Factorio: Captain Planet doesn't stand a chance.

The Something Awful Forums > Discussion > Video Games > Factorio: Captain Planet doesn't stand a chance.

zedprime: Jun 9, 2007; yospos

Agent355 posted:

I want some way to measure total throughput of a particular place and total input capacity of another so I can be more optimal.

Its called a notebook and a stopwatch. But that's getting closer to Factory than Factorio.

# ¿ Mar 15, 2016 23:13

Adbot: ADBOT LOVES YOU

# ¿ May 11, 2024 14:17

zedprime: Jun 9, 2007; yospos

Everyone has different triggers for oh god what a mess, but its hard to argue with slamming stuff down the belts faster than you can consume it, but that shouldn't discourage a few critical looks especially if you are sticking to crazy spagehtti. There's some tricks to debottlenecking stuff that I didn't figure out until way too late, like needing 2 long inserters to keep up with 0.5s crafting speeds.

By the time you have blueprints and a backpack of bots you can have the basic building blocks for building stuff automated and if you so desire you can use the old dirty slambus factory to build a new pristine factory where the lines are load balanced with the splitters and the trains run to perfect precision. And then blueprint stamp it into the horizon.

# ¿ Mar 15, 2016 23:45

zedprime: Jun 9, 2007; yospos

Default biters seem really accommodating as long as you don't end up in a treeless wasteland. It probably helped that I learned through blue science in a passive game, but I don't think I especially rushed laser turrets but had them set up around the time mediums started showing up, and especially never needed to worry about ammo belts to turrets, just shoved a stack of whatever was current into every new one and they never seemed to go dry.

You can go on the offensive a lot earlier than you might first expect. Like if there's a nest early on giving you grief, just the SMG is good enough at the beginning. Then the buggy outranges small worms and with a few gun upgrades and AP ammo, tears through nests quick enough that it's not impossible to clean house depending how much pollution you are billowing. The biggest barrier to early game nest clearing are medium worms which mean you just need to leave a few nest clusters alone until you get a tank and poison capsules. But if there's only a few even that's doable with a buggy by using a sacrificial gun turret to tank damage.

Even if at default they aren't the biggest deal, it just feels like they're keeping me honest since defense goods become such an important cornerstone of your factory.

# ¿ Mar 17, 2016 12:18

zedprime: Jun 9, 2007; yospos

I'm about to build a rocket silo and my electric furnace count just graduated to needing to be counted on two hands.

# ¿ Mar 17, 2016 18:32

zedprime: Jun 9, 2007; yospos

I don't doubt I am leaving production on the table. But also I seemingly set up plate production futureproofed for the speed at which I build things up until rocket stuff. I don't tinker a lot with old proven lines and just build something new off to the side because I have the room. Hence finally needing some new furnaces and they finally get to be electric.

There was only one ingredient I needed to not build this silo straight off, and it may surprise or make perfect sense that it is raw steel.

# ¿ Mar 17, 2016 18:59

zedprime: Jun 9, 2007; yospos

A lot of really pretty factories and cool expert setups getting posted, so I thought I'd just leave this here as an example that you, too, can launch a rocket with the conveyor belt vomit method.

# ¿ Mar 18, 2016 20:25

zedprime: Jun 9, 2007; yospos

Concrete, because what else are 400 construction bots going to do because it sure as hell wasn't blue printing organized manufacturing lines.

6x exoskeletons on top of concrete was actually more of a problem than a help, I couldn't really input movement commands short enough to nudge myself off conveyor belts when I wanted to stand still somewhere specific. Moving through the base was a lot like plinko.

# ¿ Mar 18, 2016 21:35

zedprime: Jun 9, 2007; yospos

RattiRatto posted:

Here you go, thanks.
So the assumption "don't destroy their annoying base until when it's absolute necessary" holds?
From there i get that it only matters if i actually destroy their nest, but the indiscriminate genocide of biter do not affect the growth, right?

Don't ignore that eating pollution is a constant tick once you aggravate a spawner. Its a far less magnitude because its constantly ticking over on eating pollution. If the spawner destroying evolution punishes anything, its striking way out beyond your smog cloud to kill bases that aren't aggravated yet. If you have the time and inclination, trimming down bases inside the smog cloud isn't the worst idea in the world.

