Print

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

[UristMcVampire]

Guys I just want this SIMPLE change. I just want trying to go to the edge of the world sending you to the opposite side because planets are round.

[Shonai_Dweller]

Guys I just want this SIMPLE change. I just want trying to go to the edge of the world sending you to the opposite side because planets are round.

Yes. It's planned. More complex than you hope, perhaps, and each world will be different. Some will be flat on the back of turtles maybe, some might be crazy tubes, but yeah, round (not spherical, it's hard with square tiles) presumably wraparound worlds will be a thing.

[NJW2000]

Guys I just want this SIMPLE change. I just want trying to go to the edge of the world sending you to the opposite side because planets are round.

Planets aren't donuts any more than they're squares. Realism ain't so easy.

[Starver]

Guys I just want this SIMPLE change. I just want trying to go to the edge of the world sending you to the opposite side because planets are round.

Except this would (as pointed out) make the world a donut, not a sphere. Walk too far north on Earth (and/or take a boat or whatever is amphibious enough to handle the ice/no ice fragmentation of the arctic) and you don't come up from the south until you've gone all the way south on the other side. The side you could walk east or west to.

If you wrap north-edge to south-edge without doing that (maybe still start south, but not anywhere you could walk east/west to), you're on a toroid. If it's an upright toroid then you don't reach the antipode in east/west direction, if it's a horizontal one you go to a north/south antipode that isn't east/westable. If your toroidal "north" is at an angle other than that, you helically walk around on lines 'parallel' to you, but the same line, until (after N smaller radii, and M larger radii, where N and M could be absolutely anything greater than 1, depending on the angle) you get back to your origin. (Ok, depends on if you navigate by rhumb lines or not, because there's alternate paths that do it even crazier, including recrossing your old path at sone point(s) at a differing angle.)

On a sphere (or oblate spheroid, with caveats) one complete absolute (or somewhere between major and minor axis version) circumference non-rhumb 'straight' line will take you back to where you started and still heading in the same direction. And if you then turn to any other arbitrary direction, you will do the same revisiting only the antipode (or thereabouts, for spheroids) for each and every option.


Anyway, donut or sphere(/oid), the big thing is that 'tiles' towards the polar axis (for a sphere) or 'inner equator' (for a donut) are smaller in width if your 2d map is laid upon a truly 3d object in (flat) 3d space. And drilling down has a similar arrangement (but both axes).

The donut-world space of the classic Asteroids is a flat surface that must therefore wrap in a distorted and non-flat 3d space. If it were laid out in one. (Or special squeezing used to prevent you noticing.) The world-map in the Civilization series (at least the first few, not sure if it continues that way) distorts the map to allow east/west wrapping but expands the polar regions (which exhibit an edge) so that a quick skip around the pole is as awkward as going a similar longitudinal distance around the equator. Which is why building (and controlling!) an Arctic Express railway is very useful, when rail travel is a 'free move'.


Anyway, FYI.

[voliol]

It's only active at the edge. There might be edge lag, but not much else.

To be at a representative scale it would have to be generating millions and millions of Z-levels and close to infinite tiles on the X and Y axis.

According to the wiki, a Large world is 257*257 region tiles. A region tile consists of 16*16 local area blocks, and each local area block of 48*48 tiles each.
If we crunch the number, we learn that a large world is (48*48)*(16*16)*(257*257), or 38 957 285 376 tiles in total.
A tile, by the way, is according to the projectile physics 2m*2m, or 4m².
So the total area of a large world is...  155,829,141,504 m², or to use a more proper unit, 155,829.141,504 km².

This means a large Dwarf Fortress world, with one predominant continent, can be compared with the island of Ireland, and its surrounding sea, as it is 84,421 km², and DF continents tend to take up a bit more than half of the world's area. The Earth is way up there, with its surface area of 510,072,000 km², about 3273 times that of a large DF world. If we consider a quadratic DF world as tall as it is wide, and with the surface area of Earth, it would have to be roughly 2746*2746 region tiles.

I'm not sure what to make of these numbers, but I figured they would be nice to have here.

[voliol]

Wasn’t there some talk about this in one of the myth talks? Or maybe it was a FoF shortly after them being demonstrated. I think the answer was that depending on the myth and parametres, the world could be spherical, torus-shaped, or something akin to balancing on some elephants or entangled in a world-tree. It’d be nice to have the actual reply dug up though for my words here to mean anything.

Also, all this would have to wait until the map rewrite, but luckily that should be the very first part of the big wait, so it’s not that far into the future.

Looking at FoFs from about a year back (though I maybe should have done more), this is what I found on the subject:

I saw in the July Future of the Fortress reply that you can't make a spherical world with square tiles.  There's actually something we make in 3D art called a "cube sphere".  Basically, imagine drawing the grid lines onto a cube, and then puffing out the cube into a spherical shape.

You told something about cylinder worlds, will the game support toroidal planet? Stuff like this thing presented here http://www.aleph.se/andart/archives/2014/02/torusearth.html ?

