|
DaveMcW is the GOAT of Factorio 
|
#
?
Apr 12, 2023 21:33
|
|
|
|
| # ? May 23, 2024 05:37 |
|
|
Vizuyos posted:embrace the spaghetti  
|
|
#
?
Apr 12, 2023 22:27
|
|
|
celestial teapot posted:DaveMcW is the GOAT of Factorio Dude is both brilliant... and kind of insane. https://www.youtube.com/watch?v=XiMJ156RZWs
|
|
#
?
Apr 12, 2023 23:04
|
|
|
I finally had some silly reasons (build four reactors only, wood for burner power as I get out of coal) to build simple circuits and they're pretty great. Is there any super good guide to practical setups for these? Google has found a lot of very very specific examples that usually assume you already know what stuff does and how it fits together. Looking for more of a learn-by-doing thing. I'm now sorted for 2.6GW so after setting up another four iron ore outputs, two copper ore outputs, beaconed smelters to match, four more beaconed green farms, a few beaconed oil setups, 2-3 beaconed plastic, and central lubricant... I should be at ~26-32 reactors and finally ready to start making red circuits. On the plus side this will get me immediately into blue circuits and then I can start building things. It feels like I'm building pretty big right now and... *looks at past couple pages* Jesus loving christ
|
|
#
?
Apr 12, 2023 23:42
|
|
|
Some practical combinator projects I can think of off the top of my head: - A setup for monitoring fullness and ratios on a sushi belt. Useful for when you have a large number of relatively low throughput intermediates and don't want many lanes of mostly empty belts. - A setup for filling & summoning a build train determined by deficits of logistics system at destination outpost - A setup for controlling station train count limits and stackers to optimize train counts without requiring LTN or other mods - A setup using lights for a dashboard showing stockpiles of raw materials, intermediates, beakers etc giving an at a glance view of base health - A setup allowing you to throttle fuel cell input to reactors
|
|
#
?
Apr 13, 2023 00:00
|
|
|
The wiki has some great sections on circuits: a detailed overview, some tutorials, and a page of recipes for different purposes. https://wiki.factorio.com/ In the Expert section.
|
|
#
?
Apr 13, 2023 00:02
|
|
|
Mailer posted:*looks at past couple pages* Lol... Megabases are... mega? Here is the solar power and "mall" for my 2700 SPM base. If I remember right this is about 45GW of solar. Those 'short' pieces are 8.6GW each. 
|
|
#
?
Apr 13, 2023 00:04
|
|
|
I also like to use solar power. It's great for the environment and also great for not pissing off aliens! It also requires you to put panels on 80% of the planet's surface!
|
|
#
?
Apr 13, 2023 00:30
|
|
|
feed biters that tasty tasty gas that helps them grow big and strong = they get mad and break my stuff take up all their land with big ugly panels = eh, no biggie really and people pretend we're the bad guys
|
|
#
?
Apr 13, 2023 00:44
|
|
|
maybe they like the shade 
|
|
#
?
Apr 13, 2023 00:59
|
|
|
Mailer posted:I finally had some silly reasons (build four reactors only, wood for burner power as I get out of coal) to build simple circuits and they're pretty great. Is there any super good guide to practical setups for these? Google has found a lot of very very specific examples that usually assume you already know what stuff does and how it fits together. Looking for more of a learn-by-doing thing. One thing that's pretty widely applicable is "wire up the inserters supplying your train station's so that the chests fill evenly". A guide for how to do it, with spoilers if you'd prefer to work it out yourself:
That's about the most complicated thing I do with circuits generally. Aside from that, it's simple math problems like measuring my stockpiles of two goods that share a major ingredient, determining which one I currently have less of, and then directing the ingredient to the assemblers for that item.
|
|
#
?
Apr 13, 2023 01:59
|
|
|
The Locator posted:Dude is both brilliant... and kind of insane. Check out this writeup on DaveMcW's 1 RDPM base all the way back in 0.11. He does all kinds of amazing stuff that he doesn't publish to youtube, which is a big shame. Here's his winning submission for a competition to do chemical processing with best performance. Blows my mind. I asked him if he would stream on twitch or something but he said it would just stress him out. 
|
|
#
?
