Solving bots vs. belts... with pipes?!?!?!

Post your ideas and suggestions how to improve the game.

Moderator: ickputzdirwech

Post Reply
bobucles
Smart Inserter
Smart Inserter
Posts: 1669
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Solving bots vs. belts... with pipes?!?!?!

Post by bobucles »

Did you like the clickbait title? I liked it.

Once upon a time the devs thought about dirty mining, but then they added infinite productivity research and the problem was solved. Unfortunately this created a secondary issue where belts could no longer keep up with the high output of mines. Miners can be casually boosted to double or triple output even with simple speed modules, and at that point belts can no longer keep up with the massive flood of ore that pours out. The problem can't be solved with more belts because the footprint for mining is very limited and is already at peak capacity. Players instead ""solved"" this issue by dropping belts entirely and using bots to mine. That's a shame. But what if there was another way?

Let's look into pipes. The current item standard is to treat 10 fluid as 1 "item"; so when 20 petroleum turns into 2 plastic it's the same as saying 2 petrol "items" will create 2 plastic. Pretty straight forward. A standard pipe can move a rather generous amount of fluid per tick over a reasonably large distance. Consult the chart below:

Code: Select all

Number of pipes  : Maximum flow (fluid/sec)
1	3000
2	2200
3	1860
5	1560
8	1380
12	1260
16	1200
23	1140
41	1080
**166	1020**
209	960
293	720
359	600
459	480
759	300
It's a very safe bet to say that you can move 1000 units (100 "items") of fluid/sec over any reasonably sized distance without issue. But wait! That's 2.5 times as fast as blue belts! What IF the ore output was created as a liquid? This would massively increase the ability to move ore. That means more throughput, and bots struggle to move fluids with the same ease as belt items. That means pipes can be the superior version instead of using bots.

Fluid ore offers some nice advantages over solid ore:
- It flows super fast, faster than belts.
- It automatically balances between tanks. No belt balancers needed.
- Merging and splitting pipes is far simpler than handling complicated belt setups. Thus pipe based mining outposts can be set up easily.
- Fluid loads into and unloads from trains extremely quickly.
- Fluid Miners look much cooler!

Obviously an ore mine can't just output a fluid, it'd need some kind of input fluid to function. This primary ingredient can be as simple as ordinary water or it could use some "washing" material like sulfuric acid. Don't ask me I'm no metallurgist. Ultimately the demand for more fluids will place more fluid trains on the train network, which the devs have stated as a goal.

Very few parts of a factory need a dramatic increase in throughput. Generally the most demanding parts come from the mining outposts and smelting arrays. Fluid ore and fluid smelting offer a way to dramatically increase the throughput of these operations to keep up with the rest of the factory. Naturally once the liquid smelted ore is out it needs to be transformed back into plates. A simple assembler recipe can handle this quick process, or perhaps a few alternate low tech recipes can use the input directly (like using liquid iron to make iron gear wheels).

There is also another extreme throughput demand in the form of Green Circuits, but I don't think liquid circuits will work very well. :roll:

User avatar
Deadlock989
Smart Inserter
Smart Inserter
Posts: 2528
Joined: Fri Nov 06, 2015 7:41 pm

Re: Solving bots vs. belts... with pipes?!?!?!

Post by Deadlock989 »

Why do you need infinite bandwidth out of a mine?

What's wrong with having a limit on it?

Edited to add, I'm not just being bloody-minded, at least, I'm not *only* being bloody-minded. So what if a mine has a cap on output? It just means it lasts longer. If you want more ore at once, you might have to build another mine, tragically. If people build bot farms to minmax ore output all because they don't want to have more than one mine on the go, that's their choice, isn't it? So what if belts can't keep up with that, they're like a hundred times cheaper?

Why does everything have to "compete" with everything else, why can't things just be different?
Image


dood
Filter Inserter
Filter Inserter
Posts: 360
Joined: Wed Mar 21, 2018 8:36 am
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by dood »

Ah hell no, not with those wonky pipe physics that literally nobody of the devs even knows how it works anymore.
Besides, pipes are an UPS murderer thanks to those calculations that ensure their throughput drops like a stone if you place like 4 pipes in a row with no pump because that's how pipes work, right guys?
You don't want that shit to take the place of newly performance optimized, sexy looking belts.

