Friday Facts #430 - Drowning in Fluids

[TheBuzzSaw]

I just wanna say LOL at all the people who are crying over this new system and now defending the old system. "It's a little quirky but can be managed."

The old system literally doesn't work. Random machines get starved. Pipes randomly stop flowing. And they're sensitive to placement order? No thanks. I'll take 100 consistent rules over the RNG of the old system any day.

[grandexar]

I was confused by the use of the word “extents” because my understanding is that word means “to the degree” like “he went to great extents to buy the rare collectible”. I looked up the definition and it also means “area” so when the devs said “pump extents” they meant “pump areas” talking about the 250x250 grid.

——

To people complaining about it not matching the tiling grid, did you forget that electric poles don’t match the tiling grid? Neither do turret ranges. And you can’t turn inserters to 90 degree angle operation. It’s all meant to force creative solutions.

[Tealtanium Golem]

This Is going to make huge defense walls with flamethrower turrets much more expensive to run and set up. Not only do we need power for the pumps to keep all the flamethrower turrets supplied at all, but now I'll have to use pumps where, as it stands now, I don't have to use any. A simple requester chest of oil barrels being unloaded into the pipeline was all that I needed to keep even a relatively large wall fully stocked with oil at all times. Even with the low throughput of the current system. Could flamethrower turrets themselves refresh this restriction if it must stay in place? I am also in favor of the 256x256 change if the system must stay.

[KuuLightwing]

I just wanna say LOL at all the people who are crying over this new system and now defending the old system. "It's a little quirky but can be managed."

The old system literally doesn't work. Random machines get starved. Pipes randomly stop flowing. And they're sensitive to placement order? No thanks. I'll take 100 consistent rules over the RNG of the old system any day.

I don't know how I played with old system for 2000 hours when it "literally didn't work" then. And besides, yes, old system isn't great, but this is just not better. Old system had quirks, this one has arbitrary limits because reasons. One step forward, two steps back.

[raiguard]

And here we have another example of an "optional" quality.

I hope there soon will be a mod that sets all/selected items to a pre-selected quality level before the game start, and hides the research and quality related visual nonsense from UI. I would rather play with such mod and without achievements.

I looked at several vanilla megabases and none of them even came close to needing 12k/s. Because of how the old system worked, even placing a single pipe entity would halve the throughput to 6k/s, and placing more would make it even lower.

You only need quality pumps if you are trying to feed quality machines.

[ArEyeses]

Does this mean my walls with flamethowers now need pumps every 250 tiles (~20 flamethowers) otherwise the very small less than 1/s of oil they use won't be able to get through?
This means flamethower turrets now need power which is something they previously didn't

[KuuLightwing]

I looked at several vanilla megabases and none of them even came close to needing 12k/s. Because of how the old system worked, even placing a single pipe entity would halve the throughput to 6k/s, and placing more would make it even lower.

You only need quality pumps if you are trying to feed quality machines.

Vanilla megabases didn't have fluid based metal processing though, did they? Also, how many of those used coal liquefaction for oil products thus avoiding fluid trains entirely?

[Theikkru]

The Water to Steam ratio change makes it more possible to supply power or nuclear setups using fluid wagons or even water barrels.

The reason I never even tried this is not throughput limitations; it's because there's no unpowered way to get fluids from barrels or wagons into pipes.
It feels like a broken catch 22 that you need to bring solar panels to power the pumps (for wagons) or assemblers (for barrels) needed to feed water into the boilers to power the pumps or assemblers needed to feed wa...anyways, please make at least one of these methods "gravity feed" a small amount of fluid into pipes, even without power. I think it would also be a nice niche use case for barrels if they could be unloaded without power, albeit very slowly.

[Tealtanium Golem]

did you forget that electric poles don’t match the tiling grid?

Big electric poles are being changed to match the grid / extended to have 32 tiles of range. They mentioned that way back with the rail changes in https://factorio.com/blog/post/fff-377

Unless they backtracked this change somewhere along the line, then ill eat my words.

