Developing "Advanced Hydrodynamics"

[Thorfinn]

1 hour ago, Krougal said:

As far as the water wheel, yes and no, you need to be near a stream at least, but you can build a raceway to move the water a bit, or a dam. We should have the capability to build these things in the game. We currently don't. We don't even have rivers & streams.

Oh, definitely. That's what @Katherine K was saying, too, and the whole reason behind the aqueduct discussion. You should be able to create millponds. Overshoot mills should be possible without just digging out everything below the pond level. I just think the source block carryover from Somewhere Else(TM) was a mistake that is kind of hard to recover from. I've seen the FPS death in Dwarf Fortress, so while light aquifer seems like a good idea, it's not very workable in 3D. In 2D, like Terraria, it can work, but it's also easy for that to lag heavily when draining from or to irregularly shaped reservoirs, or even regular shaped ones that extend outside the spawn range. 

@Katherine K makes a good argument for the porosity mechanic proposed here, but the unintended consequence is that spawning anywhere other than sedimentary rock layers is a serious disadvantage. Even moreso than it already is, because you could remove the top sedimentary stone and have standing water, but it wouldn't work that way in igneous. But with that also comes a serious challenge to mining, like you see if you accidentally dig into the bottom of the aquifer in DF.

Someone proposed wells, which sounds like a good idea, too, but requires a sedimentary aquifer.

Rain barrels?

I don't know what the answer is. The elimination of transferrable source blocks was a step in the right direction, but it nerfed too much that is almost necessary. Your only option for irrigated fields is building into lakes? Or digging irrigation trenches 4 long all the way around the lakeshore? Maybe I need to spend a little time going over that Hardcore Water mod, since I've been told it has been expanded from simply getting rid of transferrable source. Maybe however it addressed things resolves the problems.

Edited August 11, 2025 by Thorfinn

[Teh Pizza Lady]

20 hours ago, Thorfinn said:

Iron is essentially an uncountably finite resource. Charcoal is effectively infinite.

Just a small gripe here from a math minor with a BS in Computer Science degree.

Because you cannot have fractional nuggets of iron, it is said that the number of iron you have can be counted, even if doing so would take an unreasonable amount of time, in that you can assign a countable number to each nugget you get from the iron ore. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and so on. If you were allowed to have fractional amounts of iron such at 1.0275, then it would be said to be uncountable because there is no counting number (1, 2, 3, 4, 5, 6, 7, 8, 9 and so on) that can be used to described a fractional amount.

Thus whole amounts are countable and fractional amounts are uncountable.

Iron is a countably finite resource. Charcoal is a countably infinite resource.

The amount of f___s you're required to give is zero.

[Krougal]

8 minutes ago, Thorfinn said:

Oh, definitely. That's what @Katherine K was saying, too, and the whole reason behind the aqueduct discussion. You should be able to create millponds. Overshoot mills should be possible without just digging out everything below the pond level. I just think the source block carryover from Somewhere Else(TM) was a mistake that is kind of hard to recover from. I've seen the FPS death in Dwarf Fortress, so while light aquifer seems like a good idea, it's not very workable in 3D. In 2D, like Terraria, it can work, but it's also easy for that to lag heavily when draining from or to irregularly shaped reservoirs, or even regular shaped ones that extend outside the spawn range. 

@Katherine K makes a good argument for the porosity mechanic proposed here, but the unintended consequence is that spawning anywhere other than sedimentary rock layers is a serious disadvantage. Even moreso than it already is, because you could remove the top sedimentary stone and have standing water, but it wouldn't work that way in igneous. But with that also comes a serious challenge to mining, like you see if you accidentally dig into the bottom of the aquifer in DF.

Someone proposed wells, which sounds like a good idea, too, but requires a sedimentary aquifer.

Rain barrels?

I don't know what the answer is. The elimination of transferrable source blocks was a step in the right direction, but it nerfed too much that is almost necessary. Your only option for irrigated fields is building into lakes? Or digging irrigation trenches 4 long all the way around the lakeshore? Maybe I need to spend a little time going over that Hardcore Water mod, since I've been told it has been expanded from simply getting rid of transferrable source. Maybe however it addressed things resolves the problems.

Aquifer would be great, water ladders for everyone, no bucket required!

Yeah, the FPS death is horrible. I've built obsidian generators in Terraria that lag like hell too.

Dana tweaks actually does make barrels collect rain water.

  

3 minutes ago, traugdor said:

Just a small gripe here from a math minor with a BS in Computer Science degree.

