Jump to content

Strange timing


Recommended Posts

There is something wrong with creatures feed timing, on x3 speed they feed much slower then x1

Новый проект.gif

Cause of this, sometimes they can't eat their 100 kcal per cycle, and start starving with full feeder O_o

I thing starving pufts in hi-pressured dirty oxygen is the same reason, creatures just do not hawe enough time to feed, starving is faster.

Dev's, can you compensate this?

Link to comment
https://forums.kleientertainment.com/forums/topic/142727-strange-timing/
Share on other sites

Just now, network_flows said:

I mean the total number of critters over the map. If you clear the map, the pacu will eat quicker.

Have you watch the main post video? He basically record the same setup, which basically in the whole map the amount of critters remain the same, and even though, in x3 speed, the pacu eat slower than the x1 speed.

@Tranoze, of course we watched the video, such statements seem to do more harm then good.

 

The point here is that oni is deliberately reducing the number of cpu cycles spent on creatures when fps suffers. This leads to this behavior either when you have many creatures at low speed or few creatures at high speed. So in a way, the observed behavior is by design and thus not a bug.

Nevertheless, I'm quite sure that the devs want to minimize the discrepancy. Obviously, in an ideal world both simulation speeds should give the same results.

11 minutes ago, Jann5s said:

@Tranoze, of course we watched the video, such statements seem to do more harm then good.

 

The point here is that oni is deliberately reducing the number of cpu cycles spent on creatures when fps suffers. This leads to this behavior either when you have many creatures at low speed or few creatures at high speed. So in a way, the observed behavior is by design and thus not a bug.

Nevertheless, I'm quite sure that the devs want to minimize the discrepancy. Obviously, in an ideal world both simulation speeds should give the same results.

Indeed, so i hope devs can compensate this with boost kcal per eating depend of time speed, or something of that.

If a designed behavior leads to two different results, how is that not a bug? Also a video may show the problem but doesn't really help. A save file and full pc specs may help more. That is why a bug tracker exists where you can upload all info and describe the problem. 

There it is, i made new experiment with different count of critters. For correct, i clear all other critters on asteroid.

 

2 pacu.gif

300 pacu.gif

This means big farms get more unstable than more you hawe critters. This is sad.

So if i have farm with 200 pacu, 150 hatch, 200 pufts, some dreccons etc, then all they must starve cause there is issues in game mechanic or what, you expect farm must contain 2-10 creatures and no more?, devs can you comment this please? May be has a way to avoid that "checking all critters" every time after feeding?

17 hours ago, Tranoze said:

Have you watch the main post video? He basically record the same setup, which basically in the whole map the amount of critters remain the same, and even though, in x3 speed, the pacu eat slower than the x1 speed.

Sorry I didn't mention that. The stuck time I mean is REAL-WORLD time, not in-game time.

Suppose your CPU can only handle(or status check) a certain number of critters every sec, but there are N critters on the map. It takes real-world time to complete the loop for all critters. If real-world delay is denoted as t=f(N), then in-game delay is ~n*f(N) at nx-speed, or 0 when paused. This may account for the result.

To decrease in-game delay, you can reduce N(kill critters) or reduce n(switch to 1x game speed)

 

 

11 hours ago, saraconor said:

There it is, i made new experiment with different count of critters. For correct, i clear all other critters on asteroid.

 

2 pacu.gif

300 pacu.gif

This means big farms get more unstable than more you hawe critters. This is sad.

So if i have farm with 200 pacu, 150 hatch, 200 pufts, some dreccons etc, then all they must starve cause there is issues in game mechanic or what, you expect farm must contain 2-10 creatures and no more?, devs can you comment this please? May be has a way to avoid that "checking all critters" every time after feeding?

No, starving only happens to frequent eaters like pufts and algae-fed pacu. Hatchs, dreckos and seed-fed pacus eat much more each time and at a lower frequency, they should not suffer from starving.

1 hour ago, gabberworld said:

and now the big question, what is your fps at 1x and 3x speed

as at 3x speed fps usually drops, and that makes also times longer

more about timings issues info you can find here

 

I have 80 fps on 1x and 60 fps on 3x speed.

On 8/21/2022 at 9:23 AM, saraconor said:

There it is, i made new experiment with different count of critters. For correct, i clear all other critters on asteroid.

 

2 pacu.gif

300 pacu.gif

This means big farms get more unstable than more you hawe critters. This is sad.

So if i have farm with 200 pacu, 150 hatch, 200 pufts, some dreccons etc, then all they must starve cause there is issues in game mechanic or what, you expect farm must contain 2-10 creatures and no more?, devs can you comment this please? May be has a way to avoid that "checking all critters" every time after feeding?

What your experiment proves is that the code is optimized to skip processing in order to allow for useful game play.  Back before Spaced Out, you could slow your machine to a crawl by having too many critters in your farm.  Also, when you run at faster speeds, it means that there are more processing ticks per second, which can quickly bottleneck the CPU/RAM pipeline.  The solution, then, is to skip some tasks once a certain bandwidth threshold is reached.  There is an archived post somewhere that I wrote a long time ago, showing how open-room hydrogen/oxygen generation would affect game play vs doing the same thing in a closed room where there was less dispersal of the gasses and heat.

On 8/22/2022 at 1:57 AM, gabberworld said:

and now the big question, what is your fps at 1x and 3x speed

as at 3x speed fps usually drops, and that makes also times longer

more about timings issues info you can find here

 

Cool. I hadn't seen a recent one since the debug code was removed.  Thanks for pointing this out.

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
  • Create New...