This will be of practical interest to anyone who is stuck watching aquatic wildlife sitting in one spot on the ocean floor, of theoretical interest to those interested in the details of fluid behavior, and entertaining to those who like to hear about oceans being drained.
So I was recently trying to get a walrus breeding program going, because seriously, what's cooler than walruses?
After a bunch of fiddling around with worldgen to make a map with a lot of arctic oceans, I managed to get a site with some walruses.
The walruses immediately marched into the ocean and found a nice spot on the ocean floor to sit and breed for a year. They just all sat on their tile and didn't move. Meanwhile, no other wildlife entered the map while the group of walruses was there.
I determined to capture the walruses by any means necessary.
I mined beneath the ocean, put an up staircase in the tile directly beneath the unsuspecting odobenidae, and then dug DIAGONALLY to make a hall full of cage traps leading to a drain into an aquifer. The plan was that I would send a miner to breach the ocean floor by channeling the floor out from beneath the walruses, causing them to fall into the hole and then be washed by the local water flow into the cages.
Unfortunately, it didn't work. Walruses can swim. They stayed in the same tile even with no floor beneath them.
So I raised the stakes. I dug out the corners in the walrus trap to allow water to flow through it orthogonally. This time I imagined that the entire ocean would drain into the aquifer, and then the last bit of flow on the ocean floor would wash the walruses into the hole and thence into the cage traps.
Well, you have not lived until you have drained an ocean into an aquifer. I have drained an ocean using the fire bin method, but it's totally different.
When you drain an ocean using a fire bin, only a few tiles full of water can be destroyed in a given frame. In each frame, the bin destroys all the water near it, and a finite amount of other water teleports to the now-vacant space, ready to be destroyed on the next frame. The ocean drains gradually, and demonstrates the order in which tiles are evaluated for water motion, since it's always the first few remaining tiles that end up moving on each pass.
When you drop an ocean into an aquifer, you learn something else about the way water movement is evaluated: the water tiles take turns, and only certain ones are evaluated on each frame. You learn this because the aquifer can destroy arbitrarily many tiles of water per frame; tiles can pressure-path into the aquifer and vanish without occupying any space to prevent the entry of more water. And it takes a few frames for the ocean to vanish; they don't all go on the first frame, and you can see the order they go in. You can see it very clearly because each individual frame takes about 30 seconds to happen, as thousands of water tiles path to their demise.
Anyway, it still didn't work. After the ocean finally drained, the water entering the map from the exposed map edges kept the entire ocean floor level, as well as the trap hallway, filled to 7/7. Any further water landing on top of that would of course find its way into the aquifer and vanish, but any vacancies were immediately filled. And this, of course, does not cause any flow that would push walruses into my trap. No good.
For my third attempt, I dug a tunnel around the map edges. I actually went two levels below the ocean and dug a lot of ramps to remove the walls above so I could avoid the endless damp stone messages. I then built drains from this tunnel into the aquifer. Then I went up a few levels and built a ring of suspended floors exactly over the tunnels, themselves separated from the ocean only by floors.
I pulled the collapse lever and again waited for the ocean to drain, and oh how it drained. With the continuous drain wrapping around all of the oceanic map edges, newly-entering water would always find its way into the drain instead of refilling the middle of the map. Finally I reopened the walrus trap. With no pressurized water to instantly refill vacant tiles, the walruses immediately fell into the hole, and were swiftly washed into the cage traps by the water's diffusion.
I had captured the entire herd of walruses.
When I had done this, other wildlife finally began to enter the map. My breeding program now has large herds of elk and muskoxen as well as walruses, and a few wolves. I also have two female polar bears.
Eventually, however, another group of walruses entered the map and settled into place on the ocean floor, thereby blocking all other wildlife activity.