|
FISHMANPET posted:Why would you want to do that exactly? That swarm of bots going along a single path is going to backup every roboport in range as they'll all be recharging at once, rather than having it spread out. I just don't like the idea of my guys getting sent out every tick to do poo poo. I like watching the big line fly through.
|
#
?
Aug 31, 2016 19:39
|
|
|
|
| # ? Apr 29, 2024 13:58 |
|
|
Ratzap posted:Don't keep as much acid. You really do not need 10,000 as a buffer. Disconnect input to the acid storage and let the production drain it down then remove some tanks. 1 tank should be plenty and you'll want to divert some sulfur to explosive production (for cluster grenades) but it won't take much.
|
|
#
?
Aug 31, 2016 19:40
|
|
|
Cosmik Debris posted:I just don't like the idea of my guys getting sent out every tick to do poo poo. I like watching the big line fly through. Philosophically, batching is usually a last resort. You batch processes when there's no other choice. A lack of capacity, the nature of the process, so on. In Factorio any lack of capacity is your own personal fault because its fairly cheap to build new belts or robots or what have you until everything moves continuous like the Lean god intended.
|
|
#
?
Aug 31, 2016 19:45
|
|
|
Doesn't it waste their capacity though if a robot that can carry multiple things keeps getting assigned to carry 1 thing every time an item is removed from a requester chest? Shouldn't they ideally trigger a request when X items have been removed, where X is the maximum capacity of a robot?
|
|
#
?
Aug 31, 2016 19:58
|
|
|
Cosmik Debris posted:Doesn't it waste their capacity though if a robot that can carry multiple things keeps getting assigned to carry 1 thing every time an item is removed from a requester chest? why is it even an issue man I always have like 1000 logistic bots, 1 single bot wasting it's time with 1 item isn't a big deal. Especially due to the fact I don't keep items that are used together on opposite sides of the factory. Everything is centralized so even if a bot is carrying 1 single item, it takes less than 30 seconds.
|
|
#
?
Aug 31, 2016 19:59
|
|
|
It's real dumb and there's no good reason to do this, but I GUESS you could: Have your requestor chest that requests 1000 steel. Put 2000 steel in it. Have an inserter connected to the chest with circuit wire and have it only pull items from the box when it has more than 1000 steel. Put another inserter with another chest. Connect it to your requestor chest with circuit wire, and have it pull everything out if the requestor chest has less than 1000 steel. Put a chain of inserters and chests so the second chest can empty back into your first chest when the requestor chest is over 1000. So basically you have 1000 dead steel moving around to "trick" the requestor chest. It's really dumb and you really shouldn't do it. E: They always carry a full load, which is 4 (unless the container they're pulling from has less than 4 of the item in question) Try this: Pick something you have a lot of in your logistics network that has stacks of 50 (coal or ore, not plate or steel). Put a requestor chest and request 50 items. Wait and see how many get scheduled for delivery. It's not 50, it's 52. Because it delivers 13 full loads, not 12 full loads and one half empty loads. Your bots aren't buzzing around carrying single items. FISHMANPET fucked around with this message at 20:03 on Aug 31, 2016 |
|
#
?
Aug 31, 2016 20:00
|
|
|
Cosmik Debris posted:Doesn't it waste their capacity though if a robot that can carry multiple things keeps getting assigned to carry 1 thing every time an item is removed from a requester chest? My impression has been that bots seem to go pick up their inventory limit of whatever material and then drop it off, often leading to requester chests that are 2-4 items over their request limit.
|
|
#
?
Aug 31, 2016 20:02
|
|
|
zedprime posted:Sulfuric use is vanishingly small, and sulfur in general is a constant demand beside weapons and accumulators depending how you handle power gen capacity. I usually limit acid production to a few hundred buffer in a tank, and let sulfur build in a chest when plastic is experiencing a downtick. Bleh, thanks.I'll cut down the sulfuric acid tanks and put a circuit limit on the one left when I get home. These ratios are things I wish I had a better grasp on, the answer of "how much production can one sulfuric acid plant support" would have been great to know before I built 10ku buffer for it.
|
|
#
?
Aug 31, 2016 20:18
|
|
|
LonsomeSon posted:My impression has been that bots seem to go pick up their inventory limit of whatever material and then drop it off, often leading to requester chests that are 2-4 items over their request limit. oh ok that makes more sense then
|
|
#
?
