Factorio Help my newbie [friend] broke his tutorial |
- Help my newbie [friend] broke his tutorial
- The Itchy Guide to keeping trains simple with station limits for absolute beginners, and what good they actually do over not using them
- When growing the factory takes ALL the roobits
- imagine wasting resources on artillery shells to make a smiley f-
- DJ Spidertron - Can't reach
- Spaghetti?
- People really liked my 120 train depot before the post was removed, so here are the rest of the designs I've made during the week
- Hi! I'm new in this game these are my first hours what do u think about my base? :)
- Clicked on the wrong desktop icon this morning!
- I just realized something very wrong with the sulfuric acid barrel
- Upcoming mod (Preview open to suggestion)
- 64 hrs in. Haven't played any other game since. Am I doing this right? Any tips?
- I'm pretty new, what's wrong with my oil setup? I'm trying to load the fluid wagon, but the oil is spread very thin
- MegaSimple long Train base.
- I wonder what will happen when i launch the rocket.
- I know it´s weird, but i like it.
- What do you most hope for in the upcoming expansion?
- Version 1.1.25
- What do you think of my circuit-free fool proof kovarex set up?
- Where is all the Station Limit hype???
- Base Planner/Mod Question (Factorio Planner)
Help my newbie [friend] broke his tutorial Posted: 18 Feb 2021 02:57 AM PST
| ||
Posted: 17 Feb 2021 11:02 PM PST Hi, I'm itchy, most people can't pronounce my username, so I just tell them to call me "itchy." And today, I've decided to put together a simplified guide to how to leverage your trains properly now that we have the AMAZING "Station Limit" feature. The core idea behind the station limit, is that you will have a finite number of trains. With multiples of each stop, such as several "Iron Plate Pickup" stations, and several "Iron Plate Drop" stations, how do the trains know to go to the farther station, and not the nearer one? Or, "Why are my trains still only going to the nearer station even though the far one is open?" That's the core concept of limits: You limit the number of trains which can seek the nearer station, so other trains are forced to go to the farther stations. So lets start at the beginning, and work our way up. I will be using creative mode and loaders to simulate production and demand, in order to have a simplified setup which is easier to understand. Trains are set to the simplest schedule, and refuelling is done at the "unload" station, as it is assumed the unload will be closer to your base than the pickup. I have a need for two stations to each get 12 belts of iron plates. What for? Doesn't matter, one's doing green circuits, the other's doing gears, lets say. Two stations, 12 bluebelts each. I know they're not fully compressed, don't worry about it, test example. The amount each station needs individually isn't so relevant, only the total demand for all stations combined. I use the term "12 bluebelts" to represent 540 items per second, a blue belt is 45 items/second. You do the math for your own factory's needs. The amount you actually need per station doesn't mean as much because trains will self regulate the time they spend at a station by using "Fill" and "Empty" conditions. Just make sure total production >= total demand, and let the trains sort things out. I'd also like to note that I have placed them directly adjacent to each other to demonstrate how powerful of an effect distance actually has on train pathing, and why limits are so important. I've added a simple counter, and wired each station so it puts train count on C: this will allow us to see how much time the trains spend at each station, though I will say that it will track trains that are seeking, as well as in the station. The stations have no limits attached to them right now, it's just your basic loop setup, with two trains, as there are two stations. Lets let it run for a while, and see if the second station gets ANY plates. Hmm, trains seem to have spent 20,000 ticks (about five and a half minutes) seeking the closer station, and exactly 0 seeking the farther one. What the problem is? Two trains two destinations, why are they at the other end? Aha, here's the problem, production isn't sufficient! 