|
That didn't really clear it up. Perhaps you could quantify how far apart you think bot vs. belt throughput is right now, and how much that gap would be narrowed by more customizable inserters?
|
#
?
Dec 30, 2018 09:08
|
|
|
|
| # ? Apr 28, 2024 10:47 |
|
|
He's saying right angle inserters let you get more dense production. Which ok maybe but short of death world I'm not sure anybody is really considering density to the tile, and you can generally beacon a decent density even while still using belts.
|
|
#
?
Dec 30, 2018 15:21
|
|
|
 Third time attempting it, I did the seablock. I left it running, I increased game speed in the terrible early game, overused warehouses+loaders for huge buffers, but I did the seablock. Learned a bit more about circuit networks (S/R latches are great). Realized that circuit wires in blueprints placed by bots don't actually use or require wire items. Too late in I realized that ratio math and all that garbage I could write down with constant combinators.
|
|
#
?
Dec 30, 2018 17:38
|
|
|
Jabor posted:That didn't really clear it up. They did a whole Factorio Friday Facts about it! Part I and Part II. Edit, from Part II: Bots did a throughput of 16k plates/min, while Belts only did 9k. With bot speed upgrades, they get close to 20k. The conclusion kovarex makes at the bottom is that "Bots need a nerf." Of course, those FFFs were a long time ago (almost a year) and there was plenty of discussion about it on the forums at the time. 
XkyRauh fucked around with this message at 18:04 on Dec 30, 2018 |
|
#
?
Dec 30, 2018 18:01
|
|
|
More importantly, bots are less costly in terms of UPS than belts. Trains > bots > belts still though. 
|
|
#
?
Dec 30, 2018 18:30
|
|
|
corgski posted:Bots have way more throughput than belts in a given amount of space and a lot of that comes down to the limitations of vanilla inserters. Right angle inserters would allow increased density that would make belts competitive with bots (which really wouldn’t benefit from right angle inserters) Bots also use inserters, so any adjustments on that end isn't likely to close the gap. There's been a lot of words written about the gap between belts and bots and the blogs on the topic are good summaries, but if you want an even shorter summary about the problem of closing the gap between bots and belts, here it is: Belts have a maximum amount of throughput based on available physical space and the efficiency of your design - you can't realistically infinitely tile belt-based construction. Bots have a maximum amount of throughput based on the number of bots you construct. You can infinitely tile bot-based construction. In a game about automating production, one of these is self-reinforcing and the other is not.
|
|
#
?
Dec 30, 2018 18:36
|
|
|
Falcorum posted:More importantly, bots are less costly in terms of UPS than belts. This is why loaders should have been added. They're a tech upgrade that takes the hassle out of loading/unloading trains and feeding high-volume late-game consumers using belts, and they reduce the UPS cost by reducing the number of inserters that high throughput needs. Pigbuster posted:No reason it couldn't be post-bot tech. Unless you are worried about the verisimilitude of being unable to design an inserter that only moves half as much without giving it sentience. Why should it be a post-bot tech? Or even locked behind tech at all? Inserters take from a tile (fixed at time of placement) and drop in another machine/tile/lane (also fixed at time of placement). Right-angle or near-lane insertion doesn't change this paradigm as it's still taking from one fixed, adjacent position and dropping in another fixed, adjacent position as long as it's powered and the drop location has room. GotLag fucked around with this message at 20:03 on Dec 30, 2018 |
|
#
?
Dec 30, 2018 19:51
|
|
|
I just completed my first oil refinery setup. Each layer of complexity is deeper than the last.
|
|
#
?
Dec 30, 2018 20:56
|
|
|
I think the most elegant solution is to limit the number of active pickup/dropoff requests a chest can have. Currently, with enough bots distance increases latency, but has no effect on throughput. This change would make distance have a increasingly negative effect on throughput, giving bots a role as short range/low throughput logistics. This would make pure bot factories less optimal, while keeping some important niches alive and would be easy to mod away if you really really want full bots back. Bot train stations would be pretty much unaffected, as the provider and storage/requester chests are close by so the bots would complete jobs faster than requests build up. Bot mining would also be unaffected, as each group of miners are pretty low throughput. Dense, heavily beaconed bot factories would be hit hard, as you would either need to belt materials close to keep up throughput, or make requester chest 'antennas' beside each assembler ruining your density and optimal beacon placement. Construction bots could be exempt from the limits, as they help you build the factory not replace it.
|
|
#
?
