[raiguard][2.1.9] Fluid is destroyed when a tank is removed from a pipeline

[nalhedh]

I removed a tank that was part of a group of tanks and the fluid inside the tank was deleted instead of being distributed to the other tanks.

I had a tank that was part of a group of 6 tanks, each of which had around 20k thruster fuel in them, and when I deleted one tank, the 20k fuel inside it was not redistributed to the empty space in the other tanks. I expected the other tanks to go up to 24k (to maintain a total of 120k fuel), but they stayed at 20k, and the pipeline total dropped to 100k.

I reproduced the issue using tanks and water pumps on Nauvis:

1) Put an offshore pump next to any body of water, and connect it directly to a tank. Let the tank fill up to 25k water.

2) Remove the offshore pump.

3) Place a second tank connected to the first tank. Once placed, the water will split in half between the tanks, and each tank will have 12.5k water.

4) Remove the first tank. The second tank will still have 12.5k water in it.

I expected the second tank to fill up to 25k water, since this is how fluids inside a pipeline worked in 2.0 - if you remove a tank, its contents are distributed to connected tanks.

[dewred]

This happens only when you use the deconstruction planner or remote view to delete pipes or tanks. Removing pipes with the player character still has the expected behavior, where fluid is pushed into nearby pipes so long as they are not full.

[nalhedh]

This does still seem like a bug, since it used to work fine with the deconstruction planner before

[Shemp]

This is probably related to how this bug, 132492, was fixed from 2.0.

When using the Deconstruction planner in 2.0, all of the fluid would spread into the rest of the segment if possible, or be voided. So a storage tank marked for deconstruction would always be completely empty by the time a robot reached it, and the event would have no way of knowing how much fluid was inside before.

In 2.1, it disconnects the tank from the segment but keeps all of the fluid. I'm guessing you would like fluid redistribution to happen before that disconnection occurs?

My goal is ultimately to be able to work out exactly how much fluid is being lost due to a deconstruction event. But it seems like fluid behaviour is in a state of flux at the moment.

[nalhedh]

That's interesting, I didn't know this change was related to a bugfix. I do want the redistribution to still happen, yes.

That post does mention that the bugfix is to fix voiding of fluid before the tank is removed, but it doesn't mention redistribution being a bug, and I am hoping that the redistribution is a feature.