12 red belts of plates is only 8 blue belts, and we need a total of 24 blue belts. Okay, lets fix that, lets up our production and see where that gets us, we need 24 total, so lets add two more. Great, now we have 24 blue belts of iron plate production, and two trains to deliver to two stops, so lets see if that fixes the issue! Well here's some good news, at least the trains are going to the further pickups. And they're also going to the further drops! Lets see how much iron makes it to each station, because we're surely still not getting 12 bluebelts out just yet, the trains aren't always unloading like they should be. Nope, the first station is getting more, even though it's barely closer. Ahhh, here's the problem: The trains synchronize themselves, and the nearer stations always end up seeking the nearer stations. Check it for yourself, but with only 2 trains and 4 stations (or more), one of them will tend towards the inner loop, one will tend towards the outer loop, and when their timings are perfect for it, they'll flip, and the inner one will do the outer, outer will do the inner. So the inner stations will simply see more traffic. The distance is the reason for this, trains having to go further means they take more time and the schedules are not perfectly synced. The real problem though, is that no train EVER visits the farthest pickup station. One option for controlling this behavior, is to measure the amount of resources present in the chests at pickup stations. Only when there's enough do you turn the station on, and allow a train to seek it. The problem with this is, that it won't turn on, until a train should already be full. Also, you're doing work, where you should be letting the trains do work for you. You're adding combinators and conditions, when the trains are already doing all that work, now that we have limits. You can adjust the amount with circuitry so trains come earlier, but this is a very roundabout and complex solution, when limits work better. So lets add in a train so we have a train that's picking up from each station, and see if three trains is sufficient to get our 24 bluebelts of iron. Reset the iron plate counter, and see how our drop stations fare. WHAT?!?!? Adding a train seems to have not only not helped our problem, it seems to have exacerbated it! Now the first station is being preferred even worse! But the real problem has been highlighted in red: You're not getting 12 bluebelts of plates yet, trains are not entering the station quickly enough to support 12 blue belts. The third train is moving more resources along the rails, yes, but adding the train didn't exacerbate the problem, it demonstrated it more clearly. Trains REALLY prefer nearer stations. That's not a problem, that's a really good thing, because it means trains will attempt the shortest paths. You want trains to seek the nearest station, so if your pickups and drops aren't clumped like in this example, trains will prefer to pickup from their nearest, and drop at their nearest. That's great! So how do we make it stop starving the far station? Well, we solved the first issue, which is more production, now we have the second issue, more trains. Stack inserters move items into and out of chests more quickly than on or off belts, which allows us to buffer some production into steel chests and load the trains more quickly than we can load the chests, which helps because the trains won't always be in the station loading. That is the current constraint, with only 3 trains, and 3 loading stations. If you want to get all 8 bluebelts out of each loading station, you have to have a train present or moving into the station basically at all times. So lets add three more trains, and see what happens. That way there will be a train waiting to get into the loader, behind the one already in it. The rail system has completely seized because of slightly improper signalling, insufficient stacker space, and trains being stupid without the limit. The full train which is waiting for a drop, came from the third pickup. But because the train waiting for the second pickup has committed to the second, it can't go to the third. Thus, none of the empties have a place to go: Yes, I have intentionally signalled this "badly," but this problem is incredibly common and has to do with trains not being able to pull off somewhere safe. This concept is known as a stacker, and can be avoided, if you wish. However, you must still find some way to not stop trains from getting where they need to be, and in order to get full throughput, you need trains already on the rails seeking their destinations, before the train at said destination has left. To show this, lets set the train limits to 1 for the pickup stations, and fix the logjam. I'm going to change NOTHING else about this design, just set the train limits for the pickup stations to 1. Ta-da! Not only is the next train sitting there waiting, but the trains are delivering to the chests, faster than the belts can empty the chests. Therefore train throughput is greater than 12 bluebelts. So we've got what we want now, we have 3 production stations each doing 8 bluebelts, and the trains pick whatever one has an empty slot, IE there is no train currently loading. The train already moving has already decided to go to the third station, because the third station's train was full, and wanted to leave. The second pickup has just finished its fill, but does not yet have a train attempting to seek it. The next empty train will seek it. With a station limit of one train on the pickups, the only way a pickup station is a valid destination for a train, is if it does not have a train already filling up at it. This is the exact same behavior as wiring chests