Loopoo posted:

With the MKII Power Armour, will I be able to wade into enemy nests and kill them with blue shotgun shells? Cause with the first power armour, I die real quick, even with a couple MKII shields and lots of battery.

Its got more slots for exos, rectors, and shields so to a point, yes. But you're still really reliant on chucking in poison capsules and distractor capsules, maybe graduating to destroyer capsules too if you have the research for a personal robot army and the materials to supply a line cranking them out. Nothing really stands up to those bases with a sea of large worms except the indirect methods.

zedprime fucked around with this message at 13:54 on Mar 21, 2016

# ¿ Mar 21, 2016 13:44

zedprime: Jun 9, 2007; yospos

Travic posted:

All this talk about nests is making a lot of sense. I'm starting to reach the point where talking them out is very difficult. I'm cruising around in mk2 armor with two arc reactors and laser defense for the rest. I'm also using a tank with explosive shells and neither is cutting the mustard anymore. Are destroyer and poison capsules the last thing to add? Or is there more?

Combat gets a little weird by the time your factory is set up. The tank isn't a tank, its a battering ram. You can run it through as many mobile enemies as you'd like and about half a dozen spawners or worms before needing a repair. But by the time you have mk2 armor that's getting a little old and crusty for methods because you or your bots need to get out and repair it more often than you probably want to at that stage of the game.

So you load up with a comfortable amount of exoskeletons and shield with 2-3 reactors. Take a combat shotgun with a load of AP shells. And get some poison capsules and distractor capsules cranking out at the least, destroyer capsules only if you already have the research that allows extra combat followers. Kite horde around while you rain shotgun fire into whatever is in range and use poison to soften up worms. Combat shotgun is amazing, but there's no such thing as enough shield for large+ spitters and large worms so you need to kite and get some indirect stuff like poison and distractors going.

e. alternately you go full weirdest and just use large power poles and laser turrets. A little more planning, some WWI style battle lines moving forward, but you can just leave the junk there as forward batteries that stop things from spawning so near in the future.

zedprime fucked around with this message at 00:58 on Mar 22, 2016

# ¿ Mar 22, 2016 00:55

zedprime: Jun 9, 2007; yospos

The health gauge comes up so little I'd imagine I'd just end up ignoring it if not committing suicide every half hour. What I really need is a mod with a peripheral that hits me on the nose with a newspaper every time I manually craft something my factory is already tooled to make automatically and I just haven't place the assembler yet.

# ¿ Mar 22, 2016 14:40

zedprime: Jun 9, 2007; yospos

Travic posted:

Ok I'll give that a shot. I haven't used robots much.

One other thing. I ran a test to see how much power a single solar panel produced. In one day it filled 4 accumulators (20MJ?). The wiki says you need 0.84 Accumulators per solar panel. Even with the panel powering the base and charging accumulators it seems like you'd need more than that.

PS I did make sure the test bed was not connected to my main power grid.

Under a real load, the solar panel needs to multitask. You can think of the power accounting as it is supplying the day time power to consumers simultaneously with the night time power to accumulators for those same consumers.

Depending on the laser stress on your power grid you can float more accumulators per panel, but just adding more of both in the nearest ratio to the ideal is just as easy since accumulators are usually the limiting component if you have anywhere near the same love/hate relationship with oil as I do.

# ¿ Mar 22, 2016 16:20

zedprime: Jun 9, 2007; yospos

Loopoo posted:

I don't understand accumulators. How can I accurately determine if my stored up energy is enough to keep laser turrets firing at peak performance for 20 seconds? They're both in different units. Watts and Joules.

Beside the cornerstone everybody mentioned, if you're looking at accumulators to soak laser shots:

An unupgraded laser turrets can shoot 3 times per second at 800kJ, or a power draw of 2.4 MW

An accumulator can provide energy at a rate of 300kW, up to a total of 5.0MJ

To keep up with the power draw without touching standard generation, you need 8 (300kW*8=2.4 MW) accumulators per turret. This gives you 40MJ of resevoir, or 50 shots, or 16 seconds of uninterrupted firing.

Upgrading to 20 seconds, then you need a bigger resevoir. So 2.4MW*20s=48MJ. 48MJ/5MJ=9 or 10 accumulators per turret.