Dec 30, 2018 20:58
|
|
|
Configurable inserters are pointlessly fiddly and don't actually let you do anything you can't already do. There's no factory design that you can make with right-angle inserters that you couldn't make (just a little bit bigger) with the existing restrictions. Long inserters are useful because they solve a problem you can't solve otherwise. And then later on, when you have plentiful underground belts to give you other ways to solve the same problem, their usefulness drops off (because they don't have fast or stack variants). This is a good thing! I don't use the inserter configuration stuff when I play groggy stuff like angelbobs either. Yes, I know it's technically better throughput to manually configure all your inserters to only make 16-degree movements, but that's fiddly bullshit that doesn't add anything meaningful to the game. You just have to micromanage every single one of your inserters if you want to be ~optimal~ and it's incredibly stupid.
|
|
#
?
Dec 30, 2018 23:27
|
|
|
Jabor posted:Configurable inserters are pointlessly fiddly and don't actually let you do anything you can't already do. There's no factory design that you can make with right-angle inserters that you couldn't make (just a little bit bigger) with the existing restrictions. This is exactly why side inserters are perfect for a mod but don't need to be added to the main game.
|
|
#
?
Dec 30, 2018 23:32
|
|
|
I fell off this game around the time I just got to/past oil, and found myself picking it up again with a fresh save to more or less relearn the basics. I think what i usually struggle with is getting to the oil; my last game I regenerated until I had fairly close oil, this time I didn't but it's probably still not that far.  What would be a good way to get this oil? I have a tiny patch to the west, running power to that should be fairly simple, but there's no local resources, should I pipe it back? Or make poo poo over there and belt or train it? Last time I tried to train stuff it felt real messy.
|
|
#
?
Dec 30, 2018 23:58
|
|
|
Pumpjacks to a tank, pump into undeground pipes, then a lot of underground pipes + a pump in the middle. Until they redo fluid mechanics each pipe segment counts as one volume and underground parts don't count at all, so for each underground pipe pair you cover 11 tiles for the cost of 2. For most normal fluids its not until you get past 200 pipes that you really have to make considerations for flow rate. So 200 underground pipes can cover 2200 tiles before you really need a pump. Even then 1000 pipe segments (11,000 tiles) will still flow 200/sec with just a pump at the start so depending on how much oil the patch is putting out you may not need a pump at all. But at that point is is materially far cheaper to use trains and fluid wagons.
|
|
#
?
Dec 31, 2018 00:22
|
|
|
Well, Its 2.75 iron plate per rail segment which covers two tiles, and 15 iron plate per underground pipe which covers 11 tiles, so 1.375 iron/tile for rail vs 1.36 iron/tile for pipes but a train doesn't have to worry about flow rates.
|
|
#
?
Dec 31, 2018 00:27
|
|
|
Oxyclean posted:I fell off this game around the time I just got to/past oil, and found myself picking it up again with a fresh save to more or less relearn the basics. Oil can be piped over. It's not ~optimal~ but it works more than good enough until you're building a mega factory. The next simplest (and also the optimal solution) is to move it in liquid train cars to wherever you want your refinery to be. I suggest refining at your main base.
|
|
#
?
Dec 31, 2018 00:58
|
|
|
Cool - Im trying to get trains to work, I've build a single line (kind of like a u shape) but I'm having trouble getting it to recognize the stations. "Go to station" seems to be working in one direction but not the other, I tried putting a locomotive on both ends too.
|
|
#
?
Dec 31, 2018 01:13
|
|
|
Oxyclean posted:Cool - Im trying to get trains to work, I've build a single line (kind of like a u shape) but I'm having trouble getting it to recognize the stations. "Go to station" seems to be working in one direction but not the other, I tried putting a locomotive on both ends too. Train engines look at the right side of the rail for signals and stops, and can only look forward--not backward.
|
|
#
?