Because you cannot have fractional nuggets of iron, it is said that the number of iron you have can be counted, even if doing so would take an unreasonable amount of time, in that you can assign a countable number to each nugget you get from the iron ore. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and so on. If you were allowed to have fractional amounts of iron such at 1.0275, then it would be said to be uncountable because there is no counting number (1, 2, 3, 4, 5, 6, 7, 8, 9 and so on) that can be used to described a fractional amount.

Thus whole amounts are countable and fractional amounts are uncountable.

Iron is a countably finite resource. Charcoal is a countably infinite resource.

The amount of f___s you're required to give is zero.

Nerd!!!

Edited August 11, 2025 by Krougal

[Thorfinn]

Oh, the example the prof gave way back when was the Sahara Desert. He claimed it had uncountably finite grains of sand. I take it that is not correct?

[Krougal]

8 minutes ago, Thorfinn said:

Oh, the example the prof gave way back when was the Sahara Desert. He claimed it had uncountably finite grains of sand. I take it that is not correct?

It's all kind of moot, because for all intents and purposes, everything in VS is infinite, especially with default worldsize.

You'd die of old age (IRL) before you could mine all the iron.

Now you guys are being pedantic, but it's Traugdor's fault

[Teh Pizza Lady]

12 minutes ago, Thorfinn said:

Oh, the example the prof gave way back when was the Sahara Desert. He claimed it had uncountably finite grains of sand. I take it that is not correct?

They are countable, but I pity the fool assigned to count them.

[Parenn]

23 minutes ago, Thorfinn said:

Oh, the example the prof gave way back when was the Sahara Desert. He claimed it had uncountably finite grains of sand. I take it that is not correct?

How could they possibly be? The mass of the Earth is not infinite, so there can’t be an infinite number of material things on the Earth. Either you misunderstood, or he just meant “lots” and was being very sloppy, or he meant “It’d be really hard to count them.”

Edited August 11, 2025 by Parenn

[Katherine K]

2 hours ago, Thorfinn said:

@Katherine K makes a good argument for the porosity mechanic proposed here, but the unintended consequence is that spawning anywhere other than sedimentary rock layers is a serious disadvantage. Even moreso than it already is, because you could remove the top sedimentary stone and have standing water, but it wouldn't work that way in igneous. But with that also comes a serious challenge to mining, like you see if you accidentally dig into the bottom of the aquifer in DF.

World gen would have to change quite significantly. Since simulated weathering/erosion would be part of the generation step, I suspect a lot fewer places would end up with straight up exposed rock, and most of these that would are already not great for settling.

I don't think a bit of travel to find a suitable river to settle on is unreasonable. I do agree that this doesn't work great with the current respawn mechanic. But then I'm of the opinion that there need to be better ways to set a permanent spawn point anyways.

As for mining, I think it actually needs a bit more challenge. I know collapse risk option already sort of gives you that, but flooding is a real danger in any mining operation. What this really says is we need ways to manage that risk.

I suspect, we'll also end up with a lot more caves partially flooded as a result. Again, that is something you can account for in world gen, adjusting where the carvers are applied and what materials are added.

 

Part of the problem is that a lot of these interactions are complex enough where it's hard to say how it'd impact the game and how to balance it without trying it out. I should have a bit of time I can dedicate to a mod coming up around mid-September, depending on how a few other things land. The core ideas behind aquifiers should be relatively quick to put together. Rivers and free-flowing water are a lot more work, but I think just having porous/cracked rocks able to hold water, replenish with rains, and fill voids is enough to test out key ideas to see if this has legs. If it works, I can play around with rivers/channels later. I have no idea if I'll have time to take it beyond proof-of-concept, but at this point I'm willing to give it a shot even if just to satisfy my own curiosity.

I've done water physics for games before. But not quite like this. Could be very interesting.

[Echo Weaver]

2 hours ago, Thorfinn said:

Someone proposed wells, which sounds like a good idea, too, but requires a sedimentary aquifer.

Hydrate or Diedrate took on the aquifer question. I'm curious how they did it. I started to look at the code and then had a bit of a talk with myself about what I wanted out of my life. I think the answer is not actually writing a well mod right now.

I think one could handle a pretty slapdash aquifer system when the system you're looking to replace is water source blocks. Almost anything is more realistic than that. I even kicked around not having any actual water reservoir, but simply flagging areas that meet the requirements of an aquifer and treating them as one if the user digs down into them with, say, a well-digging axe or something. If it were a mod, I'm antsy about messing with worldgen directly, but mostly that's because I haven't looked at how extensively that breaks other mods or the save itself if the mod is removed.

That proposal is incredibly shoddy, but it seems like it would also have a small resource footprint and, as I mentioned, it's still much more immersive than self-replicating water blocks. It would also avoid miners drowning themselves all the time when they dig into aquifers.

[Krougal]

@Katherine K Yeah, those are a lot of good points.