This all changes when you change the firing speed. So you can do the math for the new firing speed, or just slam down solar fields everywhere because space is cheap at default biter settings.

e. Strict power accounting is neat but flying by the seat of your pants is really fun, because when you end up shorting power for laser turrets at night the game turns into a horror movie. The lights start flickering with each laser shot, assemblers grind to a hault, and inserters get all herky jerky and that's when you know the bugs are coming

zedprime fucked around with this message at 16:42 on Mar 22, 2016

# ¿ Mar 22, 2016 16:37

zedprime: Jun 9, 2007; yospos

Travic posted:

Ok. I'm trying to plan how many solar panels I need to run my base. My laser turrets aren't used much since I've been very aggressive about removing nests that are affected by my pollution so attacks are rare and small.

So a solar panel produces about 20MJ during daylight hours.

A day lasts 417 seconds.

I add up all the kW in my base and that's how much power I need per second?

My secondary resource base: 12 electric miners. 12*90*417=450,360kW per day. 450,360/20,000=22.5(round up to 23) solar panels to power them all day and night? With 20 accumulators?

Working off the math cited for the accumulator ratio which should still apply since day/night hasn't been changed apparently. There's an even easier rule of thumb, and it puts 23 solar panels as kind of lowballing (it puts a theoretical day closer to 15MJ per panel so not sure where the discrepancy was vs your practical test, but whatever, the engineering bottom line is use what works even if the math is against it): for a totally solar operation, you want solar generation = ~~1.67~~ 1.42 * average consumption, with accumulators ratioed to 0.84/panel from there. So for the miner example closer to ~~30-31~~ 25-26 panels and ~~26-27~~ 21-22 accumulators.

Or else just

mastermind2004 posted:

You need more solar panels. That's always how many solar panels you need. More.

e. fixed way after the fact for fat fingering the calculator

zedprime fucked around with this message at 21:46 on Mar 22, 2016

# ¿ Mar 22, 2016 20:01

zedprime: Jun 9, 2007; yospos

Loren1350 posted:

You want one solar panel per 42kW needed, and one accumulator per 50kW needed.

These are the correct numbers and I fat fingered my last example which should now be fixed.

# ¿ Mar 22, 2016 21:47

zedprime: Jun 9, 2007; yospos

hannibal posted:

Is anyone doing anything interesting with circuits? By which I mean the green/red wire, combinator stuff. I'm almost at the end of a fairly vanilla sandbox game (first playthrough) and haven't used them at all. (Trains either, for that matter)

Also, I'm always looking for more recommendations for quality of life mods. I'm not using a lot right now, just:

- Flare Stack
- Flow Control
- Fluid Barrels
- Side Inserters
- WaiTex

and I'd use Rail Tanker if I used trains, I guess.

The most common use for circuits seems to be light and heavy oil stock control. Hook up lubricant storage and a pump on the line from heavy oil to the plant cracking heavy to light, set a floor on lubricant that pumps excess heavy to cracking. Hook up light oil storage and a pump on the line to the plant cracking light to petrogas, set a ceiling on light oil (or a floor if you are deep into solid fuel production) that will pump light to cracking.

Since they are cheap and early there's some utility in setting up some faux logistic network conditionals. Like I keep a chest of straight pipe to feed into underground pipe and set the inserter not to pull below however much straight pipe I want to keep around at short notice. You can keep solar panels and accumulators ratioed with a ceiling of however many you expect to place at once. That utility wanes obviously with a mature logistics network cause its easier to set inventory control on a wider level without stringing everything up.

Most of the logic controls are for insane people, most everything you want to do involves smart chests, smart inserters, and your choice of wire color.

# ¿ Mar 22, 2016 23:43

zedprime: Jun 9, 2007; yospos

Loopoo posted:

Thing is, I gotta lay down laser turrets in my outpost. Gun turrets just get obliterated before bots can repair it from spitters. Plus, I hate the logistics of having to visit the outpost myself to restock the belt that feeds gun turrets with ammo (I have a chest that I fill with ammo then an inserter that puts it onto a belt that loops round the outpost). The reason I'm so stingy with laser turrets is because I'm in a constant battle with oil, so I prefer keeping laser turrets for my super important stuff (ie main base).

I'll double-layer gun turrets in my outposts from now on and that should hopefully keep them protected.