Dec 31, 2018 01:20
|
|
|
You can have an unlimited size single block with one train on it gojng either direction. If you want more than one train on a connected track you need signals to break it up into blocks. Think of Signals like one way gates, a train will only pass through them if they’re green I the direction they’re traveling. If you put Signals on both side of the same track segment you’ll get a gate that’s passable in both directions. Two way tracks are a pain IMHO because while they require less rails and space Signals aren’t as cheep as rails and you have complications of trying to allow controlled passing on two way track.
|
|
#
?
Dec 31, 2018 01:26
|
|
|
Oxyclean posted:Cool - Im trying to get trains to work, I've build a single line (kind of like a u shape) but I'm having trouble getting it to recognize the stations. "Go to station" seems to be working in one direction but not the other, I tried putting a locomotive on both ends too. if it's a single length of track, the easiest way is to have your train use a locomotive on both ends and your stations set on the right-hand side in the direction of travel. However, this is pretty inefficient because you have to fill twice as many locomotives and they add a bunch of dead weight to the train and slow how quickly it can brake. A slightly more advanced method is to use a teardrop switchback to allow a single-locomotive train to loop around and go back and forth. However you have to be careful about positioning, and might need to use signals to keep the train moving the correct direction. It's a good stepping stone to trying new strategies with signals and such. I recommend trying the tutorials, they're much improved from when they were introduced.
|
|
#
?
Dec 31, 2018 01:33
|
|
|
Train station at the small one and at the large one. Train stops at small to pick up from the tank (hold for 5 sec then go) then heads to the large (wait until full then go). Have a rail blueprint that includes a To and Fro track and a large tower, and make your blueprint tile to the tower's wire range.
|
|
#
?
Dec 31, 2018 04:03
|
|
|
After getting distracted for a couple of weeks I finally managed to wall off the main base. It's a good feeling, one of security. It was a pretty narrow thing there for a bit, the starter saphirite patch was down to about 15k ore by the time I was able to clear to the closest patch, which was luckily right next to some coal. I don't usually play with such huge amounts of biters since the fun of the game it designing stuff, but deathworld is a pretty interesting experience. It really changes the priorities and usefulness of various game items. Like, usually I just try to jump straight from modular armor with nightvision into power armor with a bunch of personal roboports. Using personal solar cells to power a single roboport is pretty pointless in my usual games since I can just carry around roboport buildings and 7 stacks of construction bots. But in this game the bots themselves are such a huge investment I found the solar powered personal port indispensable for quickly setting up strongpoints as cleared, and building my defense walls. Another thing that it really hammered home was how much worse laser turrets are compared to flame turrets and gun turrets with red ammo. I was forced to use them for a while since I was so short on iron I could barely keep my tank stocked up with red ammo, let alone a huge number of turrets, and man are lasers feeble compared to bullets. The usefulness of poison capsules was also a big surprise. When I'd tried them out before it always seemed pretty pointless, in terms of speed and killing power, but as it turns out they're one of the best ways I've found to clear biter nests since you can kill a dozen worms with one toss. From here on out I don't really expect any more serious worrying about being overrun. The main issue I see now is that I don't have any jivolite inside my perimeter. I guess my new goal is gonna be to push to artillery since uranium seems kind of hard to come by in angelbob games.
|
|
#
?
Dec 31, 2018 12:02
|
|
|
Galvanik posted:From here on out I don't really expect any more serious worrying about being overrun. The main issue I see now is that I don't have any jivolite inside my perimeter. I guess my new goal is gonna be to push to artillery since uranium seems kind of hard to come by in angelbob games. Alternatively there is always the option to mineral sludge in whichever raw ore you are short/lacking. Reverend Dr fucked around with this message at 18:06 on Dec 31, 2018 |
|
#
?
Dec 31, 2018 17:31
|
|
|
Ended up using a switchback, think last time I did this I had an engine on both ends so this is much nicer for having less clutter. I'm now at purple science, and to start working on prep for blue circuits and yellow science. Scaling things up to keep everything supplied is becoming the tricky part. Are modules worth it at this stage?
|
|
#
?
