|
The biggest Factorio pro-tip is to press alt.
|
# ¿ Sep 17, 2017 15:00 |
|
|
# ¿ May 9, 2024 12:10 |
|
Coal isn't really the issue - more important is the fact that production is being capped at half a belt worth of iron, since there's only half a belt of iron ore going in. If you split the inputs first, and only then put them onto the same belt as the coal, you have a full belt worth of input ore (and hence can produce a full belt of output). As an aside, lane balancing a single belt doesn't really have an practical purpose, it's purely aesthetic. Once you start having multiple lines of something and want to merge them down it starts actually mattering.
|
# ¿ Sep 21, 2017 00:10 |
|
Thotimx posted:But there's half a belt of ore on each side. That's more than enough to supply the furnaces to saturate a single output belt, it's not capped at anything other than the speed of the belts(once enough furnaces are in). To put it another way, the furnaces will run out of room to put plates on the central belt before they run out of ore supply. Immediately before the splitter, you have half a belt of ore. It doesn't matter how much available throughput there is after that, the bottleneck is what limits the overall throughput of the system. For a similar reason, lane-balancing a single belt doesn't help with throughput. It's still limited to the throughput of the belt at the moment just before the lane balancer. Lane balancing becomes useful when you're combining multiple partial belts together, since you then don't have that bottleneck.
|
# ¿ Sep 21, 2017 00:41 |
|
Thotimx posted:Totally agree as far as it goes, except for the very last part. What limits the throughput of the system isn't the ore/coal belt at all; it's the amount of smelted iron plates that can fit on the one belt being filled from both sides. One ore turns into one plate. If you have less than a full belt of ore, you're not going to get a full belt of plate out of it. It might look visually like you have a full belt once everything backs up, but that's a purely aesthetic thing - once you start actually using iron, you won't be able to continuously supply more than half a belt of it. quote:I think maybe I'm using the wrong words here. Possibly throughput isn't the right one. Let me explain differently what I mean(and then feel free to tell me I'm still wrong :P). What you're describing is buffering - making it so that more currently-unused stuff can fit on the belt. Note that your throughput of coal is still just a single miner's worth. The thing about buffering is that it's actually a bad thing in most cases - for example, suppose you add another consumer of iron plates, that means you're using up more iron than you're producing. Without buffering, you can visually see that now you're not getting enough iron, prompting you to go and improve it. If you have a large buffer, though, what happens is that your buffer silently starts shrinking, and after you go off and do something else you come back and wonder why nothing in your factory is getting done. Small buffers are useful to smooth things out if you have inconsistent supply or demand (think buffer chests at a train stop), and larger buffers can come in handy if you explicitly want them (for example, you can buffer coal for your steam engines, and use combinators + a speaker to give an alert when it starts decreasing, giving you enough time to fix it before you go into a full brownout spiral). But buffering more on belts just to make them look even isn't a good thing. Now, if you're doing a speedrun, big buffers are really helpful because you know exactly how much stuff you need and you want your machines working all the time to produce that stuff. But that's not what most people are trying to do in the game.
|
# ¿ Sep 22, 2017 01:37 |
|
Thotimx posted:I'm very aware of that. Went back and read some things though, and I think I understand where the screwup was. Changed the setup to this: Yep, that's much better - no bottleneck any more, you have a full belt of ore -> two half-belts of ore -> a full belt of plate. I'm sure you'll be able to compact that more when you need the space for more blocks of furnaces (unless you were going to rush electrics?) I wouldn't worry about the lane-balancing thing too much - it's mostly just a pet peeve of mine, and I was mainly objecting to the claim that lane-balancing a single belt improves throughput. Certainly, lane-balancing doesn't actually hurt throughput at all, and if you're okay buffering a little more for aesthetic purposes that's just fine. (As an aside, lane-balancing a single belt is actually useful later on - but I'll mention it again when we get that far, unless someone else points out out first.)
|
# ¿ Sep 22, 2017 02:58 |
|
The ore patch has the same amount of ore in it regardless of how you space out the miners. There's some merit to spacing out the miners on your first patch, since that means you need fewer miners (and hence a smaller resource investment) to cover the whole thing and then you can forget about it. Once your factory is big enough that you're expanding to extra ore fields, though, getting twice as many miners on a field basically means you need to expand to half as many fields (or at least, you get to put it off until later). The big upside to dense miners, though, is that they're way less fiddly to place.