Aug 31, 2016 21:19
|
|
|
Ciaphas posted:Bleh, thanks.I'll cut down the sulfuric acid tanks and put a circuit limit on the one left when I get home. All the information is available for you in game: - Sulfuric acid takes 5 sulfur + water -> 5 Sulfuric acid in 1 second - Chem plants have a crafting speed of 1.25, so that's 5 acid in 0.8 seconds (6.25/s) OK, but how many sulfur plants do you need to support that? - Base recipe: 3 petro gas + water -> 2 Sulfur in 1 second - At 1.25x crafting speed, that's 2 sulfur in 0.8 seconds (2.5/s) - The sulfuric acid plant needs 6.25 sulfur/s, so slap down 3 sulfur plants (7.5/s) to feed that one sulfuric acid plant with a bunch of sulfur left over, perhaps for making explosives (which needs 0.25/sec) - Or you could limit it to 2 sulfur plants (5/s) and get 5 acid / sec. The 1 sulfur : 1 sulfuric acid ratio in the acid recipe makes it really easy to scale as needed. So what can you do with 6.25 sulfuric acid per second? - Each battery factory (in a chem plant!) needs 2 acid / 4 sec, or 0.5/sec. I have a quite neat-looking blueprint for a 2 battery/sec layout that uses 2 sulfur plants : 1 sulfuric acid plant. Throw in another sulfur plant and 4 more battery plants and you get 3 battery/sec easily. - Each processing unit needs 0.5 acid / 15 sec (0.03/sec), or 0.5 acid / 20 sec if you're using Assembling machine 2s and 0.5 acid / 12 sec if you're using Assembling machine 3s. You really don't need a lot of sulfuric acid if you aren't crafting multiple batteries per second or an obscene number of processing units. Of course, you can avoid all this and just use a tool like Factoratio or Factorio Production Planner. Solumin fucked around with this message at 22:19 on Aug 31, 2016 |
|
#
?
Aug 31, 2016 22:16
|
|
|
I suppose I meant more that I don't have a head or the attention span for the numbers, rather than their lack of availability  Also my lack of immediate knowledge of what each item is an ingredient for--I was telling myself "what if sulfuric acid is used in a lot of poo poo", when it turns out it's only good for batteries in bulk and in small amounts for processors.Thanks for the explanations, though. I'll try to focus more on that stuff in the future. Some ten years ago I would have been a loving wizard at mathy stuff like this 
Ciaphas fucked around with this message at 22:30 on Aug 31, 2016 |
|
#
?
Aug 31, 2016 22:22
|
|
|
Ciaphas posted:I suppose I meant more that I don't have a head or the attention span for the numbers, rather than their lack of availability Don't worry about it too much. It's good to have a general idea of ratios in the game (like that a sulfuric acid plant will make a fuckton of acid) but sitting down and doing the math is tedious. Usually I'll start off doing that for a factory, and then it becomes "Oh I need to mass produce <something>, I'll put one assembler here and this there..." and just slap things down with abandon. What I'm trying to say is there's nothing wrong with caring about the numbers or not, but at some point it helps to have a general idea so you can figure out why you have bottlenecks in your factory.
|
|
#
?
Aug 31, 2016 22:29
|
|
|
Before I left for work I resolved to go find more oil to hook up, on the line of thinking that "I have enough petroleum to fully support either sulfur production or plastic production but not both, and I don't have enough incoming oil to support more refineries for more petroleum, so MORE OIL" which would solve the bottleneck just as well, I reckon. But yeah I can certainly take acid production offline for a long time at this point, now that I have a better idea what it's used for and in what quantities. I only ever seem to remember particular ratios, for some reason. I always remember 1 pump->14 boilers->10 engines, 2 copper cable->3 green circuit, and 1 copper cable->8 red circuit, for example.
|
|
#
?
Aug 31, 2016 22:33
|
|
|
I also need to work out how to do train junctions. My loop is getting rather large and unwieldy, and some trains certainly don't need to travel all of it. So I'll probably be spending part of tonight having a goddamn fit over how to do the signaling
|
|
#
?
Aug 31, 2016 22:37
|
|
|
I only ever worry about ratios on a set of core material sinks. That is, red science, green science, and blue science in the first wave, and rocket parts in the second wave. To that end I have spreadsheets I use set up for the key bottlenecks in operating 5 red science assemblers, 6 green science assemblers, and 8 blue science, as well as some consumption numbers for rocket parts for the second wave. I once started to do the math to perfectly ratio out what I would want for building materials, but its such an easy thing to eyeball because generally one assembler making belts or a couple making solar panels or accumulators or whatnot makes it faster than I can consume, so I just make sure there's a full bus of essentials and carry on. It lives on next to the bigger deal math for the science packs, but its more of a benchmark if I look and see golly this factory isn't working quite right. By the time you are talking about power armor gear, its a temporary sink that you'll demolish before you blink so even if its starved who cares.
|
|
#
?