Dec 31, 2018 20:47
|
|
|
Oxyclean posted:Ended up using a switchback, think last time I did this I had an engine on both ends so this is much nicer for having less clutter. Production modules in your science labs are always worth it, it's free science. Efficiency everywhere if you need to reduce pollution. Speed is nice for kovarex setups but other than that not super useful until you start beaconing.
|
|
#
?
Dec 31, 2018 21:01
|
|
|
Early game you can mostly ignore modules, really. Like was just said some production in your labs is good. It slows them down but you can just puke out more labs. It's also handy to stuff a few productions in really expensive, high end stuff like the more complex bottles. If you plan on launching more than one rocket your rocket silo should absolutely have production 3's in it. Note that using modules other than efficiency ones massively increases your energy needs. That won't be an issue if you set up a big, fat nuclear plant or vast fields of solar panels but it can be a pain on just coal. One of the biggest questions is how long you figure on playing; if you're planning on a long haul then puke a bunch of production 3's all over just like, everything. Megafactories benefit greatly from production modules just kind of being everywhere but some things take literal days of play before you get any return on the modules.
|
|
#
?
Dec 31, 2018 21:09
|
|
|
Oxyclean posted:Ended up using a switchback, think last time I did this I had an engine on both ends so this is much nicer for having less clutter. It only takes about half an hour for tier 1 productivity modules to pay for themselves in electronic circuit assemblers and other moderately expensive intermediates. So yeah, tier 1 modules are usually worth it as soon as you unlock them and can afford to make energy for them. Science assemblers (and labs I think) also benefit a lot from modules. Tier 3 modules take tens of hours to pay for themselves in most assemblers. Using modules on smelters takes a lot longer to pay off, and I usually don't switch to electric smelters until I set up my late game smeltery anyways. On the other hand, I've launched rockets without making any modules before so you don't need them if you don't want to deal with them.
|
|
#
?
Dec 31, 2018 21:09
|
|
|
Once you get basic logistics services up you can just throw old modules into logistics to be delivered for upgrade, which means once you have the tech and infrastructure to slam Productivity into science and then high-level parts you should and then upgrade them when you can.
|
|
#
?
Dec 31, 2018 21:10
|
|
|
The Factorio Cheat Sheet has a great breakdown of where you should put productivity modules: https://factoriocheatsheet.com/#productivity-module-payoffs
|
|
#
?
Dec 31, 2018 21:12
|
|
|
ToxicSlurpee posted:If you plan on launching more than one rocket your rocket silo should absolutely have production 3's in it. Prod3's pay for themselves with just one rocket 
|
|
#
?
Dec 31, 2018 21:14
|
|
|
Evilreaver posted:Prod3's pay for themselves with just one rocket They do but if your plan is to launch a rocket, declare victory, then quit then you don't really have to worry much about that.
|
|
#
?
Dec 31, 2018 21:59
|
|
|
ToxicSlurpee posted:They do but if your plan is to launch a rocket, declare victory, then quit then you don't really have to worry much about that. Wait I forget, do you need prod 3 research for rocket silo research. That would be the only reason, if its optional path gotta amortize that research in but meh, feels soo good to watch an extra rocket part appear.
|
|
#
?
Dec 31, 2018 22:19
|
|
|
Every speedrun has prod 3 modules in the silo.
|
|
#
?
Dec 31, 2018 22:43
|
|
|
zedprime posted:The idea of them paying off in <one rocket is it takes net less resources to launch even that first rocket by slamming prods in. Yeah, you do need prod 3 research to build the silo in the first place and they do return their cost on way less than one rocket. So yeah, go for prod 3 and stuff that silo full as it goes!
|
|
#
?