and only turning a station on when a certain amount has been reached, the timing is just different. Stations only "turn on" (Go white instead of blue, blue=limit full) when they don't have a train at or seeking them already, rather than when they have enough cargo. Alright, lets step things up a bit. Double the demand and make it farther away, don't double production though, lets see how things shake out. Production will be insufficient to meet demands, as demand is now 48 bluebelts, and production only 24. Innnnnteresting, so it seems the farther stations are getting some traffic, because the nearer ones have empty trains saying "Destination full." What this means, an empty train in a drop station which says "Destination full," is that you do not have sufficient production. The train has been emptied, but has no pickup available to it. Production not enough? DOUBLE IT. Not enough trains and the inner stations are being preffered? DOUBLE THAT TOO! Except it actually jammed for the same reason as before, long before I made it to 12 trains, it jammed with only two extras. "Hey Itchy I thought the limits were supposed to stop that!" Sure, but we only used limits on one end. I did this very simple track to demonstrate trains competing with each other being a problem, and the reason for stackers. So lets make it real easy, and just put one linear stacker behind each of the loading stations, see how that goes. That way, the train can get off the tracks that others want to use at least. Okay, so 12 trains and with stackers, they aren't jamming each other anymore. The trains just get to where they're going and when the full train still sitting in the station leaves, the empties pull forward to be filled. We have enough production and we should have enough trains, it was enough last time, 3 pickups 6 trains, now we have 6 pickups 12 trains. So why are the far drop stations not receiving sufficient plates, and how is there a destination full message?? This is actually a combination of two issues, both of which are common to all rail designs. The first issue is, that trains take time to get somewhere. The reason we went up to two trains per pickup station, was so that one train could be loading, and one train on the way. With a station limit of one, that's not possible, however now that we have linear waitspace, lets bump the limit to two, one train loading, one train on the way. I'll also add just a few chain signals along the upper rail to demonstrate the other problem more cleanly. Only one train can use any track at a time. The thing that's slowing down this system the most right now, is that there's only one track in both directions. I added the chain signal after the unloads, such that once a train is done unloading, it leaves the station: As you can see, there are trains waiting to use the return line. Because the rails themselves cannot handle the train throughput(because they're VERY poorly designed lol, the return line is one block, but the concept is, all rails have a limited throughput), the trains aren't making it to the farther stations before the nearer stations have an empty spot available. So lets add another track on both sides to increase throughput, and actually signal stuff better, and see how 12 trains does. It seems that there's still a very strong preference for near stations, at the expense of farther ones. A train would rather wait at a near station with an open slot, than go to an open farther station. What this means is that the far station is not actually committing its full production to the rails; Without trains to pick it up, the buffer chests become full and sit there. Decreasing the station limit to one, would function here, but it is not the desired behavior. We DO want a train waiting or en route and one loading in all of the pickup stations. We know there will be a space there, as soon as the filling train leaves. So 2 is the correct (minimum) limit. There are plenty of exceptions to "2 minimum," but for most general purpose uses of "N pickups, M Drops," you want at least two trains seeking per station. So what is the actual problem? Trains take time to get to their destinations. We fixed the throughput issue, but trains still take time. We just need more trains! We want two to be at all of the pickup stations at a time, but at least one train has to be unloading at a time, right? Well we have 6 pickups with 2 trains each for 12, and 4 drops for 4 more, lets see how 16 trains do. That doesn't seem right either? Why did those trains commit to the farther stop? The nearer stop emptied first? And more importantly why does it KEEP doing it? I added a fifth unload to drain the system of excess buffer and have since turned it off by setting the station limit to 0, so why would it do such a thing? Because for one of many potential reasons, the train decided, through its pathing algorithm, that the farther stop was more appetizing. Trains repath based on certain conditions, and that train decided it didn't