Print

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

[NW_Kohaku]

Except each cart only carries 2/7 magma, still involves fluid movement at both ends, and causes heating calculations as it moves through tiles.

A pump stack shouldn't have any space for random flow. A proper bottom-up pump stack keeps the stack in same state (7/7 magma) each tick, avoiding constant heating calculations (the cause of the FPS hit.)

That's not how I've seen it play out, pump stacks are a real FPS hog due to random flow because you need at least a 3x3 area on each floor to avoid problems, whereas minecarts don't calculate heat just for a minecart with magma traveling through it. (In fact, dwarves can pick up magma minecarts with their bare hands, and you can actually use minecarts practically as buckets for magma.  If you don't mind the magma moving slower, you can honestly just set up dwarves to drop a cart on the edge of a magma pool with tracks, push the cart through it, grab the cart on the other side, have them carry it to the surface, and then dump it in any arbitrary location via track stop designations.  Again, slow, but I find it the easiest way to handle things like surgically obsidianizing streams in the current game, since it's a lot faster to designate than a magma "spigot" to pump magma onto a stream.)  The reason is that, much like buckets, it's not actually a fluid water or magma inside the cart, it's an item called water or magma much the same way that alcohol is treated in the game, and it's only after being dumped out that code is triggered to specifically turn water or magma back into fluids again. 

The 2/7-at-a-time issue is almost never a factor.  You can't Boatmurdered, but you can push things up to one cartload per 11 ticks if you just stuff enough carts on the tracks, which is FAR more magma than you'll ever need for virtually any other application.  I get my magma forges running and have a nice magma moat just using one or two carts at 2/7 every couple hundred ticks just fine.  Even if you did need to expand magma throughput, cart elevators are fairly compact and efficient, and you can just put an arbitrary number of parallel elevators in place to up the flow.

This is moot in a suggestion thread. The need to pump magma up isn't going away.

If this is moot, then all the pump and liquid exploits can't be counted upon, either...  Toady went fairly far out of his way to specifically code mechanics for magma to be cart-able.  A big part of the reason that nobody thinks water exploits are going anywhere is because the mechanics that exist were a sort of compromise Toady had to make, and making a better system would require far more complex (and processor-intensive) mechanics to calculate something like rate of water flow (and fluid mechanics are already an FPS hog) to regulate power generated by waterwheels, and in turn, rate of flow from pumps.  Likewise, the unusual behavior of minecarts is based upon the properties of the tile-based system and his attempts to make sub-tile movements possible.  Toady knew he was coding carts that were blind to most of the track positions, and allowed a cart to enter a tracked tile from a direction that had no track leading to it, but had no other recourse because there wasn't an elegant solution.  Saying that minecarts "won't last forever" is something said when you don't understand the underlying logic, and I find is generally said because they don't WANT to understand the underlying logic.

What I expect to actually happen long-term, is that Toady will end up adding pipes that pressurize fluid based on the number of pumps in the system. Pipes themselves are already on the dev list.

Possibly, although I haven't seen such.  I also seriously doubt you can pressurize magma in any meaningful way (at least, not using hypothetically medieval-level machinery that would bend to the pressure before the magma would) since Toady already specifically exempts magma from ordinary pressure mechanics.  I tend to get the idea that pipes are more for allowing dwarves to walk over flows of liquids without drowning risks, or allowing flows of liquid to move without requiring building whole aquaducts filled with 7/7 water.

[anewaname]

With either the pumpstack or minecart method, the FPS hit is not that important unless you need a constant flow of the material. The amount needed might be the most important factor in choosing the method.

[Bumber]

That's not how I've seen it play out, pump stacks are a real FPS hog due to random flow because you need at least a 3x3 area on each floor to avoid problems, ...

Looks like you can get away with 3x1. I would think a bottom-up pump stack would replace the magma before random flow could occur, but I have only secondhand knowledge of pump stacks.

... whereas minecarts don't calculate heat just for a minecart with magma traveling through it. (In fact, dwarves can pick up magma minecarts with their bare hands, and you can actually use minecarts practically as buckets for magma.

Not true. I've had wooden wheelbarrows disintegrate while carrying magma carts sufficient distances. I can't imagine what the bare-handed dwarves are suffering through.

If this is moot, then all the pump and liquid exploits can't be counted upon, either...

Rivers and underground lakes remain, however. You're left with logistics of pump stacks, or logistics of rollers (or FPS forbid, dwarves.) Not sure if magma pistons will remain or not, given a cave-in rewrite is due.

Likewise, the unusual behavior of minecarts is based upon the properties of the tile-based system and his attempts to make sub-tile movements possible. Toady knew he was coding carts that were blind to most of the track positions, and allowed a cart to enter a tracked tile from a direction that had no track leading to it, but had no other recourse because there wasn't an elegant solution.

This doesn't explain an increase in velocity.

I also seriously doubt you can pressurize magma in any meaningful way (at least, not using hypothetically medieval-level machinery that would bend to the pressure before the magma would) since Toady already specifically exempts magma from ordinary pressure mechanics.

Well, it currently can be pressurized using pumps, for whatever reason.

[NW_Kohaku]

If this is moot, then all the pump and liquid exploits can't be counted upon, either...