Apr 13, 2023 02:36
|
|
|
I like to have a global circuit network where each loader train station puts how many trainloads of product it has available on one wire, and each unloader station puts how many trainloads it's short on the other. I don't actually do anything with these signals, but it's nice to see at a glance where my bottlenecks are (e.g. "there's 28 trains of iron ore available but the smelters are still 10 short, I either need more trains or to clear a jam somewhere"). Throttling reactor fuel cell inserters is also something I do sometimes but usually I get kovarex so quickly I don't have to worry about it. My self-supplying train defense wall is probably the most involved circuit thing I've built from scratch, it's upthread somewhere if you want to look for it. Basically each station has a local logistics network, to cover the section of wall nearest the station, and a desired amount of supplies (turrets, ammo, pipes, etc. - everything needed to rebuild the defenses if most of it gets eaten by biters). It's great because I only need to stamp the blueprint and then send a spidertron with roboports, rails, signals, stations, combinators, and power poles (and one inserter and provider chest). Since the blueprint has the station name/conditions saved in it, as soon as the combinators get built the station is activated and calls for a train to bring everything else needed to build the wall segment, which includes bots, so the train comes, unloads some construction bots first, then starts unloading supplies and the wall builds itself. Meanwhile the spidertron has moved onto the next wall segment and this can continue indefinitely. One out of every 5-10 stations (depending on my current level of artillery range) is an artillery station too, and I expand outward in a spiral so there is just a continuous expansion of the perimeter.
|
|
#
?
Apr 13, 2023 02:59
|
|
|
Dear modules and beacons: I'm sorry I every said you weren't worth using. I didn't realize how useful it would be to cut the amount of copper my megabase uses in half without massively increasing the number of assemblers needed. Thanks, modules. Thodules. Throduction thodule three.
|
|
#
?
Apr 13, 2023 07:10
|
|
|
Solumin posted:Dear modules and beacons: I'm sorry I every said you weren't worth using. I didn't realize how useful it would be to cut the amount of copper my megabase uses in half without massively increasing the number of assemblers needed. Productivity modules can be an insanely good deal in the right machines. I think beginners don't use prod 1 nearly enough. Check out the payoffs: https://factoriocheatsheet.com/#productivity-module-payoffs If you have prods throughout your supply chain, the savings compound! That's not being tracked on this chart, so it's actually way better than it looks. Consider that 4x productivity mk3 in a rocket silo pays for itself on literally the first rocket launch. Speedrunners always use prod modules! That should tell you something. I always get Prod 1 in all my labs ASAP also. It's a no-brainer. ALT-F4 #25 has an awesome visualization of the power of Efficiency Mk1 also. If you play with biters enabled, you should consider putting them in your remote drills and other highly polluting machines.
|
|
#
?
Apr 13, 2023 16:09
|
|
|
You need prod 1s for purple science anyway, and speed 1s for rockets and assembler 3s, so I always overbuild both of those to have a steady supply for the endgame transition. (Specifically I make sure to use the full belt of red + green chips going to the purple science, with half of the excess going to more prod and the other half to speed.) These all go into every assembler making science packs, 1 of each into a mk2 assembler, for a roughly 20% increase in science output (on top of putting prod1s into all the labs). It's not a good payoff but I put them in every science pack assembler, even red and green, just to keep the ratios even. I'll get the modules back later anyway. Then when I upgrade to mk3 assemblers I just add 2 more prods to each, which ends up being only 2% less total output than not adding any more modules, but at 8% less material cost. By the time I'm upgrading to level 3 modules I'm tearing all this out and building somewhere else anyway, so that's as far as I upgrade my main stuff. Once I have stable red chip production I also have a couple efficiency assemblers churning them out for mining outposts. You'd think it would be good to switch to speed in miners at endgame but what ends up happening is once you have some decent mining prod research done, you end up running into belt throughput limits, so efficiency in miners all the way. Only exception is when I'm trying to get the last dregs of a patch mined up so I can reuse an area, then it's beaconed speed3 directly into chests.
|
|
#
?
Apr 13, 2023 17:07
|
|
|
Mailer posted:I finally had some silly reasons (build four reactors only, wood for burner power as I get out of coal) to build simple circuits and they're pretty great. Is there any super good guide to practical setups for these? Google has found a lot of very very specific examples that usually assume you already know what stuff does and how it fits together. Looking for more of a learn-by-doing thing. Things I've used circuits for, and can provide blueprints for if necessary:
|
|
#
?
Apr 13, 2023 19:16
|
|
|
Like a lot of starter players, I too didn't use beacons and modules but once I tried Space Exploration, with its beacon changes and 9 levels of modules, I became a big fan. And I also avoided using circuits/logic in vanilla, but you are almost required to use it for SE.
|
|
#
?
Apr 13, 2023 20:28
|
|
|
A Bakers Cousin posted:Like a lot of starter players, I too didn't use beacons and modules but once I tried Space Exploration, with its beacon changes and 9 levels of modules, I became a big fan. And I also avoided using circuits/logic in vanilla, but you are almost required to use it for SE. Yeah, same. Vanilla beacons bother me because I have a strong impulse to come up with the best solution and there's always some better way to arrange them, SE beacons are great because you design good setups around them but they're nice and organized and tileable.
|
|
#
?
