[2.0.11] Thruster swapping Fuel and Oxidizer (mixing/fluid) (MR)

[Zaflis]

When I changed the recipe from electro science to the lightning collector it stopped working. When I tried to move it down so I could add a tank, the pipe fills with holmium solution...

I could be wrong but if i guess correctly then the machine's internal fluid buffers do not reset when you change recipe. Additionally it will use different fluid connections to each recipe, so at the point you resetted it, it could have already been connecting to wrong fluid that came in and you rotated after?

So far it is working as intended for them, but who knows if they will still change it some way if they find time out of other million issues You can fix it easily be emptying the fluid pipe contents with the trashcan button.

If you change a recipe for building that uses fluids:
1) Disconnect all pipes.
2) Set new recipe and remove the building and place it back down with cut and paste, Ctrl-X, Ctrl-V.

In the future please don't change recipe like that, make new buildings for each task just to make things easier for yourself.

[SamsaraRazor]

When I changed the recipe from electro science to the lightning collector it stopped working. When I tried to move it down so I could add a tank, the pipe fills with holmium solution...

I could be wrong but if i guess correctly then the machine's internal fluid buffers do not reset when you change recipe. Additionally it will use different fluid connections to each recipe, so at the point you resetted it, it could have already been connecting to wrong fluid that came in and you rotated after?

So far it is working as intended for them, but who knows if they will still change it some way if they find time out of other million issues You can fix it easily be emptying the fluid pipe contents with the trashcan button.

If you change a recipe for building that uses fluids:
1) Disconnect all pipes.
2) Set new recipe and remove the building and place it back down with cut and paste, Ctrl-X, Ctrl-V.

In the future please don't change recipe like that, make new buildings for each task just to make things easier for yourself.

The issue here isn't the EM plant whose recipe was changed, it was the electrolyte machine. The issue persists even with bot rebuilding. Try it yourself please, it is a little frustrating to be fighting to be believed. However, if you can see an obvious error I am not seeing, please let me know after you've tried and identified it.

Yes, it is possible to fix using the delete button in the pipes. After 800h in regular factorio, I am not used to it, and I don't think a building outputting the wrong fluid into empty pipes is intended operation. Somewhere I believe there is a crossed wire, as you are also suggesting, but I believe it is outside of my control, and it sits inside certain fluid buildings. It is as if there is a hidden pipe inside the building that outputs fluids arbitrarily. (I know it is ruled by logic, being a computer game, but the logic is either unintended (bug) or not logical to me, the user)

It also doesn't explain my second edit, how does a building whose recipe does not accept holmium solution, output holmium solution when you rotate it and place an empty pipe on its output?

[Muche]

What might be unintuitive is that if all stashed liquid is hidden in machines, trying to place a pipe won't show mixing, nor is blocked from placing it.
The only case it's shown and blocked is if one pipeline is all hidden and one pipeline has a piece of pipe showing the different fluid.
(Tested based on Spaceage_fluid_mix_fail.zip, with 2.0.15, gotta retest with 2.0.16 with the fluid mixing change mentioned in the changelog.)

FluidMixing-AllHidden.jpg (68.15 KiB) Viewed 1063 times

FluidMixing-OneVisible1.jpg (61.37 KiB) Viewed 1063 times

FluidMixing-OneVisible2.jpg (63.02 KiB) Viewed 1063 times

FluidMixing-BothVisible.jpg (63.2 KiB) Viewed 1063 times

Also Ref: 122199 Electromagnetic Plant outputting heavy oil through different output/input pipe.

[Zaflis]

The issue here isn't the EM plant whose recipe was changed, it was the electrolyte machine. The issue persists even with bot rebuilding. Try it yourself please, it is a little frustrating to be fighting to be believed. However, if you can see an obvious error I am not seeing, please let me know after you've tried and identified it.

Yes, it is possible to fix using the delete button in the pipes. After 800h in regular factorio, I am not used to it, and I don't think a building outputting the wrong fluid into empty pipes is intended operation. Somewhere I believe there is a crossed wire, as you are also suggesting, but I believe it is outside of my control, and it sits inside certain fluid buildings. It is as if there is a hidden pipe inside the building that outputs fluids arbitrarily. (I know it is ruled by logic, being a computer game, but the logic is either unintended (bug) or not logical to me, the user)

It also doesn't explain my second edit, how does a building whose recipe does not accept holmium solution, output holmium solution when you rotate it and place an empty pipe on its output?