Rivers and underground lakes remain, however. You're left with logistics of pump stacks, or logistics of rollers (or FPS forbid, dwarves.) Not sure if magma pistons will remain or not, given a cave-in rewrite is due.

However, pump stacks are going to be impractical if you don't have access to tremendous numbers of waterwheels exploiting DF's lack of Conservation of Energy.  Rollers individually aren't a major power hog, but if you rely upon them in numbers, it's just as bad as a pump stack. 

Not true. I've had wooden wheelbarrows disintegrate while carrying magma carts sufficient distances. I can't imagine what the bare-handed dwarves are suffering through.

I don't remember that from my playthrough, but it's possible there was an update, or that my dwarves were using wheelbarrows, since it's been a year or two since I last tried...

This doesn't explain an increase in velocity.

It's on a "down ramp".  Again, the problem is that the game can't look at where a cart has been to make logical sense out of the path it is traveling.  If a cart is on a tile with tracks and a ramp that "points downwards" in a certain direction (that is, there is a ramp on the tile, and at least two directions the tracks travel on the tile, such as "North-South", and all but exactly one points to a wall, then the game recognizes it as a ramp that points in the direction that doesn't have a wall), then velocity is added in the direction of the "downward slope".  The game doesn't check for how a cart enters or where it came from, if a cart is on a tile with a track, and passes onto a tile with a track, it counts as being along contiguous rails, even if you're traveling east on strictly North-South tracks. 

To use the "straight track" example:

Code: [Select]

z+0    z+1    z+2    z+3
░░░░░  ░░░░░  ░░░░░  ░░░░░
░░░░░  ░══░░  ░▼▼║░  ░░░▼░
░║░░░  ░▼░░░  ░░░║░  ░░░▼░
░║▼▼░  ░▼░░░  ░░░░░  ░░══░
░░░░░  ░░░░░  ░░░░░  ░░░░░

░ : Wall
║,═ : Track/Ramp
▼ : Down Ramp (empty space)
Let's say the cart starts from a stop at the bottom left of the z+0 floor.  It's on a ramp that "slopes north", and generates northward velocity.  Until it leaves the tile, it builds momentum, and from a stop, can generate a pretty decent amount of velocity.  When it enters the next tile, the middle-left of the Z+0 floor, it's on a "slopes south" ramp, and without the checkpoint bug, would lose momentum as it traveled (but still have just enough momentum to get over the hump).  The game does have the checkpoint bug, however, so this tile is actually almost entirely skipped, because of the change in direction of the ramps it is on, which means no momentum is lost.  Either way, this "impulse ramp" design would work, however.  (Even if "impulse ramps" were "fixed" and needed "real" ramps, this design would still work, since there's no situation where the cart would go south on the "slopes north" ramp on the Z+0 floor. You just need to dig out the wall for the ramp to lead to a dead end.) Because of how carts travel, the cart is rolling "downhill" half the time, generating momentum to go uphill... but the "downhill" is always on the same z-level, while the uphill actually goes up a z-level. 

After making it up the ramp, the cart would be on the top-left of Z+1. The northward momentum it has will make it crash into the northern wall, but that's irrelevant.  Because it's on a tile that "slopes east", it starts building up eastward momentum from having a zeroed momentum, and carries on from there. 

The transition from a north-south to a east-west set of tracks is something the game isn't really capable of understanding.  As some of the testers have pointed out, the only time track direction matters is for ramps, curves, and guiding dwarves.  Carts ignore anything about tracks except for curves, ramps with "slopes", and the existence of a track at all. 

Well, it currently can be pressurized using pumps, for whatever reason.

It isn't really "pressurized" by the pumps, the pumps simply push magma out of their way to add new magma in.  (On a game code level, pumps will add fluid to their target tile, but if it is full, search for the next nearby tile to add their liquid into, so long as it's part of a contiguous channel on the same floor.)  Magma pushed like this only fills its own Z-level, and never follows the rules for emulating hydrostatic pressure that water, which involves pushing water "upwards" when a body of water is detected as being higher than a place the water can flow into.  (Whose code functionally teleports water from the highest Z-level with water into the area found with a void.)

[Bumber]

However, pump stacks are going to be impractical if you don't have access to tremendous numbers of waterwheels exploiting DF's lack of Conservation of Energy.

You get 9 pumps/z-levels per wheel. Hardly tremendous. Windmills can support 2 or 4, if you get wind.
Minecart rollers need axles and gear assemblies to transfer power to each other, which puts them at only slightly less power. (On the other hand, you can't use glass like with pumps. But I digress.)

It isn't really "pressurized" by the pumps, the pumps simply push magma out of their way to add new magma in.  (On a game code level, pumps will add fluid to their target tile, but if it is full, search for the next nearby tile to add their liquid into, so long as it's part of a contiguous channel on the same floor.)  Magma pushed like this only fills its own Z-level, and never follows the rules for emulating hydrostatic pressure that water, which involves pushing water "upwards" when a body of water is detected as being higher than a place the water can flow into.  (Whose code functionally teleports water from the highest Z-level with water into the area found with a void.)

Due to the overly technical nature of your description, I can't tell if you're aware that pumps will teleport magma to areas on the same z-level, even if it has to go through a U-bend.
(See: Magma landmines. Around 1:51 is a good view of pumps and magma.)
For all intents, it's pressurized by pumps.