Pufts always starving

[Sturm58]

Hi guys,

I have a big problem with pufts. They live in a polluted oxygen air, +-3kg, temperature is right, tamed, not overcrowded and yet they are always starving. This is most probably related to the game speed and amount of calculations that happen. My guess is the game optimizes by skipping critter activities like eating. This breaks the game for me. I have killed a ton of creatures to fix the problem but it does not help. Any ideas what to do?

(ignore the missing rancher icon)

[SharraShimada]

Your assumption is correct. If the computer is already struggling with the performance critters will suffer first.

You may try the Mod "Fast Track". It solves many problems (but its not perfect)

Other than this, there are only two solutions. Reduce gamespeed, or get a better CPU. Sorry.

[asurendra]

Reduce calculations. Sweep all debris. Fill all unused space with cells. Remove unnecessary critters.  Check  your dupes pathing

[gabberworld]

i call this as game bug or unity engine bug, because when you todo game restart it may go away. its like they to lose some information's from game objects 

[Atro]

Multiple layers of game login are running out of sync. I'd consider this a software defect. I've also looked for a solution to this problem because I like critters but haven't found a real solution. It's sort of gamebreaking because critters are essential for many resources like slime or coal.

If the game is running too fast critters will lag behind. In extreme cases they die. You can easily test this by spawning thousands of critters in sandbox mode until they start lagging. The critters will each wait for their turn. Sometimes you can even see the order they are processed while the game is paused as they snap to a direction they want to move to because it runs entirely decoupled from the rest of the game logic..

While investigating this I've found that if you slow the game down below the slowest speed with mods the problem goes away. You can also pause buffer to let the critters catch up with the rest of the game. This only works if the critters are actually lagging behind. Constraining critter movement to one tile didn't help. The only thing that helped is slowing down game speed or removing critters.

[gabberworld]

3 hours ago, Atro said:

removing critters.

hmm. well people not buy game for start remove them. if there some animation sync issues or clock then this should be resolved from developers some different way.

i notice one thing tho. at some-point they start stop fly around when this constant hungry starts happen

[Atro]

16 hours ago, gabberworld said:

hmm. well people not buy game for start remove them. if there some animation sync issues or clock then this should be resolved from developers some different way.

i notice one thing tho. at some-point they start stop fly around when this constant hungry starts happen

I really like critters I don't want to suggest removing them from the game. Critters are mostly inactive as it is. They only do something every couple seconds like find a spot to move to or eat. If the game can't update all other critters in the world before the first one is ready to to perform its next action again it'll begin to lag behind because while the critter waits it is still losing calories. At that point the critter will start metabolizing less material than it potentially could and eventually it'll eat so infrequent that it starves. The problem is worse of course when you run the game at faster speeds. Currently you can remove some critters so the remaining can update in time or you can slow down the game.

I think it's a bug or maybe a difficult software problem. A quick way to fix it would be making critters increment calorie consumption in fixed bulky increments only when it's their turn to perform an action or maybe even have critters metabolize mass based on the duration since their last action to compensate for the lag.

 

 

[SharraShimada]

On 2/5/2023 at 11:18 PM, gabberworld said:

i call this as game bug or unity engine bug, because when you todo game restart it may go away. its like they to lose some information's from game objects 

Nope. Imagine the system as your desk in an office. Everone loads off their work onto your desk and you cant do it all in time, so it piles up and up... Then you leave the office and come back, and the desk is clean again. The huge pile of work is gone, and you start again with a clean desk. But in time all of the others will start loading off on you again. 

If the computer cant handle all of the tasks in real time, the engine queues them up, until the game decides to ditch less important stuff in favor of getting the list not any longer (because it cant geht any longer at a certain point).

The devs decided to cancel critter stuff first, because it wont matter if they are starving for a few seconds without any update. The problem is: The queue is not getting solved in time its really clogged up. So they may starve to death. 

You may solve it by letting the game run in 1x speed for a few minutes, until the queue is empty. Unfortunately you have to guess. There is no visual indicator if its empty, or if it just does not matter at the moment as much as a few seconds before.

[gabberworld]

2 hours ago, SharraShimada said:

Nope. Imagine the system as your desk in an office. Everone loads off their work onto your desk and you cant do it all in time

 

yeah im totally agree with you. but now comes this thing. multicore as that what it was destined todo first place.

[Farsight]

Multicore isn't a magic wand. Many of the game's systems depend on each other, so can't run in parallel. 

As you explore and build more and more, you eventually surpass your CPU's ability to update everything. The only solutions are:

1) Buy a better CPU

2) Hope for optimizations in a future update

3) Simplify the asteroids to reduce what the CPU has to do.