want the closer stations, it wanted the farther ones. The train pathing algorithm uses A* and has several variables it looks at, including how long trains have been stopped on a path. It could shift for many reasons, but the important thing to note is that it does shift at times, and that this behavior is less than ideal: An empty unloader isn't receiving the goods it should be. So how do we fix this? Same thing as before with the pickups, linear wait space(though in practice, I STRONGLY recommend parallel stackers. If you need some help with the stacker concept, many others have done a better job of explaining it than I could), force trains to commit to their next stop, and limit the number of trains which can commit to any given stop, then make sure there are enough trains to saturate the rails. Again we'll do two trains per station, one unloading, another on the rails to get to that station. AHHHHH STILL???? Yes, but we're very close now. The trains are behaving correctly: There just isn't enough of them! That train which is waiting for an unload is doing what it should be, I WANT a train waiting there. So if the production is fine, and the rail throughput is fine, the problem is that there just aren't enough trains. Ahhhh, finally.. We have our unload stations being kept busy, and our load stations being kept busy. So what conditions were necessary, to keep all of the stations busy at all times? The simplest answer, is N-1 trains. There are four drop stations, each with a two train limit, and you can see there that I have 19 trains with that station. While five of that station actually exist, one has a limit of zero right now, four have a limit of two. There are six pickup stations, each also having a train limit of two. That means that there are 10 stations with 2 spots each, 20 total seek spots. There needs to be one open spot on the circuit though, or the entire circuit jams completely, almost instantly. Think of it like a sliding tile puzzle: The trains need an open spot. You need a missing tile in the sliding tile puzzle, or nothing moves at all. They don't actually LEAVE their station, until they have a destination. This is a good thing, because it means you won't have a train sitting in the station waiting, but not being counted as "in the station." There are two primary ways to think about limits and N-1: The first is, N-1 is set already. You have determined that (for example) you want eight total trains to do a given task. Those eight trains should be sufficient, you have decided, so you set the limits in order to force the trains to seek farther stations. A great and common example of this, is when you know a train will make the trip from the drop to the pickup, before you've produced enough to fill it: Setting that pickup's limit to one makes a lot of sense as you gain nothing from having a second train seek, while one is already loading. It will still have to wait when it gets there for the resources to be produced. The second, is to set the limits first rather than the total, and determine how many trains you'll need (N-1) based on how many trains you want for each station. Either way works, they're just different ways of coming to the same conclusion of N-1 total trains. The TL;DR version: 1) Station limits should be set based on how many trains you want on the rails headed towards the station, the farther the distance, or the greater the demand, the more trains necessary. For static limit systems, I recommend include at least +1 for the train already in a station. 2) All trains which can seek a station, should have a safe place to get out of the way of other trains, IE stackers. This means that if you have 6 parallel unload stations with 5 trains each, and one with an additional three, then every single one of those trains needs to be able to get OFF the "main rails" and sit somewhere not in the way of other trains. If you think you can maintain demand and/or supply without every having even a blip of a drop in it, go ahead and ignore this one, but I sincerely wish you the best of luck. You'll need it. 3) You must have N-1 trains in a static limit system, where N is the total sum of station limits for that circuit, in order to properly saturate the rails. Fewer causes starvation at one end or the other, which ends up being both ends because you are not getting full throughput on your production. More causes the system to instantly seize. 