Before i look more, lets be clear we are on the same page. This is known behavior for fluid mechanics:
(In/out connections are not filtered, same type fluid connections are connected, fluid backflows from pipe to internal buffer)

Fluid mechanics.mp4

(9.15 MiB) Downloaded 26 times

[heggart]

I noticed on the lateatest release they modified the prio a fluid gets into a system

120390

This probably will fix our issues

[BlueTemplar]

Fluid mixing will prefer the fluid with more volume and discard the other.

Well, sounds like a radical solution ?

What is somewhat baffling to me, is to why Space Age was released as it was, meaning :

[...]
I don't think a building outputting the wrong [for what they show as allowing for that input/output as a] fluid into empty pipes is intended operation.
[...]

considering the existence of Space Platform Thrusters and Fulgoran Electromagnetic Plants...

Was it so hard to forbid specifically this from happening, regardless of what fluids machine's fluidboxes might already hold ??
(Note how this goes against all other Factorio experience, in particular filter inserters.)
(Yes, I know about the pre-SA Steam Engine passthrough tricks, doesn't mean this should have stayed in SA, again considering the new machines.)
Have no betatesters, especially those without prior Factorio experience, stumble on this ??

[SamsaraRazor]

Before i look more, lets be clear we are on the same page. This is known behavior for fluid mechanics:
(In/out connections are not filtered, same type fluid connections are connected, fluid backflows from pipe to internal buffer)
Fluid mechanics.mp4

Known intended behaviour or unintended behaviour?
I finally found what you're refering to, FFF 351. So it seems that the idea is that you should still be able to mix fluid in the pipes, but to borrow the belts analogy:

1. Assemblers with recipes do not accept items they are not working with from inserters. (And as an extent it does not output its product to machines that does not accept it.)
2. You do not have to filter output inserters, because they only take the output, not the ingredients.
3. The item buffer for assemblers is displayed (except what it currently being used for making), and it is therefore transparent what is happening.
4. When you have a mix on the belts, the solution is to scoop it up, as stated in FFF 351. So why is the solution to mixing fluids to delete them, and therefore losing the resources? In the case of pipes, you can deconstruct from one end and have it get back into tanks, but in the case of machines, it fills an unspecified buffer?

That buffer can't be viewed, so you can't know that a building is fluid mixed until it either runs dry of something or you make a change, and then you risk that it contaminates every other machine with some fluid? Which means that that fluid in the buffer of every machine is lost.

It doesn't seem to be the intention of the change to cause grief as a gameplay feature, so it would maybe be a good idea to look this over again?

The issues I had with the EM plant, may be due to the same bug as mentioned 119600 Which you've also contributed to and dismissed as not really an issue, and that it can be fixed with deleting the fluid in the pipes.

Don't get me wrong, having people like you here on the forum is great, but this issue, intended or not is important to solve in such a way that it makes sense to people playing. It doesn't make sense to me, others too. Possibly some issues are bugs, some are intended/unintended and it is now such a mix of fluid related issues (hah) that I'm struggling to keep them apart in my head. This topic is viewed far more than anything else, and I think that may be an indication that others have this experience and are waiting for some kind of solution/statement other than what is propsed here.

[BlueTemplar]

I finally found what you're refering to, FFF 351.

https://factorio.com/blog/post/fff-351# ... Prevention
I don't see any reference to fluid in machines there ?

Also, that's not mixing, mixing would be something like oil and sulfuric acid turning into lubricant if they happened to touch and/or share a fluidbox, Factorio doesn't have different fluid mixing and never had.
(Or a bit how steam used to be really hot water, and Factorio does mix fluid temperatures... at least in a single fluidbox.)

That buffer can't be viewed

Yeah, as I pointed it to be (one of the) main issue(s) in another similar thread.
(I should have realized what was going on, too, but only remembered about that passthrough Steam Engine trick after the fact...)

[SamsaraRazor]

I finally found what you're refering to, FFF 351.

https://factorio.com/blog/post/fff-351# ... Prevention
I don't see any reference to fluid in machines there ?