#3 is usually the easiest option.

I do wish there were a Gravitas Device called the "Planetary Shifter", that we could use to blink an entire asteroid out of existence. The regolith and Moo asteroids have to be hard on the CPU, and I barely ever touch either of them. I'd use a Gravitas Death Star on both of em in a heartbeat.  

[SharraShimada]

3 hours ago, gabberworld said:

yeah im totally agree with you. but now comes this thing. multicore as that what it was destined todo first place.

I´m pretty sure, you´ve been told already, multicore is NOT the solution for every problem. But i´m also positive, you´ve always ignored it. So i´m done here. 

[gabberworld]

11 hours ago, SharraShimada said:

I´m pretty sure, you´ve been told already, multicore is NOT the solution for every problem. But i´m also positive, you´ve always ignored it. So i´m done here. 

well, i to know that some stuff is kind off impossible add as multicore. unity engine not even allow todo that for some stuff

[SharraShimada]

22 minutes ago, gabberworld said:

well, i to know that some stuff is kind off impossible add as multicore. unity engine not even allow todo that for some stuff

Thats wrong too. Unity is modular. Its right the engine wont come in a state to do so on its own for everything. But you can always write your own plugins for such things, or buy them from third party. 

[gabberworld]

6 minutes ago, SharraShimada said:

Thats wrong too. Unity is modular. Its right the engine wont come in a state to do so on its own for everything. But you can always write your own plugins for such things, or buy them from third party. 

you know what that 2d game reminds me. sonic or supermario or what ever similar game

[Tytan]

On 2/4/2023 at 7:21 AM, SharraShimada said:

Fast Track

link?

[SGT_Imalas]

 

 

[Sanchozz]

I'm curious.
Pufts starve because they eat several times less often than expected ? or because they are stuck and stand still and do nothing ?
in the first case, this is quite easy to fix,  just need to increase the suction rate in proportion to hunger. I would even make a mod.
in the second case, it is quite difficult. perhaps only the developers could fix it.

[JRup]

13 hours ago, Sanchozz said:

Pufts starve because they eat several times less often than expected

I'm experiencing exactly something like this, pufts feed but the amount taken in is hardly enough to balance against hunger. I'm on an old 10k+ cycle save and what can currently be defined as an old computer. (CPU is a 4770K ...)

The save does not have a tame slickster ranch so I couldn't compare it with the only other critter species that consumes gas.

Edit: Loaded save file and observed that calorie intake amount was insufficient after critter has fed.

[Farsight]

Pufts eat 50k of gas a cycle. If their room has 1k of gas pressure, they're going to need to eat at least 50 times a cycle if they only pull from one tile, or 10 times a cycle if they suck air like a gas pump (5 tiles).

Logically, you can probably reduce their starvation by increasing the pressure in their room.

If a mod could increase the number of tiles they suck up to eat, that could reduce their starvation as well, at the cost of seeing them vacuum out a chamber each time they eat.  

[JRup]

11 hours ago, Farsight said:

Logically, you can probably reduce their starvation by increasing the pressure in their room.

It's a nice thought, however, I do try to keep pressure at least at 3kg/tile going up to 4kg/tile so I can reliably assure you pressure is not a factor.

This puft only gained about 5 calories after inhaling...

[gabberworld]

13 hours ago, Sanchozz said:

or because they are stuck and stand still and do nothing ?

that one is correct. they stand at same place and doing nothing, that activates the hunger

[asurendra]

Pufts breaths are extremely inefficient. They consume only few hundreds grams, and this amount is not depend on air pressure. So when game reduce they feeding rate starving comes. Slicksters has the same illness, but less intensive as they need less gas per cycle

Increasing that amount may help, but it has some limitations. With hundreds critters and weak PC sometimes puft doesnt have a chance to make inhale for entire cycle... 

We need critters AI rework for sure

[Atro]

1 hour ago, JRup said:

This puft only gained about 5 calories after inhaling...

I have observed a Puft which isn't hindered by lag under ideal conditions will inhale frequently enough to easily remain above 1,080 calories. Then stop inhaling until it's at 1,080 calories again. Below 1,080 calories it'll drop about 3 then gain 5 calories, like you mention. So it seems to me the Puft in the image is slowly starving due to lag unless it started out starved already.

[Sanchozz]

I made a mod to increase the gas suction rate depending on hunger. Now we need to test it.

[gabberworld]

1 hour ago, Sanchozz said:

I made a mod to increase the gas suction rate depending on hunger. Now we need to test it.

add your mod to the github, i use epic,  i cant take it from steam