Anyone else get a little anal when they check turrets and see some with 0 kills? I end up wanting to remove them but then worry that maybe one day biters will break in through that place.

It sounds like you've caught yourself in a catch 22: you need laser turrets to get sustainable oil set ups, but you need a sustainable oil set up for laser turrets. Don't be afraid of shuttering other battery users to mass produce turrets until your outposts are safe, especially blue science. Use blue science to rush advanced oil processing, then mop up red and green techs for a while to let you flood the battery users that let you ultimately make more batteries like laser turrets. Laser turrets are probably more important than robots and blue science at that point, maybe equal importance for accumulators. Your battery factories are solely for the defense industry until you've got outposts sewn up.

# ¿ Mar 23, 2016 13:06

zedprime: Jun 9, 2007; yospos

RattiRatto posted:

Well the basic is the decider. You can input any signal, compare it to something and output a signal to a smart inserter. If the smart inserter is connected to a wire, it always works when the constrain you set is true.

What i use it for?
the Smarter Circuit mod lets you have a sensor to get how many of a item is inside a container. It outputs it as a signal with a item and a number -> attach it to a decider, if it's smaller than whatever you need -> activate smart inserter to feed your assembly line.

That's all in the base game if I'm reading you right. Connect a smart chest and a smart inserter by circuit wire. Set inserter to pull when item < N. No mod, no decider.

That's probably the only useful use of circuits before you get the logistic network going, if you want to keep stocks of things that are less than 1 slot. If your stock is in multiples of slots, don't forget about slot restrictions in chests.

SneakyFrog posted:

Soooo... anyone have a link or a good explanation for the "smart" devices and the logic thingies?

Essentially Id like to do batch manufacturing to where all the separate parts were counted/collected into a batch and then manufactured.

... or i could just be overdoing it because... well.. because.

I too have dreamed of a smart pull/made-to-order manufacturing line because the push manufacturing from a bus is totally wrong. But its not actually wrong because there's very little punishment for work in progress as long as you set up inventory control on your most expensive items.

Doing more complicated than the smartchest-smart inserter loop and you need a grasp on algorithmic math which is better taught in other circumstances. Including games if you're into that. Its also, as mentioned, completely superfluous.

# ¿ Mar 23, 2016 13:40

zedprime: Jun 9, 2007; yospos

Solumin posted:

Please don't you'll make me sad

Real talk, there's a lot of things your can't craft by hand, like engines.

Pull-based automation is a neat idea that you could approximate to some degree by pulling components for one assembler's product directly from the assemblers that make them. No bus, no intermediary storage, just directly pull from one assembler into another.

Or just give in to the beautiful world of belts and buses!

But you could end up with WIP in the assemblers! That is not tidy! :spergin:

Using chest limiters or circuit/logistic conditions to create a finite pull at the end is as much pull as you ever need because plates and circuits sitting on the belt is a pittance of extra pollution for the convenience factor, and as you say more complicated constructions shuffle materials from assembler to assembler. If I knew anything about scripting and thought it'd be balance-able in some way, figuring out a punishment for WIP would be an interesting experience compared to the buses of popularity.

Even without that, a made to order robot system seems like it'd be fun to make. Just setting up robot makers to dump everything into roboports is good enough since you can limit or debottleneck the rate as you choose from some of the slow steps, and who doesn't want extra robots? But imagine if you had a highly parallelized robot maker line with logic set up that let you order 500 robots at a time.

# ¿ Mar 23, 2016 14:27

zedprime: Jun 9, 2007; yospos

Solumin posted:

Achieving one logistics robot per second isn't even that hard -- it's a bit bulky, but everything besides advanced circuits and electrical engines is basic components.

I'm not sure what you mean about WIP. There's nothing wrong with having stuff lying around!

Ratzap posted:

0.13 will make doing this easy because they're adding switches. So build all the precursors and when all the signals add up correctly, activate the switch which powers the factory block until the desired output count is met and it switches the power back off.

Very sane responses. But I'm talking insane here. Its not about the speed of robots. Its about making robots with nothing left on the belts or in assemblers afterward. Turning off the base power draw is awesome, but leaves WIP in assemblers. Kill all WIP.