Aug 31, 2016 22:54
|
|
|
Any particular reason you say 5, 6, and 8 assemblers for red green and blue? Something about their consumption rates or production speeds?
|
|
#
?
Aug 31, 2016 23:02
|
|
|
5-6-12 is actually the better ratio for science. While there are a variety of recipies with varying levels of needs for specific types of science, aiming for recipies that require equal amounts of each is a good baseline. So red science takes 5 seconds per pack, green 6 seconds, and blue science 12. So if you want 1 item/second, you want 5-6-12. And when you get to purple, a single assembler with a T1 speed module in it.
|
|
#
?
Aug 31, 2016 23:05
|
|
|
Ah, that makes more sense. I thought blue were consumed more slowly than green and red when zedprime said 5-6-8. New note to self on list, redesign the science labs to adhere to this ratio too God drat this game, I spend way too much time on it every time i dive in  
|
|
#
?
Aug 31, 2016 23:07
|
|
|
Ciaphas posted:Any particular reason you say 5, 6, and 8 assemblers for red green and blue? Something about their consumption rates or production speeds? 5-6-12 is the equal production ratio when everything is in assembler 2 or assembler 3. e. I am probably a bad factorio player because I take so long in replacing assembler 1s that don't make green circuits. zedprime fucked around with this message at 23:12 on Aug 31, 2016 |
|
#
?
Aug 31, 2016 23:08
|
|
|
Yeah I pretty much build the minimum number of Assembler Is to allow me to build the IIs.
|
|
#
?
Aug 31, 2016 23:14
|
|
|
Ciaphas posted:I also need to work out how to do train junctions. My loop is getting rather large and unwieldy, and some trains certainly don't need to travel all of it. So I'll probably be spending part of tonight having a goddamn fit over how to do the signaling You're using 2 rails for bi-directional trains, you have plastic and therefore red circuits -> blueprints so you have all the parts. You just need to make 2 specimen junctions and blueprint them. Even if you don't have robots yet, you can use the blueprint ghost as a well, blueprint, to lay things out yourself. If you struggle with the signals, use the examples from the rail gallery (http://imgur.com/a/CxXxd) to get it right.
|
|
#
?
Aug 31, 2016 23:14
|
|
|
FISHMANPET posted:Here's pictures of the basic blueprints if you want to import them "manually" Just quoting myself for some good clear pictures of junctions
|
|
#
?
Aug 31, 2016 23:19
|
|
|
Thanks for the gallery link. When I get home if things in the factory are stable I'll probably just lay down some track and a couple engines totally separate from my rail line, put down signals, and see how the lights change as I move the engines. It seems like just putting signals down on both tracks right after a split and right before a join would cover everything, but obviously not else chain signals wouldn't be a thing. (edit) And of course I can't get to the factorio forums at work 
|
|
#
?
Aug 31, 2016 23:20
|
|
|
Basically, when you are placing a signal, you want to pay attention to the train ghost it makes. This will give you a rough idea of where to place the signal BEHIND that signal. As in, if the signal you are placing is red, this is where a train would be on the tracks that it is stopping. Placing a signal before that will ensure that a single stop somewhere will not make multiple signals red. If you apply this to your junctions, you can then see when and how one signal being red (because a train is in the block ahead) will make a different signal red, causing a block. Or you can just run trains, watch them for a while and add/delete/move signals as you find inefficiencies.
|
|
#
?
Aug 31, 2016 23:29
|
|
|
Decided to take an early out, since nothing's going on at work today, and well i am absolutely confident that this will not kill nor maim anyone
|
|
#
?
Sep 1, 2016 00:05
|
|
|
You could have some issues unless you drop in some chain signals at the entry lines there.
|
|
#
?
Sep 1, 2016 00:49
|
|
|
Here's a nice refinery layout with room for beacons:
|
|
#
?
Sep 1, 2016 00:55
|
|
|
Slickdrac posted:You could have some issues unless you drop in some chain signals at the entry lines there. I'm gonna leave it until there's an issue, tbh, because I want to SEE it happen to better understand the whys and wherefores Unless I get time to make a demonstration junction for myself and I got too many other things to do to be bothered~
|
|
#
?
Sep 1, 2016 00:55
|
|
|
My god first it was the petroleum, now it's the green circuits At least those are easy to expand, I suppose
|
|
#
?
Sep 1, 2016 01:41
|
|
|
Ciaphas posted:My god first it was the petroleum, now it's the green circuits You say that, and then suddenly you find you've run out of copper and iron.
|
|
#
?