[thermomug]

I just wanna say LOL at all the people who are crying over this new system and now defending the old system. "It's a little quirky but can be managed."

The old system literally doesn't work. Random machines get starved. Pipes randomly stop flowing. And they're sensitive to placement order? No thanks. I'll take 100 consistent rules over the RNG of the old system any day.

I agree, I suspect people are just reacting defensly and trying to forsee implications, which is rather hard and pointless IMO.

I think the issues with the old system being broken lies in the fact, that during a tick update, the fluid system is updated "in-place". This makes the order of entity updates relevant for the resulting physics. E.g. if you update fluid flow from north to south, northern pipes will get updated first and claim more of the fluid. This makes fluid-flow directional.

My proposed solution to this would be that during tick update, instead of mutating the current state in-place, one would need to calculate the new state S' from the old one S and then replace S with S'. Fluid system states would then be immutable which is also good for multithreading. The price would be higher memory consumption/allocations.
Hybrid strategies would also be possible, where each part of the fluid system (a segment) gets updated independently in the above described procedure.
God, thinking about performance optimizations for my favorite game is way too much fun xD

[raiguard]

Who TF thought limiting pumps by 10x would be a good idea??!?!

Now my nuclear reactors (who still falsely say they consume nuclear fuel (that stuff to power vehicles) instead of nuclear fuel cells... and they say that's correct... anyways...)

Now my nuclear reactors will need 5x to 10x as many pipes, as it won't be an option anymore to consolidate pipes for that reason.

That's a huge step backwards and totally uncalled for, I really hope that's just an oversight and will be reverted before the release!

You don't need 5x as many pipes because throughput through a given pipelines is unlimited, it's only limited at the connection point to each machine. Also, offshore pumps are unchanged, they have been 1200/s for a very long time. So you can just add more pumps to the same pipe for higher throughput.

[raiguard]

I looked at several vanilla megabases and none of them even came close to needing 12k/s. Because of how the old system worked, even placing a single pipe entity would halve the throughput to 6k/s, and placing more would make it even lower.

You only need quality pumps if you are trying to feed quality machines.

Vanilla megabases didn't have fluid based metal processing though, did they? Also, how many of those used coal liquefaction for oil products thus avoiding fluid trains entirely?

If you don't build quality machines then you usually don't need quality pumps, even in Space Age.

I don't know about the coal liquefaction, I didn't pay attention.

[raiguard]

So is there still no vanilla way of using three pumps on one side of a fluid wagon? And now we're even more pump limited. Unloading fluid trains in a large to mega-base scale was always a struggle and is nearly impossible now.

You don't need to pump directly from a fluid wagon into a tank any more, you can just place pipes and put your tank elsewhere. So you can very easily use three pumps on one side of a fluid wagon.

In hindsight I should not have used my 0.17 base to demonstrate these features. That was a mistake.

[raiguard]

This Is going to make huge defense walls with flamethrower turrets much more expensive to run and set up.

Good. Flamethrowers have been extremely OP for years so I think making them more difficult is nothing but a good thing.

[jamaicancastle]

Instead of using a 250*250 bounding box as the limiting factor the number of pipe segments should be the relevant number. It is far more realistic (which I know is not that important in game design) but it being realistic also makes it more intuitive.

Also instead of either 100% or 0%, the output should be a function relative to the number of pipe segments. For example 100% for 1-50 segments and then dropping near 0% at 300 segments. It might be worth to consider counting underground pipes and fluid pipes as multiple segments to balance long distance pipelines with distribution pipelines.

With that system a player could decide how much throughput their pipeline needs and therefore how often they need to place pumps.

I'm a fan of both of these tweaks. I don't like how a complex pipe manifold supplying many buildings doesn't provide any more "resistance" than a long straight pipe, it just doesn't feel intuitive. It might make builds involving a lot of fluid processing more interesting than just stamping down a long row of the same blueprint, too, as you have to take into account the size of each district and where they'll need pumps.