Also, that's not mixing, mixing would be something like oil and sulfuric acid turning into lubricant if they happened to touch and/or share a fluidbox, Factorio doesn't have different fluid mixing and never had.
(Or a bit how steam used to be really hot water, and Factorio does mix fluid temperatures... at least in a single fluidbox.)

That buffer can't be viewed

Yeah, as I pointed it to be (one of the) main issue(s) in another similar thread.
(I should have realized what was going on, too, but only remembered about that passthrough Steam Engine trick after the fact...)

I don't quite understand what you're saying. When I say mixing, I'm talking about fluid being present in a system of pipes that it would ordinarily not be, such as holmium in a electrolyte system, or oxidizer in a fuel system by some method of action that looks like it is mixing through machines/thrusters.

For example with the steam engine example, even if you contaminate one steam engine with water (which I don't think is resonable, given that it only has one recipe) you shouldn't then be able to contaminate everything downstream with water through the steam O/I in my opinion. Also, you can contaminate the second steam engine with water, have it be working by connecting a steam engine with steam, but if the steam runs out, the water then backflows into everything and it can't be made to work again without pipe content deletion. The water isn't visible on tooltips for the engines while they work with steam (or at any point). Granted, this steam engine example was contaminated using bots.

I don't really see what is added from a gameplay perspective. I need to click a pipe and click delete if I run out of things? Because I made a mistake that I didn't catch when I set it up, but it worked fine until I ran out, then the mistake actually shows? Why? It's also inconsistent with other gameplay as I pointed out previously, and inconsistent within itself. If it is contaminated, it shouldn't be possible to make it work, and it should be readily apparent on the tooltip that this container contains *whatever*. And if it is supposed to be working even when contaminated, it should be possible to make it work again by raising the "pressure" of the correct fluid. (I'm playing devils advocate here, I don't actually expect the latter to be a feature) (Edit, which it is sometimes?? Not sure about consistency)

Finally, it doesn't seem to be a consistent feature. This is from 2.0.15:

Start with a fully contaminated steam engine:

20241109162244_1.jpg (736.34 KiB) Viewed 936 times

Then force steam into it:

20241109162300_1.jpg (771.22 KiB) Viewed 936 times

Et voilà! contamination gone, this time.

20241109162429_1.jpg (799.93 KiB) Viewed 936 times

All the fluids were also passed through the steam engine, which is another chapter. Imagine having an assembler with iron gear recipe, that you can input every item from the game into a hidden storage segment inside the assembler....


It isn't very strange this is considered to be a bug, seemingly fluids appearing from nowhere and without consistent rules.
Maybe the changes in 2.0.16 helps, I haven't tried them yet. But it still seems to me that hidden buffers are going to cause confusion.

[BlueTemplar]

Yes, my issue above was with the terminology that the FFF used, «pipeline contamination» is a much better term for it than «fluid mixing».

If it is contaminated, it shouldn't be possible to make it work, and it should be readily apparent on the tooltip that this container contains *whatever*. And if it is supposed to be working even when contaminated, it should be possible to make it work again by raising the "pressure" of the correct fluid.

Yeah, and/or it shouldn't be able to dump/accept fluids different from what the icons on the pipes indicate.
(Also note that when an EM plant has no recipe set, it doesn't have any pipe inputs/outputs either.)

All the fluids were also passed through the steam engine, which is another chapter. Imagine having an assembler with iron gear recipe, that you can input every item from the game into a hidden storage segment inside the assembler....

Didn't we just get something very similar to this in 2.0 with assembler trash slots ?

Maybe the changes in 2.0.16 helps, I haven't tried them yet. But it still seems to me that hidden buffers are going to cause confusion.

Depending on how it works now in 2.0.16, extra fluids might only spend one tick there, and never get the chance to contaminate anything else...

[Ext3h]

The EM plant is a special kind of nasty, given that depending on the recipe it will even open/close the internal connection between two of the inputs arbitrarily, making it really hard to correctly flush the pipes. But that's not really the point.

The point of this thread is, that there is specifically something going on with thrusters that causes them to mix up not just tiny residual amounts from an internal buffer, but somehow establish full connections between the oxidizer and fuel lines whenever the other drops to 0, and you have more than one port of each type connected.