That love isn't ready for spreading until some hard fixes happen.

bobucles
Smart Inserter
Smart Inserter
Posts: 1669
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by bobucles »

Besides, pipes are an UPS murderer thanks to those calculations that ensure their throughput drops like a stone if you place like 4 pipes in a row with no pump because that's how pipes work, right guys?
I mean. The chart of pipe throughput is RIGHT THERE.

dood
Filter Inserter
Filter Inserter
Posts: 360
Joined: Wed Mar 21, 2018 8:36 am
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by dood »

bobucles wrote:
Besides, pipes are an UPS murderer thanks to those calculations that ensure their throughput drops like a stone if you place like 4 pipes in a row with no pump because that's how pipes work, right guys?
I mean. The chart of pipe throughput is RIGHT THERE.
Clearly, throughput was the main focus of my post.
You cracked the code, buddy.

bobucles
Smart Inserter
Smart Inserter
Posts: 1669
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by bobucles »

Come on Dood. Not every single post has to be dripping with scathing sarcasm. We get it. You're edgy. I've read over a hundred posts to that effect. It's extremely "hip" and "cool" but it's also the slightest bit rude. I'm merely proposing a different type of solution in the bots vs. belts debate that hasn't been considered before. It's okay to consider new ideas, right?

I've certainly talked many times before about how belt stacking is a great way to boost item throughput for the most intense parts of base design. But in truth bases only really struggle when it comes to moving large quantities of low tech ingredients. Changing the low tech products into fluids is a different type of solution (pun intended). It lets product move around at a much faster rate, and it also introduces new base design elements that did not exist before.

For example an ore mine's belt lane can only output up to 40 items per second. It's very easy to reach that limit with speed modules, rail settings and productivity research. If the ore output was changed 1:1 into a liquid then those pipe lanes will suddenly move an equivalent of 100 items per second. Not only is that a lot more raw mining goodness but it's easily passed over modest distances (undergrounds), automatically lane balances on its own(between storage tanks), and flows through train stations much faster than piles of stack inserters.

What advantages does using a fluid have over normal solid items? The most obvious one is that fluid movement favors pipes much more than bots. Bots can not move a fluid until it is transformed into a barrel, and it must be unbarreled on site to be used. Even after it's unbarreled it still must be piped around if you don't want a dedicated barreling station for every single thing. Bot bases get bulked up when lots of fluids are involved so they suddenly stop being the most ideal answer to base design.

Fluids can also benefit small mining operations, too. Without a fluid system you'd almost never consider connecting two ore outposts together- the belt length is simply too much. But with a fluid system it's not difficult to cross the moderate distance and tie a lot of ore mines together into a single large station.

Train networks get to have a lot more fluid trains when ore is being transformed into a fluid. Ore mines get to try a new layout that maximizes the power of pipes. Smelters can try a different layout for their liquid based input and generally it places a lot more pipes in a large base.
Why does everything have to "compete" with everything else, why can't things just be different?
But competition is fun. That's how new ideas get made, you know.

dood
Filter Inserter
Filter Inserter
Posts: 360
Joined: Wed Mar 21, 2018 8:36 am
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by dood »

bobucles wrote:Come on Dood. Not every single post has to be dripping with scathing sarcasm. We get it. You're edgy. I've read over a hundred posts to that effect. It's extremely "hip" and "cool" but it's also the slightest bit rude. I'm merely proposing a different type of solution in the bots vs. belts debate that hasn't been considered before. It's okay to consider new ideas, right?
Fine. I will reiterate for you.

I pointed out a problem. Game performance.
Not throughput or the utilitarian merits of pipes or fluids, game performance.

This approach is inherently hindered if you can't apply it to a large scale which is where it is supposed to shine.
So as long as pipes don't get the optimizations they need, this is a self-defeating feature that will be dropped in favor of the more UPS friendly thing in the scenario it was designed for.

jack/cartesiandaemon
Manual Inserter
Manual Inserter
Posts: 2
Joined: Wed Apr 11, 2018 9:51 am
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by jack/cartesiandaemon »

Or, if it seems desirable, how about a pneumatic system or monorail, to take the place of "conveyor belt but EVEN MORE THROUGHPUT"? I'm not sure if there's a need or not, but if there is, if it's end-game only, it could be implemented whatever the most UPS-friendly way is (e.g. treat the whole system as a single box with a constant short delay between "in" and "available to remove").

Heck, if you went EVEN MORE end game you could have teleporters, if robots don't keep up with throughput any more.

