodies don't move like ours, but that seems pretty weak when the mineralogy is so intricate.

[zwei]

You've got to track the apparent location of a thing that is going around another thing (the sun)  that is effectively going around you at a rate of once a day. Then you need to tell the apparent rate of movement, and the direction. I know enough about the problem to tell you that it involves running integral calculus on trig functions (consistently failing to do exactly that is what cost me my college career--that and Rainbow Six).

Dunno, you know [Xplanet, Yplanet] [sin(time)*orbit, cos(time)*orbit]

You also know [Xdwarvenplanet, Ydwarvenplanet] [sin(time)*dwarvenplanetorbit, cos(time)*dwarvenplanetorbit]

And you also know vector where your settlement is going to face (sin(time+angleFromGreenwich), cos(time+angleFromGreenwich)) - vector of its "noon"

Since HomeworldToPlanet vector is easy to compute (Xplanet-Xdwarvenplanet, Yplanet-Ydwarvenplanet)

Your apparent position on sky would be just angle of those two vectors ...

Image:

Spoiler (click to show/hide)

Purple is planet
Brown is dwarf homeworld

on right is example of how it would look to dwarf.

[Neonivek]

Hmm, the dwarf seems to have some planet on his shoe.

Either that or the dwarf is using a planet as a peg leg.

[Felblood]

Okay, so the pictured dwarf is standing on the north pole, and the yellow line separates his field of vision from the unseen stuff that is behind him. The vector perpendicular to that is the direction he is facing. To see this vista from the equator, he needs to look strait up. Is that right?

What does the variable "Orbit" represent. Is it the period of the orbit? That seems to make the most sense, but I don't know jack about astronomy.

This is a lot tidier than what I would have come up with. Does anyone have the formula for the angle between two vectors handy?

A planet is in retrograde whenever the derivative of the angle of those two vectors is negative. Since some or all of the planets may be in retrograde more often than not, this shouldn't have too extreme of a mechanical effect.

To use these equations Toady would need a math library that can approximate trig functions cheaply, but if we follow the traditional convention of dividing the sky into twelve houses, it only needs to be accurate to 1/6π radians of a circle.

If a star can be seen it needs to be within (let's say) +/-90 degrees of the dwarf's viewing direction and of strait up, on both axes. Things that elevate the horizon, like tall buildings or geography could simply penalize the window from strait up.

How quickly do planets move? Can we calculate their position in the star-field once a day, and just report which direction they are going (west like the sun and the stars, or east like some kind of astral rebel), or will they need to be tracked throughout each night?

[zwei]

@Fellblood

"orbit" means orbit radius.

trig functions do not need to be computed - you can just use precomputed tables (array with 360 entries, one for each degree or it equivalent for some subdivision of pi). That is going to be superfast.

You will need to recalculate starfield position pretty much every time you look at it - planets rotate and thus evening starfield is going to be quite different from dawn starfield. I'd suggest some stargazing trip

[Felblood]

@Fellblood

Only one "L." I don't normally correct people on this, but it has been happening an awful lot lately.

"orbit" means orbit radius.

I see much more clearly now. You realize that this yields a system of planets that all have the same orbital period. That's not going to work.

The formula needs to be:

Spoiler (click to show/hide)

x = orbit * sin(time/period)
y = orbit * cos(time/period)

I can't believe I actually remembered that!

trig functions do not need to be computed - you can just use precomputed tables (array with 360 entries, one for each degree or it equivalent for some subdivision of pi). That is going to be superfast.

Even if it's only accurate to (1/6)π, that is twelve entries. Such a table always requires some look up time, which can add up very quickly if you're calling the function a lot. I was just saying that it would be wise to compare options.

You will need to recalculate starfield position pretty much every time you look at it - planets rotate and thus evening starfield is going to be quite different from dawn starfield. I'd suggest some stargazing trip

Let me rephrase the question.

How fast do planets move across the sky relative to the starfield?

If it isn't very fast we can just calculate their position relative to the starfield daily, and then rotate them along with the starfield, over the course of said day. In effect, we treat them just like stars, except they change their position in the star field every day.

If it is too fast for that, then we'll have to recalculate them with trig calls, every time anyone looks at the sky. Moreover, a planet could do something interesting in the interim, so it's entire path will have to be calculated, if it's moving very fast at all.

Speaking of complex math problems, we still need to tackle the issue of detecting retrograde planets during wolrdgen. Having astral bodies in adventurer and fortress mode is a lot more important, but one day, we will want astrology to impact worldgen.