I agree about traveling to find a river if need be, if there were rivers. That has been always been a huge draw when looking for a place to settle down throughout history.

That is another thing about water wheels and movable source blocks; if the majority of (sane)players are going to prefer to settle down next to a river, it is a non-issue anyway. Really, it just benefits the people who care about aesthetics, since they can put their water wheel where it looks purdy.

Yeah, mining hazards...I'd play with the collapsing if it were better implemented. I have tried it, but it seems too random and I just don't like it. Flooding would be another good challenge, although I don't know if it could really be managed or would matter. 

Other than iron & quartz mines, I feel like most deposits are so small I exhaust them before I move on, so there is no long term to care about flooding. Not to mention I like to flood my mines as it is, better than lighting it up with torches to keep the rot beasts away.

It would require changes to the game to make it matter really. Like if you couldn't mine while submerged, which would be reasonable, but not fun.

[Thorfinn]

16 hours ago, traugdor said:

They are countable, but I pity the fool assigned to count them.

 

15 hours ago, Parenn said:

The mass of the Earth is not infinite, so there can’t be an infinite number of material things on the Earth.

OK, but wouldn't that mean the only uncountably finite thing on earth would be photons at rest? Everything else has a quanta mass, so they are similarly finite, and therefore countable. Why even have the term if it has so little (read "zero") applicability?

Edited August 12, 2025 by Thorfinn

[Teh Pizza Lady]

7 hours ago, Thorfinn said:

 

OK, but wouldn't that mean the only uncountably finite thing on earth would be photons at rest? Everything else has a quanta mass, so they are similarly finite, and therefore countable. Why even have the term if it has so little (read "zero") applicability?

depends on who you ask. Some would say the universe is finite, since matter can neither be created nor destroyed. However in some applications it may be better to just say things are "infinite" for a two-fold purpose:

1. To trigger the nerds

2. To avoid conversations like this.

[Katherine K]

9 hours ago, Thorfinn said:

OK, but wouldn't that mean the only uncountably finite thing on earth would be photons at rest? Everything else has a quanta mass, so they are similarly finite, and therefore countable. Why even have the term if it has so little (read "zero") applicability?

"At rest" photons also have energy and are finite. That's EM contribution to the vacuum energy. It's a somewhat outstanding problem, because upper limit on vacuum energy from cosmology is in disagreement with the lower limit we get from standard model in quantum physics, but both agree that it's non-zero but finite.

Quantum mechanics relies quite heavily on any system being very large but finite, in general. And thanks to General Relativity, that might as well be true. Even if true universe is infinite, only a finite part of it can ever influence any region of space you wish to study. So for any practical purpose, the universe is finite and contains only a finite number of particles at any given time, because there are only a finite number of particle states to begin with.

2 hours ago, traugdor said:

To trigger the nerds

Too late. My Ph.D. research was particle theory. I never defended, because I ran away to California to join a circus game development studio, but I was basically finished with the research itself and had most of my dissertation written. (Thinking I'll finish it remotely. What a naive, young fool I was... )

[Teh Pizza Lady]

2 hours ago, Katherine K said:

Too late. My Ph.D. research was particle theory. I never defended, because I ran away to California to join a circus game development studio, but I was basically finished with the research itself and had most of my dissertation written. (Thinking I'll finish it remotely. What a naive, young fool I was... )

I see my plan has worked.

[Parenn]

15 hours ago, Thorfinn said:

OK, but wouldn't that mean the only uncountably finite thing on earth would be photons at rest? Everything else has a quanta mass, so they are similarly finite, and therefore countable. Why even have the term if it has so little (read "zero") applicability?

Why would it need to have application on Earth to physical things to be useful? The set of integers is infinite, the set of reals between 0 and 1 is infinite, there are lots of useful infinities, but they aren’t physical things.

It’s possible the universe is infinite in extent, but there’s no way to know since we’re limited by the speed of light.

[Krougal]

6 minutes ago, Parenn said:

Why would it need to have application on Earth to physical things to be useful? The set of integers is infinite, the set of reals between 0 and 1 is infinite, there are lots of useful infinities, but they aren’t physical things.

It’s possible the universe is infinite in extent, but there’s no way to know since we’re limited by the speed of light.

I knew it! I'm surrounded by Assholes mathematicians!

 

[Katherine K]

5 hours ago, Krougal said:

I knew it! I'm surrounded by Assholes mathematicians!

Since there exists a natural injective mapping from the set of mathematicians to the set of assholes, namely, from each mathematician to their asshole, the latter trivially implies the former.

(P.S. Yes, I know the reference.)