Apr 13, 2023 20:43
|
|
|
Bremen posted:Yeah, same. Vanilla beacons bother me because I have a strong impulse to come up with the best solution and there's always some better way to arrange them, SE beacons are great because you design good setups around them but they're nice and organized and tileable. Vanilla can become a bit boring though, with almost every build becoming very similar with 4 prod modules and 8 speed beacons. It's hard to do something else once you're aware this option exists and learn some common belt routing tricks to work with the limited space. I'm actually looking forward to playing SE solely because of the beacon changes.
|
|
#
?
Apr 13, 2023 20:49
|
|
|
I was considering looking for a mod that gives SE-like beacons for this current base. But that was a bit too far from vanilla for what I wanted for the run.
|
|
#
?
Apr 13, 2023 21:06
|
|
|
IMlemon posted:Vanilla can become a bit boring though, with almost every build becoming very similar with 4 prod modules and 8 speed beacons. It's hard to do something else once you're aware this option exists and learn some common belt routing tricks to work with the limited space. I'm actually looking forward to playing SE solely because of the beacon changes. You end up with the similar optimized beacon free designs, its just you now need to tile it more and they're less space efficient. The system always converges on a few good solutions and that's the meta, unless you deliberately do something different.
|
|
#
?
Apr 13, 2023 22:47
|
|
|
This is why recommend doing the recipe ratio randomizer, a lot of stuff does still end up roughly the same but you have to adapt your designs and it's a new challenge every time. Especially fun when you randomize belt statistics so even your smelter ratios are different.
|
|
#
?
Apr 14, 2023 16:42
|
|
|
Xerol posted:Especially fun when you randomize belt statistics so even your smelter ratios are different. That is a bridge too far.
|
|
#
?
Apr 14, 2023 17:30
|
|
|
darthbob88 posted:Things I've used circuits for, and can provide blueprints for if necessary: I actually did these on my own and it was fun. In particular I added circuits to my standard oil (now beaconed oil) blueprint that turns on/off heavy cracking if I'm low on lube and makes rocket fuel if gas is abundant. That's a solution that'd be better solved by increased production instead of better handling of excess, but the experiment was fun. The train stuff is a bit macro and, like optimal blueprints, something I generally outsource to the internet. LTN has its own problems but it's a relatively elegant solution to just plugging in stations everywhere and not having to handle the back end logic.
|
|
#
?
Apr 14, 2023 18:12
|
|
|
This looks like a neat overhaul mod: https://mods.factorio.com/mod/FreightForwarding It seems to be a vanilla+ mod with emphasis on using a wide variety of transport vehicles to move goods around. For example, stack sizes are lower, but you can put things into cargo containers and stack those instead --- e.g. a cargo ship can carry 5,000 iron plates directly, or 100,000 if you put them into cargo containers first. Here's a reddit writeup about it: https://www.reddit.com/r/factorio/comments/12odxwv/freight_forwarding_review_tldr_its_good/
|
|
#
?
Apr 17, 2023 07:00
|
|
|
Solumin posted:This looks like a neat overhaul mod: https://mods.factorio.com/mod/FreightForwarding One of the more well-known factorio youtubers is currently doing a playthrough https://www.youtube.com/@KatherineOfSky
|
|
#
?
Apr 17, 2023 09:21
|
|
|
IMlemon posted:Vanilla can become a bit boring though, with almost every build becoming very similar with 4 prod modules and 8 speed beacons. It's hard to do something else once you're aware this option exists and learn some common belt routing tricks to work with the limited space. I'm actually looking forward to playing SE solely because of the beacon changes. Go down the rabbit hole of UPS optimization and you'll be introduced to a new world. 8 and 12 beacon is almost always wrong. Beacons themselves have a small UPS cost, but you have so many of them that you have to think constantly about how to get the machines laid out so that you can get the beacon affects you need without over-beaconing anything at the same time. And what counts as "too much" depends on what is consuming the thing you're making so it's always different. It's not for everyone but I am having a lot of fun, and when people say "Beacons make everything look the same"... I know what they mean and I empathize, but to me it's just not true.
|
|
#
?
Apr 18, 2023 18:37
|
|
|
I think the problem with beacons is they're way too visually busy. One beacon looks really cool. 100 beacons look terrible.
|
|
#
?
Apr 19, 2023 16:55
|
|
|
I've never quite liked how they looked - they look like weird sarlac pit things, I barely noticed the tower/antenna in the middle. I do like the modules are visible.
|
|
#
?