Welcome to the Clean Plate Club (full album)

Nothing on the belts. Nothing in the assemblers. It only chews through the exact number of components as you tell it to at any given moment.

The cornerstone is this mess of spaghetti wires. It picks the right amount when told by the central controller, then the central controller ticks down and it sends everything into the live system.

With a blueprint it isn't 100% crazy to create production lines this way, only 99%. Red and green wires and chest circuit "switches" are a loving shitshow for trying to program logic. Those chest circuit switches end up being a nightmare because the marker item in transit takes real time for the state to change that caused troubleshooting woe. I can't even tell you if half those wires are even necessary, but it all works and I've managed to avoid crosstalk by no doubt luck. Its kind of hypnotizing to watch work in a way that your first bus was so many gameplay hours ago.

# ¿ Mar 23, 2016 20:22

zedprime: Jun 9, 2007; yospos

SneakyFrog posted:

I like your flavor of crazy.

is there a limit as to the max distance for circuit wires?

They travel along power poles if you want to hook things up long distance. Red and Green are essentially two channels, and you broadcast and receive through the entire channel through anything attached, which is why I talk about lucking out of crosstalk by fidgeting around and luck.

# ¿ Mar 23, 2016 20:43

zedprime: Jun 9, 2007; yospos

RattiRatto posted:

Yes, but you can tide it to an electric pole and make it run through many poles. Therefore it's possibly infinite.

Zedprime good job with the requester/production of the clean belts project. How do you send the "request signal"?

There's a little more specification in the Imgur album if you haven't looked, but the circuit system is so janky its hard to describe very well.

But a request signal is as simple as >0 in the counter chest. You load up the counter chest with your inserter stack tech times the quantity you want, as long as its got something in it it says hey get going.

I run it through a decider to pare it down to 1 and run it through an adder that sums it up with anything in inventory so it only starts picking cycles when it the master signal is on and it has nothing in the picker or inventory chests. When it finishes a cycle, the same signal dumps everything out as does increment the counter chest down so by the time it tries to start a new cycle, the master signal might be off if its finished.

e. I could probably specify the mess in functional block form if there was interest

zedprime fucked around with this message at 21:04 on Mar 23, 2016

# ¿ Mar 23, 2016 21:00

zedprime: Jun 9, 2007; yospos

If what Youtube is telling me is right, several hours from now this might be stuttery mess with warped audio because I don't actually know anything about video capture and I don't have a ShadowPlay graphics card, but it will show the basic idea.

https://www.youtube.com/watch?v=mY6zsz8Okt0

e. Videos out of the oven, get it whiles it hot.

zedprime fucked around with this message at 00:05 on Mar 24, 2016

# ¿ Mar 23, 2016 21:39

zedprime: Jun 9, 2007; yospos

Ratzap posted:

I like this. Given that Factorio is supposed to be Turing complete, I'm surprised there aren't more weird control circuits around.

I imagine the draw just isn't there. Gameplay wise, a bus really is the correct answer, and where the simple implementation fails there's load balancing by belt design to do to fix it. The average person could probably benefit from a little interwoven inventory control, I know I could. But you can also just make more forges and oil refineries until you don't need to worry about robbing batteries from the laser turrets and accumulators to feed blue science and ending up eaten by biters.

Meanwhile its not an incredibly attractive tool on the theoretical side. It lacks the immediate recognition of Minecraft's circuit board like constructs, and logical constructs are kind of a mess, assuming I'm not missing any big ahas people have found out, which is possible cause I just hacked the picker out of first principles.

# ¿ Mar 24, 2016 02:14

zedprime: Jun 9, 2007; yospos

RattiRatto posted:

This is great. Thanks for the idea. Maybe you have some blueprint strings you want to share? That would cut in half the work

http://pastebin.com/8TSUhSuT
That's the basic picker block I copy and pasted around. It basically just needs a wide red network connection supplying:
A: This comes from a Decider Combinator, which is fed by enough Arithmetic Combinators to sum up all components, and the Decider checks if everything is there (set to 15 in the construction robot example). You can feed the Arithmetic Combinators summing up with the wide red network
B: This comes from a Decider Combinator checking the count chest >0

The Arithmetic Combinator placed by the blueprint needs a new Output signal (Green in the blueprint) every placement to avoid crosstalk. Change the Smart Inserters looking at steel to look at whatever you are picking, and the upper inserter value (steel=2 in the blueprint) to however many you want to pick in a batch. Check the screenshots for the resting token position of the pipe and wall piece since they don't come along with the blueprint.