People addressed most of the issues -- therahedwig's diagram is the main one with seamless loading.  It just doesn't work in 2D.  Changing over to six separate loadable areas would be a step backward and require separate and messy handling.

I don't know exactly how it'll turn out, but code-wise supporting wrapping on either edge is similar enough that it would probably be supported.

Come the mythgen release, do you think we'll get any details on what lies beyond the borders of our maps? For example, if some worlds are meant to be exactly what we see, perhaps floating on the back of a turtle?  While others are meant to be part of something bigger?

Probably?  I have various riffs on the turtle idea in the generator notes, but haven't done any of them in the generator yet.  We had some issues with having civs outside the main world that could send invaders without any ability to reply (which was sort of in a very early version), so the region-within-a-world model isn't as attractive now, even though it is in the list of default gen params.  It'll be especially weird when planar travel comes in and you can reply (if it makes sense) to planar invaders, but not to some barbarians that come in from the edge.  Though there could be room for a kind of infinite home plane as well, where it has to cope with the idea of civs going on and on forever, which is sort of interesting, but mushy.

And a discussion about sphere worlds more precisely: http://www.bay12forums.com/smf/index.php?topic=159164.msg7507635#msg7507635

[PTTG??]

Possible shapes of the planet:
- Spherical
- Flat with a Wall sealing the world
- Flat with an edge where one can fall off...

Possible planetary situations
-Heliocentric Star System
- Everything is circling the planet star system
- Animals that carry the planet (or animals that carry animals that carry the planet)
- World-Tree

Now the first two planetary situations are not so difficult to handle gamewise, but the latter two pose some problems. What happens when one falls off the planet? In an animal carrying scenario the player would actually be on the giant animal, but what if he falls off of that. In a world tree setting, can one climb out of his world to reach others?

Calling it now: The Heavenly Rhesus Macaque stole the world from the gods and is now carrying it through the heavens.

[Detoxicated]

Guys I just want this SIMPLE change. I just want trying to go to the edge of the world sending you to the opposite side because planets are round.

Your suggestion was fine indeed but these forums tend to go crazy on the smallest things as you might know.

[Putnam]

what the people are saying is that the change isn't as simple as suggested

[Starver]

To the "take a cube, blow it up into a sphere" idea, embedded in those past quotes (and probably answered by someone in that conversation, but bears repeating) imagine you're on a tile at the extreme top-right of one of those equatorial faces.

Go east. You're now on the extreme top-left of the adjacent equatorial face. Go north, you're now on a corner of the upper polar face. Presuming it's an option consistent with the last move, now go 'west'. You're on the original tile, but rotated so that local north is right, not up.

(Or when you go north, the reverse route is actually SE, although the same transition from the next corner over at that 'bottom' is SW and S in the middle of each side, even the ones over to what you currently consider as "north". There are loads of ways to deal with it, but all have logical local problems, where at least something like HyperRogue is uniformly consistent in the confusion of direction. In a 'blown cube' you have special tiles that indicate material 'stitches' in the world. When looking for a seamlessly unbounded-yet-finite world, you've failed right there.)

[Giant Ostrich Bone [29]]

I could understand changing the way ALL game/game edges are processed, but if there starts to be variation of world/planetary types between games in generation, I feel like this would be more Spore than Dwarf Fortress territory, hehe.

[KittyTac]

I could understand changing the way ALL game/game edges are processed, but if there starts to be variation of world/planetary types between games in generation, I feel like this would be more Spore than Dwarf Fortress territory, hehe.

Well, world types are a possible variable in the myth release.

[Miles_Umbrae]

It's only active at the edge. There might be edge lag, but not much else.

To be at a representative scale it would have to be generating millions and millions of Z-levels and close to infinite tiles on the X and Y axis.

*bunch of pointless numbers not related to the comment*

My argument was about the simulation of the multitude of flat-world mythology.
According to most of them there is an infinite space of nothing beyond the border/edge of the world .. not to mention how many millions of Z-levels of animal needs to be simulated carrying the world on its back some of those mythologies have, or infinite empty space others explain is below the flat world.

[Rufflikerex]

Why focus on generating a spherical world when we can just work around it? Keep the current world gen as-is, and just generate the land chunks of the opposite side of the map when the player approaches an edge of the world.

[Putnam]

It's only active at the edge. There might be edge lag, but not much else.

To be at a representative scale it would have to be generating millions and millions of Z-levels and close to infinite tiles on the X and Y axis.

*bunch of pointless numbers not related to the comment*

My argument was about the simulation of the multitude of flat-world mythology.
According to most of them there is an infinite space of nothing beyond the border/edge of the world .. not to mention how mention how many millions of Z-levels of animal needs to be simulated carrying the world on its back some of those mythologies have, or infinite empty space others explain is below the flat world.

you do know that you don't need to perfectly accurately simulate everything, yes?