Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: [40d] Multiple Water Anomalies  (Read 1140 times)

Captain Failmore

  • Bay Watcher
    • View Profile
    • http://chairangaem.blogspot.com
[40d] Multiple Water Anomalies
« on: November 29, 2008, 09:37:34 pm »

In my latest fortress, I created a system of fountains. The general design was supposed to accomplish the following:

  • Pump water from a brook up into a system of pipes.
  • Draw the water downward into a reservoir via gravity.
  • Use overpressure to push the water back up through several hollow stone columns, and out the tops of each.
  • Move the falling water out of a drainage sewer with another pump.

It only accomplished the first two satisfactorily.

What actually happened:

  • The system's in-pump moved water from a brook up into a system of pipes.
  • The piping drew the water downward into a reservoir via gravity.
  • Only the northernmost fountains started even though all five columns filled with water. While they're also the closest pair of fountains to the input pipes, this shouldn't have any effect on the system because the pressure in the water should be uniform.
  • The drainage pumps failed miserably to keep up with water removal in spite of their being an equal amount of water withdrawn from the system as there is placed into the system judging by the number of pumps that interact directly with the water input and drainage reservoirs. Flooding occurred shortly afterward.

With no other explanations I can offer, it would appear that the system is somehow generating more water than it takes in, and that overpressure flooding selects tiles the same way path finding does - non-randomly, north and northwest first. Needless to say, as my most ambitious fortress design to date, to see it fail so badly is quite disappointing. If that failure is due to the fluid dynamics not behaving as advertised, it should be addressed.

I can upload this to DFMA if you'd like, though I quit through a three-finger-salute so I could tinker with the design a bit more. Once uploaded it will be activation-ready.
Logged
A HREF="http://chairangaem.blogspot.com">LOLCHAIR ADVENTURES

eerr

  • Bay Watcher
    • View Profile
Re: [40d] Multiple Water Anomalies
« Reply #1 on: November 30, 2008, 12:13:29 am »

sadly, water in dwarf-fort is only a half-way decent approximation,
water moves through teleportation!

a bug concerning u-bends generates infinite water
(the bottom level is under pressure, and this pressure translates into real water)

what are the fountains for anyway?
Logged

Captain Failmore

  • Bay Watcher
    • View Profile
    • http://chairangaem.blogspot.com
Re: [40d] Multiple Water Anomalies
« Reply #2 on: November 30, 2008, 06:58:45 am »

Mist and aesthetics.

The same waterworks could be used for farming, too.
Logged
A HREF="http://chairangaem.blogspot.com">LOLCHAIR ADVENTURES

Captain Failmore

  • Bay Watcher
    • View Profile
    • http://chairangaem.blogspot.com
Re: [40d] Multiple Water Anomalies
« Reply #3 on: November 30, 2008, 09:18:16 am »

Additional experimentation with another fountain revealed the cause of the flooding conclusively; Water was falling onto 7/7 tiles, and instead of being moved to empty space created by the pump, it was being moved to the lip of the fountain. The water levels around the pump were lower, but water was 'piling up' around the fountain. This happened even when no additional water was being added to the fountain, in which case it started to lose water from spillover.

As it is, the water seems to act more like a thick syrup than a thin liquid. Finding a way to adjust the viscosity could alleviate this problem by making it flow and equalize quicker, avoiding pileup. I'll run another experiment to find out why the other fountains weren't working. Seeing as the whole affair has been a game-breaking disappointment for me, I might as well find out why this stuff doesn't work so Toady can fix it instead of whining about it elsewhere.
Logged
A HREF="http://chairangaem.blogspot.com">LOLCHAIR ADVENTURES

Captain Failmore

  • Bay Watcher
    • View Profile
    • http://chairangaem.blogspot.com
Re: [40d] Multiple Water Anomalies
« Reply #4 on: November 30, 2008, 02:03:06 pm »

Even further experimentation confirms that the overpressure water placement prefers western tiles, and then north.

I constructed three fountains connected to a common pressure reservoir. Each was an equal distance from the fill pipe, so I know it's not a matter of being near or far from the source. The westernmost fountain released all of the water even though all three fountains were filled and the water itself had uniform pressure. Not only that, said westernmost fountain released most of its water on the west side of itself. This contradicts the stated mechanics of the overpressure system, which is supposed to pick random tiles around the banks of a pressurized body of water and then 'teleport' pressurized water to their edges. Instead, the random placement of water tiles mimics creature path finding.

Expanding the lip of the fountain by one tile provided a space inefficient but effective solution to the 'syrup overflow'. The fountains no longer shed water. However, any possible adjustments should be made so that water behaves a bit more... watery. (Seriously, it flows very, very slowly even with the frame rate being high, and also while under pressure.) Overpressure tile placement should attempt to be as random as possible; were it not for spillover from that fountain eventually soaking the top of the whole building, my dwarfs could have fished out of the other two pressurized heads.

Edit:

Turns out it does still shed water, just not nearly as bad as previous designs.
Logged
A HREF="http://chairangaem.blogspot.com">LOLCHAIR ADVENTURES

BurnedToast

  • Bay Watcher
  • Personal Text
    • View Profile
Re: [40d] Multiple Water Anomalies
« Reply #5 on: November 30, 2008, 08:01:21 pm »

The 'syrup' effect is not really a bug, as such. From what I understand, water flows slowly to be easier on your CPU, it could be made to flow faster and more realistically but it would grind your fort to a halt.

I wish it was an option in the init - slow flows (slower then now) for people with really bad computer, normal (same as now) for a good balance, and fast (water is as realistic as possible given the 7/7 system) for people who have beasts of a computer and/or don't care about slowdown due to flows.
Logged
An ambush! curse all friends of nature!

Captain Failmore

  • Bay Watcher
    • View Profile
    • http://chairangaem.blogspot.com
Re: [40d] Multiple Water Anomalies
« Reply #6 on: December 01, 2008, 08:25:30 am »

The 'syrup' effect is not really a bug, as such. From what I understand, water flows slowly to be easier on your CPU, it could be made to flow faster and more realistically but it would grind your fort to a halt.

I wish it was an option in the init - slow flows (slower then now) for people with really bad computer, normal (same as now) for a good balance, and fast (water is as realistic as possible given the 7/7 system) for people who have beasts of a computer and/or don't care about slowdown due to flows.

Yeah, it really falls into the realm of preference and system capability. I once thought my computer was slowing down when water started flowing freely because it moved so slow, but then I saw my frame rate was totally unaffected. I have a 2008 computer, so the massive slowdowns I was accustomed to in previous versions don't happen nearly so often in general. The water just flows like molasses - or magma - and distributes pressure just about as quickly, which might keep things running smoothly but causes huge problems for my waterworks.

I concur with the suggestion to put this in the init-settings.
Logged
A HREF="http://chairangaem.blogspot.com">LOLCHAIR ADVENTURES