# ¿ Mar 24, 2016 12:04

zedprime: Jun 9, 2007; yospos

SneakyFrog posted:

.... so.. I have to role-play as trump? waitaminute...

You've been Trump from the beginning. You're a lone captain of industry, bootstrapping himself to a vertically integrated monopoly on smog, weapons, and rocket production. OK, space rockets are a little out of character but its still a small step to slaughtering the locals wholesale.

# ¿ Mar 24, 2016 14:02

zedprime: Jun 9, 2007; yospos

Who wants overcomplicated inventory control? I've got overcomplicated inventory control.

# ¿ Mar 24, 2016 15:30

zedprime: Jun 9, 2007; yospos

Node posted:

What is the basic solution to overflow on transport lines? I have a loop with coal, iron, and an assembly plant making gears, but the transport lines are going to be clogged up and have coal going to places where it won't be used (tech labs and such) and I'm not sure how to automate a solution to that yet.

75% of the time you only want one thing on each belt. The last 25% is dedicated local routes such as for science where it goes to a big science farm or production block where everything can use both things on the belt. Keep the sides segregated in that case.

Manage belt access by rushing logistics to get long inserters, underground belts, and splitter belts. Keep your forges clean: coal comes in on a belt, ore comes in on a belt (important for iron ore, KEEP IT ON THE SAME SIDE AS THE COAL BELT or else bad things happen when you research steel), a belt of plate comes out.

Keep trunks of common raws like iron and copper plate. Use splitters and underground belts to split off a branch from the trunk for a manufacturing line. If you are making an intermediate like gears, keep in mind how it might get ratioed out. Like for example if anything uses 1 gearwheel and takes 0.5 crafting time, you can make its own dedicated gear assembler and feed directly from the gear assembler to the product assembler without putting it on a belt. One gearwheel assembler can feed multiple red science assemblers, so you can stick it on a dedicated local belt headed to a string of the things.

# ¿ Mar 24, 2016 18:04

zedprime: Jun 9, 2007; yospos

SneakyFrog posted:

untill you start getting the hang of the more weird ways to manipulate belts to put things on specific sides, split/combine things and such I wouldnt mix belts for a good while unless its multiple parts for one assembler

Ask me about how somehow steel got put all throughout my loving factory resulting in chopping half the lines.

That's the safe advice as you can probably launch a rocket with no mixed lanes ever. But its worth taking your medicine sooner rather than later because automating things with 4 ingredients takes some hijinx if you don't have shared belts.

# ¿ Mar 24, 2016 18:20

zedprime: Jun 9, 2007; yospos

That's a lot of real neat high level stuff that I think I have applied accidentally in places and never learned how it might be better than ways I normally do it. I think I'd mostly settled on coal and ore co-sided because I never actually learned the trick to why that inserter placement you've got works over ones that don't.

A lot of my building ends up informed by being incredibly space conscious, maybe because I played Big Pharma first before ever touching this game. So even though I have the whole map to make big bus trunks and very orderly manufacturing lines branching off, I get really dense clusterfucks of woven belts and condensing things into the bare minimum belt footprint around an assembler. My rocket launch was almost entirely predicated around half of my factory being off at any given moment due to inventory management (the normal way with chest slots, not that logic car crash) or accidentally purposeful bottlenecks.

I've started working on an RSO game with the intent of getting really big wads of materials moving around and seeing how much ~~smog~~ ~~weapons~~ rockets I can really cram into the world.

# ¿ Mar 24, 2016 19:38

zedprime: Jun 9, 2007; yospos

Node posted:

Can biters spawn from anywhere to attack your base or will they always come from the direction of a spawner? I just lost a few buildings to them and want to know how to plan my defenses.

And I hope they don't swim.

They hammer out a few discrete paths which help with planning even if you don't have natural terrain bottlenecks. If there's multiple spawners in a nehigborhood they go to a rally point first and path from the rally point to a pollution maxima, trimming 360 degrees of possible incoming biters into a handful of approach vectors to focus your defense efforts. If they ever come across a military building like a turret or radar on their path to the pollution maxima, they get distracted and want to kill it first so you can use that to further magnetize hordes into well defended pillboxes.