Dec 31, 2018 23:35
|
|
|
If you have the production to spare (and if you don't, set up more production) you should be making and boxing a few stacks of each T1 module as soon as you unlock them. Stick T1 prods in everything oil-product-related ASAP (and T3s into plastics, sulfur/acid, and lube as soon as you can too). If it's running too slow just add more plants. Three prod1 + one speed1 in assembler 3s for yellow/purple packs. Also do 3+1 in your blue chips. Best prod available in labs at all times. Speed in pumpjacks. 3x Eff1 in remote miners, cuts down on power but more importantly pollution significantly - slows down attacks and lets you get away with lighter (but not zero) defenses in your outposts. Having a couple stacks of eff1s to throw into your T2 assemblers and various other miners during a power crisis can save a base. Four prod3s in a silo reduces the number of rocket parts needed from 100 to 72. (Actually 71 and 4/7ths, but you won't get that fraction until you build the next part.) This saves 280 of each of the rocket components. Silo speed is never an issue until you get into kspm bases, and even if you're launching just one rocket, you're probably not going to have your main production scaled up to the point of being able to make those components quickly enough to have the silo continuously running. So it is basically a 28% time save on launching. Also if doing nuclear but not running kovarex (which you only really need if you're mass producing nuclear fuel for trains or rocket nukes) sticking four prod3s in your fuel cell assembler makes the U-235 go so much farther.
|
|
#
?
Jan 1, 2019 06:19
|
|
|
I have finally found an oil patch (+4000%) on my first (non-poo poo) attempt of a game. Unfortunately, it is decently far from my base, so I need to make an oil train (first train). Are there any good guides/blueprints to help me do this right? I am really looking forward to the train part.. but really daunted by it. I am also a scrub with oil, as I have only set it up one time. (I am using Nilaus' Vanilla Done Right YouTube videos.)
|
|
#
?
Jan 2, 2019 00:43
|
|
|
dexefiend posted:I have finally found an oil patch (+4000%) on my first (non-poo poo) attempt of a game. What exactly are you looking for blueprints for? I like train intersection blueprints, and station blueprints are useful too. But for your first train stuff, you definitely don't need blueprints. Start with putting down some tracks. If you place a train station, it'll show you outlines of trains on the tracks to give you an idea of the placement. Similarly, you can place pumps to fill/empty your tanker car at the train station. That's all you need to move oil around. Don't listen to the people who will tell you to use barrels, those aren't cool.
|
|
#
?
Jan 2, 2019 00:57
|
|
|
Here's my basic oil train setup; I'm still learning, so it's probably not the best: As people suggested up-thread, you'll want a switchback on the track - the train can't drive backwards on self-drive. You could put an engine on both ends but that's a bit more fussy. You need a pump to put the oil in, and a pump to take it out. You don't need to have double tanks on the pump end; I was just fussing around with the setup, but I have the impression you'll at least want one to act as a buffer. Same goes on the receiving end, you'll probably want a tank or two to stockpile your oil for the refineries. You'll need to set a schedule;  You'll also need to fuel it with something - you can start with some belted coal, this is my secondary base and I figured I'd convert some of my heavy oil into solid fuel. Here's a basic refinery setup:  This is from my secondary base so it doesn't have much attached to it yet, but the basic idea is to space your pipes so you can attach more refinerys as/if you need, and then give yourself space for storage as well. Your refinery needs to output everything, or production stops entirely. Also, leave space to bring in water later, you can get advanced refining which is oil + water for a higher petroleum yield. e: Unrelated want to share my map: (playing on peaceful like a huge coward)  Unlocked logic requester chests, trying to figure out how make good use of those so I started building a second base. Oxyclean fucked around with this message at 01:30 on Jan 2, 2019 |
|
#
?
Jan 2, 2019 01:23
|
|
|
Also, like everything, resist the instinct to just make one refinery and wait. Make as many as you have resources to build, run them all in parallel, and link up all their outputs to common storage.
|
|
#
?
Jan 2, 2019 04:17
|
|
|
Only took like four hours, but I got my oil train working! I made some plastics! Nilaus' Blueprint set a setup for oil refineries that I used. 7 refineries, 11 chemical plants, 8 lines and sets of tanks for various fluids. dexefiend fucked around with this message at 05:00 on Jan 2, 2019 |
|
#
?
Jan 2, 2019 04:57
|
|
|
|
| # ? Apr 28, 2024 10:47 |
|
|
Solving oil is one of the most interesting problems in the game. Avoid blueprints and guides entirely, IMO. Don't be stubborn and insist on making everything perfect the first time you play - you can't unlearn it and you'll miss out on the discovery, which is the entire point.
|
|
#
?
Jan 2, 2019 06:12
|
|