4) There are only three potential problems with this setup: Not enough trains, not enough production, or not enough rail throughput. If you have empty trains sitting at the drop station saying "Destination Full," check your farthest pickups. If there is a train in all the stations, you need more production. If there is not, you need to up your limits and add trains, all of the trains are still on the rails and have not yet arrived. If you have any trains sitting around, in station or somewhere before your "main rails," and they do NOT say "Destination full," you need better rail throughput, as the trains want to leave but don't have an open path. Also, if you're noticing that previous limits which were working stopped working for some reason and stations are being starved, you've added other circuits to the same rails, and you see rail congestion, you also should figure out why your rail throughput is low. There are many potential reasons why rails might not be carrying as much as they could be, improper signalling being one of the largest, and I'm fairly sure that a properly signalled, single rail would've been sufficient in the demo case, but a second rail more cleanly demonstrates "Throughput has been increased" than a hundred extra rail signals. Please feel free to ask me any questions you'd like about how to use this feature, I am still massively on the hype train even though I've been playing with this feature since it hit experimental, and really it's changed my quality of life a hundredfold. Trains do what they should now!! In Vanilla!!! It's just a touch confusing at first, but once you wrap your head around it and keep it simple, it's more elegant of a solution than I ever could have dreamed of. Simple "Full-Empty" train cycles properly draw from Pickups, and properly distribute to drops. OK I'll stop gushing now. EDIT: I'mma add and tweak bits here and there so suggestions to clarify things or make it more readable are VERY welcome! [link] [comments] | ||
When growing the factory takes ALL the roobits Posted: 18 Feb 2021 12:18 AM PST
| ||
imagine wasting resources on artillery shells to make a smiley f- Posted: 17 Feb 2021 11:14 AM PST
| ||
Posted: 17 Feb 2021 04:54 PM PST
| ||
Posted: 17 Feb 2021 09:44 PM PST
| ||
Posted: 17 Feb 2021 01:52 PM PST
| ||
Hi! I'm new in this game these are my first hours what do u think about my base? :) Posted: 18 Feb 2021 05:48 AM PST
| ||
Clicked on the wrong desktop icon this morning! Posted: 18 Feb 2021 07:30 AM PST | ||
I just realized something very wrong with the sulfuric acid barrel Posted: 17 Feb 2021 06:27 PM PST So we all know that factorio doesn't respect the law of physics or chemistry at all, perhaps with the exception of trains that kill you. But this realization just popped up in my mind when I was doing my (online) chemistry lab: Recall from high school chemistry class, that iron will undergo double replacement reaction with sulfuric acid and form iron (II) sulfate, a process more commonly known as erosion. Recall from factorio crafting recipes, that you can make one barrel from one steel plate, which is made by smelting five iron plate. This implies that the majority of the barrel is iron, which means that a barrel of sulfuric acid should start leak fairly quickly due to corrosion. But of course, we also use iron pipes to transport sulfuric acid and casually carry uranium in our pockets while walking around on heat pipes that are heated to 1000 degree Celcius. So perhaps this is nothing at all, after all. [link] [comments] | ||
Upcoming mod (Preview open to suggestion) Posted: 18 Feb 2021 12:25 AM PST This is direct copy from the mod page, the mod is not functional yet but thought I would ask for suggestions and gauge if it's something people would be interested in. Preview open to suggestionThis mod is well underway and you can see spoilers on my discord. Also planed pinch and pinch mine as a form of defense but is intended to sap your electricity as a result Status [link] [comments] | ||
64 hrs in. Haven't played any other game since. Am I doing this right? Any tips? Posted: 18 Feb 2021 07:15 AM PST | ||
Posted: 18 Feb 2021 06:46 AM PST
| ||
Posted: 18 Feb 2021 06:45 AM PST Playing around with long trains, I've found a way to create a frightfully simplistic train setup which has really great throughput. Use around 8 trains, each 42 wagon + 10 locomotives in length (52 total). Put them on a loop which goes around to each of the patches: coal, iron, copper, stone. Important: Build all the single resource intermediates at each base to increase throughput (belts, engines, gears, steel, bricks, walls) and pipe in / train in water to build some fluid stuff (sulfur, lubricant, petro). Just as an example, 1 cargo wagon of steel is requires 10 cargo wagons of raw ore if you want to smelt it at your factory. So, yeah, throughput relies on building as much as you can at outposts, but I prefer to only build stuff mined at that particular patch (except Oil, where I pipe in water. Will train it in at some point, cause fluid in Factorio is F'd. FFF!) Critical, and this is the dirty little secret to all this: No intersections! Intersections are the devils work and just slow everything down. Yes, I know, only morons can't handle chain signals, but seriously - intersections are anti-throughput. Put two train stops at each output station and your main base - call them "TM" and "TC". This is the route your trains will have, "TM - Wait 30 seconds" and "TC", no wait condition. That's it. Your trains rotate around the loop, stop at each TM, load up for 30 