Sep 1, 2016 01:49
|
|
|
the thing about ratios is that worrying about them presupposes that the assemblers are expensive they are fuckin cheap plop down whatever and get on with it!!
|
|
#
?
Sep 1, 2016 01:52
|
|
|
Ciaphas posted:I'm gonna leave it until there's an issue, tbh, because I want to SEE it happen to better understand the whys and wherefores Three ways are slightly harder to gridlock to leave evidence. It might end up being fine in a not so dense neighborhood. Depending how many trains could end up in one direction compared to the other you might lock, but for all but the most malevolent design you will probably just have extra traffic jams depending on how well behaved traffic is downstream in any given direction. Lets look at an example of the literal worst way to build a three way  A three way intersection only services a precariously placed station on one of the directions. Trains can't wait to go and are lining up by the droves. One makes it to the station, one occupies the block just before which happens to be mid intersection, and one occupies the block before the intersection.  Oh no! A lack of chain signals shows its face. The train in the top block can't enter the intersection to vacate the top block. The train in the intersection can't vacate the intersection block because its on track to the top block. Besides completely halting traffic to the top station and back, anybody who wants to avoid this mess and go south can't because that third train is already in the way. The grid is locked and you must take manual control. Chain signals make trains more predictably dumb  If you put some chain signals at the entrance to the intersection, they take a look and say hey, are there any red lights in there? If so, think twice. In a branch they'll turn blue if a train is fine to go in one direction, but would be stopped in the other. The train pathfinds and says hey, this blue is really a red because there's a red ahead and doesn't enter the intersection. A train leaving from the top station will catch its green, make its turn, vacate the block, and once it clears the intersection our waiting train will catch its green and make it through the intersection and to its destination station. Baloogan posted:the thing about ratios is that worrying about them presupposes that the assemblers are expensive
|
|
#
?
Sep 1, 2016 02:08
|
|
|
Solumin posted:You say that, and then suddenly you find you've run out of copper and iron. This literally just happened except the coal also ran out Said coal was still feeding my power factorissimo station Needless to say that was a scrumble to fix
|
|
#
?
Sep 1, 2016 02:12
|
|
|
also just wanna say if you ARENT using factorriosssimo mod u have no soul
|
|
#
?
Sep 1, 2016 02:16
|
|
|
Ciaphas posted:This literally just happened except the coal also ran out Green circuits are a hungry, voracious beast.
|
|
#
?
Sep 1, 2016 02:48
|
|
|
Solumin posted:Green circuits are a hungry, voracious beast. Just think how he'll react when he realises each purple circuit costs 20 green plus the red and acid. Making any amount of those things absolutely devours green circuits.
|
|
#
?
Sep 1, 2016 03:04
|
|
|
I keep coming back to my existing factory save with the intention of improving it, then scaring myself back off again with the magnitude of what I need to do - I'm basically down to just "launch a rocket" stages but somehow it still seems so huge. Vaguely considering restarting, but maybe I'll start expanding resource processing in my existing game in order to get this cooking... Can anyone point me toward a straightforward, tile-able, Advanced Oil Processing layout?
|
|
#
?
Sep 1, 2016 03:26
|
|
|
Ignoranus posted:Can anyone point me toward a straightforward, tile-able, Advanced Oil Processing layout? Phssthpok posted:Here's a nice refinery layout with room for beacons: Each column is tile-able straight down. There's plenty of room on the light oil pipe to add solid fuel plants.
|
|
#
?
Sep 1, 2016 03:30
|
|
|
Ignoranus posted:I keep coming back to my existing factory save with the intention of improving it, then scaring myself back off again with the magnitude of what I need to do - I'm basically down to just "launch a rocket" stages but somehow it still seems so huge. Vaguely considering restarting, but maybe I'll start expanding resource processing in my existing game in order to get this cooking... Can anyone point me toward a straightforward, tile-able, Advanced Oil Processing layout? My first rocket was a mess of building all the parts on opposite sides of my factory and botting them to the silo. Just work on figuring out what you need to do to make each kind of rocket part and do them one at a time and you'll learn from your mistakes. Personally I think this from reddit is a pretty good newbie friendly oil setup.  Expand each line as needed rather than try and get perfect ratios. Belt the sulfur somewhere with space to make acid and batteries etc.
|
|
#
?
Sep 1, 2016 03:55
|
|
|
|
| # ? Apr 29, 2024 13:58 |
|
|
Ratzap posted:Just think how he'll react when he realises each purple circuit costs 20 green plus the red and acid. Making any amount of those things absolutely devours green circuits. I started on this pretty much as you made this post and my god what
|
|
#
?
Sep 1, 2016 04:00
|
|