User avatar
Deadlock989
Smart Inserter
Smart Inserter
Posts: 2528
Joined: Fri Nov 06, 2015 7:41 pm

Re: Solving bots vs. belts... with pipes?!?!?!

Post by Deadlock989 »

bobucles wrote:But competition is fun. That's how new ideas get made, you know.
*shrug* Seems like something that's not even a problem being "fixed" by something a bit wacky to me, but that's just personal taste I guess. New ideas aren't necessarily good ideas. Which is why mods are great, because it allows people to have what you think are terrible ideas in the privacy of their own space, instead of foisting it on everybody.

Regardless of whether it's a good idea or not (PS. it's totally not), you can nearly make this with mods - most of the engine support is there (e.g. miners that require a fluid to function), except one crucial thing - miners always output an item. In fact they're the only producer in the game that dumps items straight onto a belt without requiring extraction by an inserter or a loader. Literally everything about miners is "belt" writ large.

You'd have to request a brand new kind of entity which superficially looks like a miner but has both input and output fluid boxes and automatically seeps fluid when it's situated on ore. Sort of a cross between a miner and an offshore pump. I think your chances of getting that at this late stage of the development cycle are limited.

Edited to add, also, the magic-potion-miner would have to be more "directional" than the current uranium miner, because it has both input and output fluids. Not sure how that would work practically if you wanted to keep the pass-through function that acid currently enjoys. I guess it would be more like a boiler, that passes through water and puts out your herbal ore infusion at a right angle. Sounds time-consuming to make this intuitive and not a visual mess. Possible though. Worth it? I'd rather have icons in chat text.
Image

User avatar
darkfrei
Smart Inserter
Smart Inserter
Posts: 2903
Joined: Thu Nov 20, 2014 11:11 pm
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by darkfrei »

except one crucial thing - miners always output an item
Pumpjacks are also miners.
You'd have to request a brand new kind of entity which superficially looks like a miner but has both input and output fluid boxes and automatically seeps fluid when it's situated on ore.
https://mods.factorio.com/mod/HydraulicPumpjacks

dood
Filter Inserter
Filter Inserter
Posts: 360
Joined: Wed Mar 21, 2018 8:36 am
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by dood »

Deadlock989 wrote:*shrug* Seems like something that's not even a problem being "fixed" by something a bit wacky to me, but that's just personal taste I guess. New ideas aren't necessarily good ideas. Which is why mods are great, because it allows people to have what you think are terrible ideas in the privacy of their own space, instead of foisting it on everybody.

Regardless of whether it's a good idea or not (PS. it's totally not), you can nearly make this with mods - most of the engine support is there (e.g. miners that require a fluid to function), except one crucial thing - miners always output an item. In fact they're the only producer in the game that dumps items straight onto a belt without requiring extraction by an inserter or a loader. Literally everything about miners is "belt" writ large.

You'd have to request a brand new kind of entity which superficially looks like a miner but has both input and output fluid boxes and automatically seeps fluid when it's situated on ore. Sort of a cross between a miner and an offshore pump. I think your chances of getting that at this late stage of the development cycle are limited.
I don't see the problem here.
Why bother with miners, just make it a chemplant recipe.
Water and ore in, fluid out. Fluid in, ore out.
Done.

Edit: as a matter of fact,
viewtopic.php?f=190&t=59278

User avatar
Deadlock989
Smart Inserter
Smart Inserter
Posts: 2528
Joined: Fri Nov 06, 2015 7:41 pm

Re: Solving bots vs. belts... with pipes?!?!?!

Post by Deadlock989 »

darkfrei wrote:
except one crucial thing - miners always output an item
Pumpjacks are also miners.
You'd have to request a brand new kind of entity which superficially looks like a miner but has both input and output fluid boxes and automatically seeps fluid when it's situated on ore.
https://mods.factorio.com/mod/HydraulicPumpjacks
So they are. So you could implement this terrible idea after all. I wonder why no-one else thought of it yet.
Image

User avatar
darkfrei
Smart Inserter
Smart Inserter
Posts: 2903
Joined: Thu Nov 20, 2014 11:11 pm
Contact:

Re: Solving bots vs. belts... with pipes?!?!?!

Post by darkfrei »

Deadlock989 wrote:So they are. So you could implement this terrible idea after all. I wonder why no-one else thought of it yet.
The mod All My Fluids must do this too, but I haven't enough time for testing.

Post Reply

Return to “Ideas and Suggestions”