# ¿ Mar 25, 2016 01:12

zedprime: Jun 9, 2007; yospos

SneakyFrog posted:

so, if i have a few disposible forward attack points that will keep them busy for a bit?

If you make forward bases, you need them to win. IIRC if they run out of military targets after going into combat mode, they will eat each and every man made object between the battle and the pollution maxima.

Belting ammo out is kind of a pain for anything but the nearest resource gathering ops, but its fairly viable to large power pole power out to a laser farm 100yds downrange of a rally point and just pen them in at the source.

# ¿ Mar 25, 2016 01:35

zedprime: Jun 9, 2007; yospos

Loopoo posted:

How do I mass designate a smart chest / cargo wagon to only allow certain goods? Surely middle clicking and setting each and every box isn't the only way to do it? Is there any way I can set one cargo slot and then drag it across to paste it onto others?

Reason I ask is because I've got a coal and copper mine right beside each other, and I figure I can use one train with 3 wagons to take in both copper and coal. Copper gets dropped off on the way at my brand new electric furnace facility, and then it can carry on and drop the coal off to my central smelters in my original base.

I think easiest practice to do exactly what you want is designate the order of wagons in a dedicated way and set up the stops so the correct wagon is loaded at the correct stop.

Best/full transport tycoon practice is make a signalled train network and run a dedicated train to each resource.

Wagons should count in the logistic network? Maybe finagle a logistics network for each stop, and set smart inserters to pull below a ceiling to prevent filling a wagon.

e. wait logistics wouldn't work unless you range the load chests out of the logistics network which would be an awful pain.

zedprime fucked around with this message at 02:08 on Mar 26, 2016

# ¿ Mar 26, 2016 02:06

zedprime: Jun 9, 2007; yospos

Balancing bus trunks seems like handicapping . I suppose the point of a saturated bus system is to avoid accounting like this, but why not a tree system?

You have saturated bus lanes A, B, and C. You split A into A-local and A-continuing, both perfect halves of your total lane. You use A-local for your 6/13/20 item/s sinks in the area, maybe splitting them in half once or twice or more to ration to the local production lines as appropriate.

Now you have saturated lanes B and C next to half full A-continuing. You split B into B-local and B-continuing, same as above. But now you merge A-continuing and B-continuing into the now saturated AB and continue on with saturated lanes AB, and C.

Not as immediately infinitely stampable. But pulling from saturated belts gives you very predictable item delivery rates. So it can keep you honest about what your bus capacity really is instead of getting a few drops in and start pulling from a flaccid tailing end of a bus because your first few drops really needed all that could be supplied from a single splitter.

# ¿ Mar 26, 2016 14:12

zedprime: Jun 9, 2007; yospos

Indecisive posted:

You will never have 4 saturated belts if you're taking material off though, so why run 4 belts that are half full when 2 full belts is equivalent?

Yeah, that would be the idea, with the intention of not surprising yourself about the real throughput of your bus when oh golly the Nth green circuit machine just turned on and now the belt assemblers are starved of iron plate.

Belts and splitters entitlements can get lost in using the bus as a giant warehouse for WIP. The tree idea is over designed for the sort of production most people do, but if you find your factory calling your bluff about what can actually spit out the option is there to do the production accounting and keep everything full when a big user spins up. A tree would help production accounting because the entitlement calculation is very easy: a full belt gives half a full belt out of a splitter.

Dirk the Average posted:

Because you're not always running all parts of your factory at once. If an early part of your factory shuts down, it's useful to have the rest of the plate move on to the rest of the factory.

At the very least you have known sinks in rocket parts, and research if you have an endless research mode.

Solumin posted:

Sure, the tree could work. Is this a design you've used before?

Nah I just vomit ouroboroses with the promise of cleaning it all up later with blueprints, and then never doing so.

# ¿ Mar 26, 2016 16:26

zedprime: Jun 9, 2007; yospos

Jabor posted:

More to the point, you could just remove the excess belts without any negative impact.