Why am I so certain it's a full connection? Because it's not just residual amounts being dropped into the wrong pipe, it's actually filling up entire tanks.

By all means, it looks as if when the thruster drops resources from the internal buffer due to an imbalance, it's by accident doing so to all liquid output ports that may currently not have a resource...

I believe something similar is also happening with the EM plant, where it's unpredictable which resource gets dumped on what connector when switching recipe. (Not just due to the dynamic internal connection but in general.)

[fosterboy123]

I am experiencing the same issue as the OP.

I have attached gif, version & save file for inspection. Platform is Zaxaramas.

Cheers,
Steve

[fosterboy123]

Have confirmed that as OP suggested, the bug seems to present itself when the inputs are parallel to eachother.

This configuration does not present the problem:

11-10-2024, 13-07-02.png (1010.56 KiB) Viewed 852 times

[Zaflis]

I am experiencing the same issue as the OP.

I have attached gif, version & save file for inspection. Platform is Zaxaramas.

Cheers,
Steve

Sorry but your gif doesn't contain enough information to make it a bug. Like mentioned earlier the fluid inputs are not filtered to 1 type of fluid by design. If you have a contamination of fluid inside the truster already it will still spread out. You didn't flush the contents of the engine before the gif?

[fosterboy123]

I am experiencing the same issue as the OP.

I have attached gif, version & save file for inspection. Platform is Zaxaramas.

Cheers,
Steve

Sorry but your gif doesn't contain enough information to make it a bug. Like mentioned earlier the fluid inputs are not filtered to 1 type of fluid by design. If you have a contamination of fluid inside the truster already it will still spread out. You didn't flush the contents of the engine before the gif?

Maybe I am misunderstanding the fluid system. I have my output red and blue fuels going to the specific red and blue i/o on the thruster. I am not connecting red to blue or blue to red, so why would the blue end up in the red pipe or vice versa?

If this is not a bug, it is the first time in 300 hours of factorio that i have experienced this confusing fluid behaviour.

I have attached my save file, not just a gif, so it should be trivial to confirm if this is/is not a bug based on my post.

[fosterboy123]

I think I am starting to understand that you are saying I have pre-polluted the fluid inside the thruster itself, then when I reconnect the fluid is able to move backwards, back into the pipes.

I know you said this is by design, but I would argue that this post gaining such traction implies the system is not intuitive. To me, I would never have expected any red fluid to enter into the blue input. I would expect the fluid to be rejected by the thruster.

[The_Sane]

I know you said this is by design, but I would argue that this post gaining such traction implies the system is not intuitive.

Having hidden internal buffers is obviously a major point of confusion, especially when the thruster can somehow function for a time with the wrong fluid in these buffers, so long as the external connections have the correct fluid in them... Unfortunately there is also currently a bug that can cause fluids to enter the wrong pipelines after updating the game.

Thrusters are particularly easy to contaminate and not realise you did it, but if anyone here is absolutely certain that they did nothing wrong, have rebuilt the full system even, and then later (after an update) they find that the fluids are contaminated again then that might be the root cause, masked by the funky thruster mechanics.

[veriszg]

I have to agree with predecessors. I've experienced fluid mixing many times during the game.
At first I thought this was because mods and longer underground pipes.

I have debugged piping thoroughly and what I can tell, my pipes were full, but after loading the game... bam! water instead of light oil or water in part of the system that shouldn't contain water at all.

As a workaround I've managed to put additional pumps in place with filters set to filter out water if it happens to wrongly appear in pipes.

But there is clearly something bad happening in fluid system.

I'll post more info as soon as I trace it back more.

Just wanted to confirm these things happening.

[Zaflis]

I have debugged piping thoroughly and what I can tell, my pipes were full, but after loading the game... bam! water instead of light oil or water in part of the system that shouldn't contain water at all.

That is a different bug that happens to only some few people, so far reason is unknown but has something to with game version change between saving and loading. I haven't heard of it change fluids if you load with same version you saved it on. It could even be a memory leak in worst case.

What most people discuss here is just fluid mixing system in general that was present in 1.1 already. Also i never experienced unintentional fluid mixing in my saves from beta to present, and i always have at least 3 or more platforms that fly between planets. None of them have ever shown signs of mixing.