seconds, and then go to TC and then onto TM which will be the next station in the loop. When you add or remove patches, you don't have to change any of your trains. They will automatically path to the stop. Place the stops strategically, so when the train loads up for that particular stop, all of the relevant cargo wagons will stop by the patch and not further out - this obviates belting stuff around to load onto your train. (See example) Add train signals between the stations, make sure you leave room for 52 train length between signals. I try to make it so a train can queue up between stations, waiting for the next one to be done. I have about 8 patches going on each loop and 8 trains * 42 getting unloaded. They queue up on the loop. A long train stops in my factory every 45 seconds or so. This gives me enough throughput for around 1.5kspm. Use longer trains for more SPM or separate loops. That's it. No combinators, no complex train station limits, no wild stackers and train turnaround designs. Just dead simple massive train loop with long trains and constant, never interrupted throughput and supply. If you're OCD and need some kind of redundancy, than go ahead and build a parallel track with bypasses, but the long trains are pretty good at running over behemoths so probably not required. Use stack filter inserters and cargo train filters(use middle mouse button on inventory slot in cargo train) FTW when loading / unloading. I have the train square around my factory, but some people prefer to have it go through the middle. All up to you. Warnings: You may find yourself checking to make sure you didn't drop down infinity chests in your factory. You will find your patches getting drained very quickly due to increase throughput. This design can put you sleep as you watch it and nothing ever seems to go wrong. Where is the fun in that? example: https://forums.factorio.com/download/file.php?id=67687 [link] [comments] | ||
I wonder what will happen when i launch the rocket. Posted: 17 Feb 2021 01:06 PM PST
| ||
I know it´s weird, but i like it. Posted: 17 Feb 2021 11:35 AM PST
| ||
What do you most hope for in the upcoming expansion? Posted: 18 Feb 2021 06:13 AM PST Air enemies/units? Exploration mechanics? Ruins? Water entities? New resources? Different music? [link] [comments] | ||
Posted: 17 Feb 2021 09:46 AM PST Bugfixes
Use the automatic updater if you can (check experimental updates in other settings) or download full installation at http://www.factorio.com/download/experimental. [link] [comments] | ||
What do you think of my circuit-free fool proof kovarex set up? Posted: 17 Feb 2021 06:06 PM PST
| ||
Where is all the Station Limit hype??? Posted: 17 Feb 2021 02:16 PM PST I returned to Factorio after not playing since 2018, because of this feature. I've seen no hype for this feature, heck I haven't even seen someone mention it. This feature is amazing and the best possible solution for a problem which has plagued the best Factorio minds of several generations. The problem was, you have multiple places you're mining resources, iron or copper or whatever, and multiple places you need said resources. LTN is a mod which solves this problem. But the Station Limit also solves this problem, and IMO does it in the best way possible. Regulating the number of trains allowed to seek a station is the best possible solution because it allows you to adjust each and every station. Farther stations should have more trains seeking them, as it takes longer to get there. If it takes 15 seconds to fill the train, but it takes 60s to make the round trip(not including fill time), then you need 5 trains: One being filled, and four already on their way. If the round trip is 120s, then it's 9. I can't get over how awesome this solution is and how many mods and other massive projects that it just subverts with a "Haha nice try though." Where's the hype??? [link] [comments] | ||
Base Planner/Mod Question (Factorio Planner) Posted: 18 Feb 2021 03:55 AM PST Hi all . I got a Question and i hope someone maybe can Help me . Until now i only played Vanilla Factorio without gamechanging MoDs and uses the planner/calculator https://kirkmcdonald.github.io (awesome tool to plan 1000-10000 spm Bases ) So my Problem is with Factorio Planner i have Problems to plan my base ahead because it doesnt show the Info like Kirk (how many Ore smelter i need or how many mining Drill ect so from the Ground to the Top of all needed things . I tryed in "Factorio planner" to open all sup Categories but with that Modded Game there are so many subcategories and sometimes you misses one item and the overall Calculation is broken . The question is , is there something similar to Kircmcdonald ecept Helmod or Factorio Planner ? How are you planning a 1000 spm Base in PY wich is even more complicated than Industrial Revolution ? [link] [comments] |
You are subscribed to email updates from Factorio. To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States |
No comments:
Post a Comment