Before anyone asks what the harm is in extra belts, a fast belt is 1/1000 of the iron plate necessary for a rocket launch, and an express belt is 1/500 :spergin:

Loopoo posted:

I love this thread. Your designs are amazing. Also, I think your arguments about the 4-belt bus are invalid. The guy said he uses splitters after a belt branches off to an assembly line to even things out, and more often than not, parts of your factory work slower or shut down completely, which means you'll end up having fully saturated belts.

On top of that, if your production is more than your consumption, you'll always have fully saturated belts. If I were to use the 4-bus method or the tree-branch method, I'd go for the 4-bus.

Always is a strong word. If you have two active supply drop splitters in a row, the second splitter is seeing slack and can't reach entitlement. How big a deal that is depends on how full out you want to run your supply drops. Closer to full out means less belts, and less belts, well see above.

# ¿ Mar 26, 2016 17:02

zedprime: Jun 9, 2007; yospos

Mindblast posted:

I'm not sure how that setup properly divides into 4 lanes, since some routes through it cause the items to move through more splitters than other routes.

That's the practical method of throw splitters at it vs bulkier perfect balancing schemes as explained in this reply last night

# ¿ Mar 26, 2016 21:11

zedprime: Jun 9, 2007; yospos

Jamsque posted:

I've only been messing around with this game for a couple of days, the pictures of bases in this thread are really intimidating.

Should I be aiming for that kind of bus-centric setup from the very beginning? On my best run so far I had red and green science production going as fast as eight labs could consume it, and I started to think about replacing all of my burner mines with electronic ones and I couldn't face it. My setup of one long belt of coal going through iron and copper fields (drills feeding straight in to furnaces) and terminating at my generators didn't seem upgrade-able. I can see how having smelting separated from mining would help with that upgrade path but I can't figure out how efficiently feed smelters with coal AND ore while still having space to take out the plates.

The burner stage is a trap, use burner equipment solely with the purpose of getting materials for your first electricity generator. Research Automation and you get electric long handed inserters which should provide a major aha on how to service furnaces.

There's no penalty to unbuilding everything besides the time it takes to do so, you get everything back 1:1.

The bus is a powerful organizational tool. There's also nothing inherently wrong with spaghetti, assuming you can keep track of it and weave things where they need to go.

# ¿ Mar 26, 2016 23:05

zedprime: Jun 9, 2007; yospos

LLSix posted:

Yep, that's my problem. I'm playing with RSO so I have a million miles of train tracks. Is there any way to fix the deadlock besides adding in a rail signal every other screen or so? I think I'm going to need chain rail signals to handle T-junctions, right?

pro-tip: always unload trains L/R. If the train is aligned Top to Bottom you can't fit as many inserters and so can't support as many smelters per unloading station. I really wish I'd set my unloading station up R/L now but I didn't realize how big a difference it makes.

You've got a personal rail-layer engine for yourself on a siding, right? Catch a ride and lay down signals at 140 kph. OK, maybe not full speed, but it lets you go faster than on foot and keeps things pixel perfect unlike a buggy. In the future include signals in your blueprint.

Chain signals are best practice before a junction to keep stuff from parking inside the junction, yeah.

# ¿ Mar 27, 2016 01:12

zedprime: Jun 9, 2007; yospos

LLSix posted:

My wife and I launched our first rockets yesterday too. The rocket silo builds rockets at 1% per every set of ten inputs, so you need 1,000 of all three inputs. We just set up 5 of each type of input and let it run for a 1-2 hours.

Probably going to start a new game later this week, but I don't know if we'll use RSO again. The trains are pretty cool, but the long distances make oil a giant pain. Is there a way to increase oil and biter frequency with RSO? Biters only got up to the medium sized ones in our play through.

The surfaced settings in config.lua don't break it out into individual resources. If you get into resourceconfigs\vainilla.lua you can fiddle with the raw settings for oil until you get a map you like. Same with resourceconfigs\vainilla_enemies.lua for biters.

# ¿ Mar 28, 2016 16:16

Adbot: ADBOT LOVES YOU

# ¿ May 11, 2024 14:17

zedprime: Jun 9, 2007; yospos

The Bob's Mod production trees seem like something I would be really into in a game that had Anno trappings instead of a strange survival game cum belts.

# ¿ Mar 28, 2016 23:58

The Something Awful Forums > Discussion > Video Games > Factorio: Captain Planet doesn't stand a chance.

«‹›16 »