v0.31 Talk:Water

Tried dropping various creatures in various places in arena mode. Everything survives 1z drops unharmed. 2z is damaging. 3z is more often than not lethal. Water appears to cushion falls. I dropped things from the top of the arena into the 3z deep end of the water and they weren't injured at all, though they quickly drowned due to not being swimmers. Also tried dropping people in 3z of magma. Magma seems to provide less cushioning than water. And then they burn. Suggest making irrationally large (200-300z or so) drops into various depths of water to test cushioning effect. Rkyeun 04:44, 5 April 2010 (UTC)

It's worth noting that layers of water in previous versions did not cusion your fall so much as the depth of the water didn't increase it. I.e. the drop damage was calculated by the amount of air you fell through. Is that a possibility here? --Eagle0600 05:24, 5 April 2010 (UTC)

Contamination

I've built a very large reservoir (currently 30x25x4 approx) on some ground that had mountain goat blood on it. There is also a small amount of mountain goat and dwarf blood on the area around the reservoir. On overflowing the reservoir (which is fed from the top by a "clean" source), I get blood EVERYWHERE the water flows, seemingly on evaporation of the water. There's now far more blood spread around than the creatures had originally... anyone seen similar? --Nimblewright 21:21, 8 April 2010 (UTC)

I did some arena tests with blood and water. apparently adding a 7/7 tile of water on top of a pool of blood creates more pools of blood which are pushed out to the edges of the water as it moves. This behavior may be a bug. There is an open ticket about the subject on the bug tracker here with my full results. http://www.bay12games.com/dwarves/mantisbt/view.php?id=296 - Doctorzuber 01:31, 9 April 2010 (UTC)

Verifications of old behavior

• U-bend test from a river fills to z-1 as expected
• U-bend test from a murky pool fills to z-1 as expected
• channeling directly from a river temporarily has no flow, but soon becomes naturally flowing as expected
• screw pump to a dead end one level above a river generates natural flow as expected
• Diagonal connections neutralize water pressure as expected
• Diagonal connections do not prevent the spread of flow as expected
• Reactor#3 built as a perpetual motion test -- behaves as expected

• some flow artifacts at the bottom of both U-bends -- this was always a bit quirky, but should run more tests
• screw pump from a murky pool in a 1x18 channel which doubles back to the murky pond generates reliable flow -- possible change in flow mechanics
• Screw pump from z-2 one level below the river, pushing into a 1x100 channel generates flow only near the end of the channel where the water is still filling. Behaves as expected, the previous test appears to be a fluke.

I'm running through a battery of tests here, I will update this with my results. The one surprise so far is the murky pool test. In 40d a similar test would quickly become saturated with 7/7 water and only indicate flow at the end of the channel. need to test that more to make sure this isn't a fluke. I will continue to update this as I verify more things. Once I have a bit more information I'll start updating the water pages. Doctorzuber 01:26, 9 April 2010 (UTC)

Ran a new variation of the long channel flow test using a level drop this time to break the natural flow effect from the river. This time I got the expected result of many tiles of still water with a small amount of flowing water near the end of the channel where it had not yet filled. This does however mean I'm using pressurized water this time which might be worth a bit more investigation. This also demonstrates the simplest form of water teleportation which is what is believed to be why flow of this nature fails to appear. Doctorzuber 19:53, 10 April 2010 (UTC)