Edited August 13, 2025 by Katherine K

[Seyko]

Since there are knowledgeable people here, wouldn't it be more productive to calculate water analytically and apply the finished calculations to the coordinates of the world? I don't know, I was a C student!
What happens inside the blocks (about porosity) doesn't need to be drawn and loaded on the video card. And the water that comes to the surface would be described by an equation, for example, Navier-Stokes, and would be interpreted on the game space.
As if you take a figure and apply it to millimeter paper.
This requires knowledge of physics and numerical methods, which is difficult in itself. But would it be optimal for implementation in the game?
I don't know if the Navier-Stokes equation is something that can be useful. But on the Internet they write that it is widely used to solve various problems. In essence, we need to know the direction and speed of movement. Direction is in which direction the cube transfers water. Speed is the amount that is transferred.

Edited August 13, 2025 by Seyko

[Seyko]

On 8/11/2025 at 11:34 PM, Thorfinn said:

I've seen the FPS death in Dwarf Fortress, so while light aquifer seems like a good idea, it's not very workable in 3D. In 2D, like Terraria, it can work, but it's also easy for that to lag heavily when draining from or to irregularly shaped reservoirs, or even regular shaped ones that extend outside the spawn range. 

On the old laptop, FPS in DF was more stable than in Terraria. The video card was the weak point.

[Thorfinn]

15 hours ago, Parenn said:

Why would it need to have application on Earth to physical things to be useful? The set of integers is infinite,

Care to prove that? It is only in the special case of treating mathematics as something other than a model that you can get there. Granted, it's been a wildly successful model, but there's no way to prove the model is correct at infinity.

But that's completely irrelevant to the point I was making, which , from context was iobviously the world of VS, and that, unlike  charcoal, there is currently no "sustainable" source of iron, it is finite.

[Krougal]

4 minutes ago, Thorfinn said:

Care to prove that? It is only in the special case of treating mathematics as something other than a model that you can get there. Granted, it's been a wildly successful model, but there's no way to prove the model is correct at infinity.

But that's completely irrelevant to the point I was making, which , from context was iobviously the world of VS, and that, unlike  charcoal, there is currently no "sustainable" source of iron, it is finite.

Now one of you nerds smart people needs to go make a calculator. Some function of the world size and ore density parameters along with the generation constants should be able to give you a rough approximation of just how much iron your world will have.

[Katherine K]

2 hours ago, Seyko said:

Since there are knowledgeable people here, wouldn't it be more productive to calculate water analytically and apply the finished calculations to the coordinates of the world?

Pure analytical solution of system of equations actually take way longer than iterative methods, because you end up having to find roots of nth-order polynomial, and for 5th+ order that's still iterative and much, much slower.

Analogous problem that I have way more experience with is solving mechanical systems with constraints. Your standard games physics engine. The equations of constraints between multiple objects either connected to or in contact with each other, is actually very similar to the conservation of volume constraints for fluid flow, and the flow itself is very much analogous to the undetermined multipliers in the mechanical system - that is, the constraint forces. In other words, if you have two buckets connected by a tube, amount of water is total momentum, flow is the force, and the parameters of the tube are the constraint between the two.

In a games physics engine, you could construct the system of equations, solve it, and just apply external parameters as inputs to get the solved flow. Problem is, if you invert analytically, it's super expensive, and the moment anything changes in the system, you have to throw it all out and start over.

Iterative approach is very similar to how you solve linear systems via row elimination. (In fact, that is one of the methods, but not commonly used, because of some optimizations available in other methods.) You basically run through the matrix, and try to balance all of your constraints by pivoting the matrix. On each pass, you drastically reduce the difference between current approximation of the solution and the true inverse. So much so, in fact, that most game engines that use this method don't bother running multiple passes per frame. So you basically check for collisions, collect contact points, apply contact cache (that is, any contacts that persisted are initialized to last frame's step) and you run one iteration of the solver. Then you use that to compute forces and update positions.

With a fluid solver, you don't even need to check for block changes. You run that via block update notification. If a block was placed or removed, update relevant constraints and re-initialize relevant flows if needed. Run one iteration of the solver, and update the flows. After a few frames of that to let things settle, if water hasn't spread to new blocks, you can put solver to sleep and keep updating the flows by simply updating the inflow/outflow conditions. This would be analogous in a game's physics engine to a crate coming to rest on the floor. Once the gravity and support forces are balanced, you no longer need to keep updating them every frame. In these cases, the simulated object goes to "sleep" and is woken only if new contact points appear or some external force is applied.

[Bumber]

On 8/11/2025 at 1:38 PM, traugdor said:

The amount of f___s you're required to give is zero.

"Give ZERO f___s. Give INFINITY of them" - Statements dreamed up by the utterly Deranged