|
# ¿ Oct 3, 2017 07:54 |
|
Psychotic Weasel posted:A person I learned to play the game from used a somewhat hybrid approach to miners, I guess you could say. They didn't jam them all together but they didn't spread them far apart to make sure everything was perfectly covered, either. Each drill was placed one tile apart in each direction to form a perfect square grid with some overlap since each miner can reach one tile out in addition to whatever is directly underneath. I find it's a fairly good compromise and allows you to mine resources at a good rate without having to figure out and hand place each one. Usually you don't place drills right next to each other, you leave a one-row gap in front (for the belt) and behind (for power poles), while having them butt up against each other on the sides. Having them butt up against each other makes it really easy to place down a bunch - just click and drag. Placing extra drills is only wasteful if you're not using them (and even then, you're only "wasting" the relatively tiny amount of resources it takes to make a drill, and a little bit of idle power drain) - if you're using X drills worth of iron ore per second, it's actually less wasteful to have them all on one patch than it is to go out and find other patches for the extra drills.
|
# ¿ Oct 3, 2017 13:25 |
|
The trick for adding hysteresis is that you need some way to actually store information - the output of combinators only depends on their input, they don't have any inbuilt storage. The secret is to wire the output of a combinator back to its own input. If you pick a signal, let's call it S, then you can make a decider combinator that outputs 1 S when S>=1. If you wire that up to itself, it will "turn on" when you send it a 1, and then "turn off" when you send it a -1. This is one of the places where having multiple wire types comes in really handy - you don't want to pollute the internal logic of your combinator circuits with signals from unrelated circuits.
|
# ¿ Oct 9, 2017 01:24 |
|
Alavaria posted:Sampling with or without replacement... The number of individual "bits" of uranium in any sizeable chunk of ore is so huge that it's almost off the top of the scale of SI units. It's a mind-bogglingly large amount. Suffice to say, the law of large numbers is very much in effect.
|
# ¿ Nov 2, 2017 12:35 |
|
As long as you've got enough buffer, you're probably not going to get rng-screwed. Even less likely if you have surplus refining capacity. You could set up a reactor (remembering to leave space to expand it to at least 2x2 later on) right now to handle literally all your power consumption, and you'd still be saving up U-235 for later. The other way of making U-235 is mainly useful when you get up to the other thing that uses U-235. -- also, the ~optimal~ reactor layout is a 2xX block long enough to meet your factory's power needs.
|
# ¿ Dec 21, 2017 01:32 |
|
EponymousMrYar posted:Chain signals in the ring don't work unless your roundabout is friggen' ginormous so that they don't read both the next segment of the roundabout AND the exit. Chain signals absolutely solve this though? The whole point of using chain signals is you want them to find a clear path all the way to the exit before entering the roundabout. So when you have two trains making right turns across each other's path, one of them will go first, while the other will wait outside the roundabout (hence not blocking the first train's path) until there's room for it to make its own turn.
|
# ¿ Jan 21, 2018 14:22 |
|
Some people in here have a weird-rear end interpretation of how chain signals work.
|
# ¿ Jan 21, 2018 23:53 |
|
First image seems to be a duplicate of the second - presumably you were intending to show the pipe research option there?
|
# ¿ Jul 8, 2018 07:09 |
|
|
# ¿ May 9, 2024 12:10 |
|
Thotimx posted:So … Personally I don't see anything wrong with belts gradually getting phased out of a factory in favor of bots as you progress later into things. The reason it's terribad from a game design perspective is that it removes literally all the logistical challenge from the game, and all that's left is an endless sea of beaconed assemblers hooked up to logistics chests. In order to really scratch the engineering itch, putting time and thought into coming up with a clever solution should result in something that's better (in at least one dimension) than the no-effort obvious solution. If the no-effort obvious solution is better in literally every way than any other solution, it's all kind of pointless.
|
# ¿ Oct 9, 2018 07:45 |