I'd also second the idea of a gradual rather than abrupt shutdown. I'd lean towards an exponential falloff, something like: at X tiles in a district, you get 100% flow rate. At 2X, 50%. At 3X, 25%, and so forth. Something so that a) going over the limit by a handful of tiles isn't an instant kiss of death, and b) you could have a long pipeline with a very small amount of fluid if that's all you need.

I also really hope this information is available in the building window for the pipe.

[epartridge]

Looks like a solid update! I'm also in the 256x256 camp. Would love to see a power-of-two number that is also chunk size!

Agree with blog post that some limitation of pipe throughput is wise. This choice of limitation seems fine on the whole, but it does feel a bit artificial. Total pipe segment count (or more properly total pipe length count, to take undergrounds into consideration) would seem less artificial to me, and is still very easy to understand and represent to the user. Also, the strict cutoff would feel a lot more physically plausible as a throughput reduction that scales with how much you're exceeding the pipe network budget.

One thing I'm worried about here is long-distance throughput. If you want to get a lot of fluid from A to B with pipes, what that looks like now is a single pipe with a 'fanout and remerge' every 250 tiles that splits the single-pipe flow across 4, 8, or however many pumps you need. Would love to see some different pipe options to address this long-distance issue. Perhaps a pump that's much stronger but doesn't connect to fluid wagons? As a more far-fetched solution, would be super cool to see "large pipes", multiple tiles wide, with a construction system similar to rails, and specialized structures to import/export to smaller pipes. Really get that factory look going!

[KuuLightwing]

You don't need to pump directly from a fluid wagon into a tank any more, you can just place pipes and put your tank elsewhere. So you can very easily use three fluid wagons on one side of a train.

In hindsight I should not have used my 0.17 base to demonstrate these features. That was a mistake.

I mean before I used a design that doesn't pump directly into/from a tank, there was a pipe segment between them. Yea, it limited the output to around 6000 IIRC, but it allowed for a rather compact design with load balancing too. Here's an example.

09-27-2024, 18-51-39.png (699.88 KiB) Viewed 1721 times

Now, I will have to use three pumps to get a half of that, and because of the way pumps are with much larger and more unwieldy footprint.

[TheBuzzSaw]

I don't know how I played with old system for 2000 hours when it "literally didn't work" then. And besides, yes, old system isn't great, but this is just not better. Old system had quirks, this one has arbitrary limits because reasons. One step forward, two steps back.

K. Enjoy playing 1.1 forever.

[Hiekie]

And once again, I'll repeat my comment from "Fluids "2.0" topic - just make new system a setting or a mod to turn on/off. The fact that this simplified version of fluid handling is being forced on all players and not just the ones who'll buy SA is mindboggling to me.

I am sorry, but i agree with the last part of this post, i actually LOVE the fluid system of 1.1, only thing that maybe should be changed is that buildings do not consume fluid in the same order they are built like it does now..

So please make a setting (or a mod) to use the "classic" fluid system, this system just feels so natural.

I love all the other changes and can not wait to play to new expansion!

[Taipion]

dude, are you dense?

you did not read a single line of what I said

@mods please remove this troll from the forums...


Please explain to me how you can not understand that?!
You still need to transport the SAME AMOUNT of steam, but the pumps have 10X REDUCED THROUGHPUT, so you need 10x (practically 5-10x) more pipes to transport the same amount of steam.

I can't comprehend how one can repeatedly fail to understand this simple thing, yes insists to keep on talking.

You do understand that you will not need any pumps if you keep your pipes short enough right? You do not need 10x the pumps, you need 0x the pumps. And higher quality pumps will have a higher pump rate if you want to move steam a long distance.

what is, post expansion, the limit of what I can put through pipes without using pumps?

go on, tell me!, if it's less than 2k/s it will break reactor design in an absolutely fundamental way