Apr 19, 2023 17:07
|
|
|
The old beacon graphics were better. https://mods.factorio.com/mod/classic-beacon
|
|
#
?
Apr 19, 2023 17:10
|
|
|
At least they went away from the full height radio tower design
|
|
#
?
Apr 19, 2023 19:25
|
|
|
https://www.youtube.com/watch?v=3Z10QL4PLZo (nothing really interesting in episodes 1 and 2. 3 is when he starts showing off the actual ChatGPT outputs)
|
|
#
?
Apr 22, 2023 19:52
|
|
|
We've been playing some modded hellset of Krastorio + Space Exploration. We're at well over 100 hours on this server and the fun modular spaceships are still super far away. I spent like... 5 hours yesterday just on holmium ingot production and I have more work to do there. I also can't believe how much vitamelange we use, one automated rocket full of nuggets isn't bringing enough back, I might have to set up a second and crank our rocket fuel production way up.
|
|
#
?
Apr 23, 2023 16:01
|
|
|
Yeah, I recently went back to my K2+SE play through that I had burned out on and my rebuilt iridium factory uses 4 out going cargo rockets and 4 incoming cargo rockets while my vulcanite plant now has hundreds of centrifuges so scaling things up turns it up to 11. edit: biggest complaint is that you can't make rocket fuel without making solid fuel first, that iron plate is more of a irritation than you would think. A Bakers Cousin fucked around with this message at 23:26 on Apr 23, 2023 |
|
#
?
Apr 23, 2023 23:11
|
|
|
A fun, if offbeat use for circuit logic comes from UPS optimization too. Problem: inserters unloading steel plates onto a belt will wake up as soon as there is 1 steel plate in the furnace. But that's wasted inserter activity, which is a hit to UPS (and wastes power, if you care about that). How would you design an inserter clock to ensure that the inserter doesn't wake up until there are 12 plates in the furnace, without sacrificing throughput? If you don't want to figure it all out yourself, there's a YouTube video describing one reliable approach, and a python script that does the math shown in this video and spits out a blueprint.
|
|
#
?
Apr 24, 2023 18:14
|
|
|
I think the easiest solution is to just not care if you're "wasting" single-digits of extra steel in the furnace's output slot - just enable the output inserter when (number of input plates - (number of output plates * 5)) > 60. You do get full throughput once it's all warmed up. e: I guess you need to use a different multiplier if you have productivity modules, and you might end up having more excess in the output slot. But it still runs at full throughput once there's enough of a buffer established, and there's no need to think about rates or timings at all - it's purely inputs vs. outputs. Jabor fucked around with this message at 18:33 on Apr 24, 2023 |
|
#
?
Apr 24, 2023 18:26
|
|
|
Is a time clocked approach really going to improve UPS? It is necessarily one memory op every tick which is like a mountain of processing compared to all the voodoo going on inside inserter performance. Last I screwed with circuits a clock was a quick way to drop 1 UPS on the spot but I think it's been optimized since then. The right solution seems to keep as much back pressure as possible so the inserter doesn't have the chance to swing at full speed taking full advantage of the overly optimized factory element performance. Otherwise has anybody modded angular rate limiting to inserters yet? E. Googling it and reminding myself UPS optimization is insane and an inserter sitting over a full belt is also bad for UPS. Also that people do this and sometimes get gains (sometimes not?) and it's probably something to do just for fun considering you just got done doing every other impactful optimization first. zedprime fucked around with this message at 20:03 on Apr 24, 2023 |
|
#
?
Apr 24, 2023 19:41
|
|
|
I thought after 1.1 inserter clocking was no better or worse than not doing it. It was a big craze for a while and then the patch changed something but people keep thinking it's an improvement.
|
|
#
?
Apr 25, 2023 04:17
|
|
|
|
| # ? May 23, 2024 05:37 |
|
|
Clocking inserters is definitely not always going to help in every context, but in bases big enough to give you UPS problems, some recipes should be clocked when unloaded onto belts. Steel and advanced circuits are common examples, because the recipes are often too slow to direct insert. Belts got tons of optimizations between 0.16-0.18. Inserter activity is still a huge drain on UPS in a big base in 1.1. But you can test stuff like this yourself, and I recommend that you do. There's so much folklore out there from gurus who have been playing the game forever and so they tell you about UPS concerns that have been optimized out long ago. For instance Nilaus did a video tutorial on UPS optimization that is full of misinformation that doesn't apply anymore in 1.1. Scientific doubt is good! You should not believe it just because I told you, or anyone else on the internet told you. `factorio --benchmark` is powerful
|
|
#
?
Apr 25, 2023 14:46
|
|