|
Kenlon posted:Why on earth were you using the recipe that takes silver plate instead of using the one that uses Tin and Lead, which are both guaranteed to be in your starting location? Looking at it now? I have no idea. When I did it I was stuck on something, saw the silver plate recipe and went and got some silver.
|
# ? May 12, 2019 08:24 |
|
|
# ? May 25, 2024 13:22 |
|
My google-fu isn't the greatest and I can't find an answer. How do I turn a train station on if ANYTHING is in the chest wired to it? I want to put some "trash" stations at various places to dump stuff to send back to the main base's storage area, but I can't figure out what circuit logic to use. Would be great if there was an option for "contents >0" or maybe there is and I'm missing it? Also, wish list for trains getting a wait condition regarding fuel levels. EDIT: I'm dumb and had the alligator pointing the wrong way. I was looking for *>0 and had it as *<0. Leaving this just in case it helps anyone else *sigh* BlueOccamy fucked around with this message at 04:27 on May 13, 2019 |
# ? May 13, 2019 04:24 |
|
It may sound silly but in that type of situation I usually would attach the source of the signal [chests] to a combinator with anything>0 and then have it output the enable condition to the station. At this point I think the only thing I directly use anything>0 are on decorative colored lights and hastily made mid-game things that I fully intend to tear down. Its a little bit of idiot proofing (I know who's going to end up using it) and prevents the circuit condition from being completely open if a stray signal ends up being connected. Also it gives a good place to engineer in anything additional later, such as a timer which was incredibly important for the trash train I most recently built.
|
# ? May 13, 2019 15:12 |
|
Went back to my primary large factory and noticed that my FPS/UPS is dipping into the 47~48 range now. Kind of annoying since when I open up the monitor to check on system usage I'm running 20% or less CPU utilization, but it's completely pegging out a single core (i7-8700k @ 4.7Ghz) only and the video card is basically at idle (1080ti). Seems fairly obvious that this game needs some good multi-threading and it would scale vastly larger than my factory without any issues. I can't imagine how some of those huge mega-factories at 10k SPM must play. I tried using console commands to switch to peaceful mode and then -kill all the biters, and only picked up like 1-2 UPS after that. Guess if I really want to increase it I'll have to use that mod that reclaims all the explored chunks I'm not using. I could also delete all my perimeter defensive stuff if I'm killing off all the biters, but then how would I explore without hundreds of artillery batteries?!
|
# ? May 13, 2019 15:30 |
|
Games like this (where everything interacts) are not exactly easy to thread. They already do some, I think, for flow simulation and maybe some other spots. I think turning off pollution entirely can help a good deal with FPS, too.
|
# ? May 13, 2019 15:37 |
|
necrotic posted:Games like this (where everything interacts) are not exactly easy to thread. They already do some, I think, for flow simulation and maybe some other spots. I honestly have no idea how threading works, just know it would be cool to have the load spread out amongst my cores! Working 1 of 12 is kind of a bummer. Thanks for the pollution tip, might try that. At this point evolution is pretty much maxed out at something like -.998x, and the biters aren't any sort of threat, just a constant use of energy/ammo/crude oil as they throw themselves into their own funeral pyres at the walls, so turning biters and pollution off would allow me to get rid of my defensive walls as well to help (no idea how much it would help, but at this point the perimeter wall is huge, with lots of turrets and inserters). There are no more achievements to get that aren't already failed (the timed ones) except for the solar-only one, which I'm not trying to get since I've got 7gw of nuclear online, so losing achievements isn't an issue either.
|
# ? May 13, 2019 15:50 |
|
Redid a 2x2 nuclear design from scratch. The piping is ugly, but the circuit regulation is more sophisticated than the version I did a couple years ago. The circuits detect when the steam level in the tanks is low enough to absorb a full fuel cycle, then trips the first latch on top, which sends a pulse to the fuel inserters. On the output side of the heat exchangers I have 1 way valves from the Flow Control mod feeding pumps. The one way valves can be circuit linked to read their fluid content. So, the second latch reads the sum total of the volume of all the one way valves - when it is zero the cycle is inactive / complete, when it's a small value above zero, it is active / in progress. When the cycle switches from active to inactive, it resets the first latch. If steam levels are high enough, the reactor stays on standby, if not, it immediately starts another cycle. Bottom right is the status indicator. The 10 lights bin deciles where 0% < steam % full < 100%. The top and bottom ones are for <=0 and >=100% which should never happen. Color indicates reactor status: Green is active cycle, yellow is standby, red is failure to generate requested steam (maybe the fuel cells ran out), purple is reactor is active but insufficient steam capacity to absorb the cycle (ie wasting fuel.) There's a couple sirens to signal conditions red and purple, which have yet to occur other than during testing. Normal operating behavior is the reactor is on standby (yellow) > steam levels reach 52% > reactor initiates fuel cycle, cells turn green > steam levels rise to about 95% depending on load > reactor returns to standby. So in the second picture, it's about a minute or so away from starting a new cycle. I also built 3x as many turbines as the sustained output so that the reactor can spike power in case of a mass biter attack or something like that. Steam tank capacity is just over 2 cycles so it can run at 1.44 GW for about 2 minutes before bottoming out steam. Max sustained is 480 MW. I know there are a lot of these available, and surely better ones, but I really love the feeling of ownership you get over your designs in this game. I made a hilariously overdesigned kovarex processor with the design goal of keeping zero excess U235 in the centrifuges. Of course, it just has a lot of excess on the belts... metasynthetic fucked around with this message at 18:41 on May 13, 2019 |
# ? May 13, 2019 18:34 |
|
The Locator posted:I honestly have no idea how threading works, just know it would be cool to have the load spread out amongst my cores! Working 1 of 12 is kind of a bummer. The devs have talked about it at length in a few FFFs and the occasional forum post. There are reasons they don't, but my brain usually turns to mush after the first couple of sentences, but I think the upshoot is that it would be a crazy amount of rewriting/work and it ain't happening any time soon, no before 1.0 at least. These posts usually rapidly devolve into internet experts telling the devs why they're wrong.
|
# ? May 13, 2019 18:56 |
|
Gravy Jones posted:These posts usually rapidly devolve into internet experts telling the devs why they're wrong. Lol... this is not shocking to me at all. The internet never fails to deliver this sort of thing.
|
# ? May 13, 2019 19:19 |
|
The Locator posted:Lol... this is not shocking to me at all. The internet never fails to deliver this sort of thing. Gamers are some of the worst about it. Similar thing comes up with Dwarf Fortress pretty often, which would be infinitely harder to parallelize than Factorio.
|
# ? May 13, 2019 19:27 |
Gravy Jones posted:The devs have talked about it at length in a few FFFs and the occasional forum post. There are reasons they don't, but my brain usually turns to mush after the first couple of sentences, but I think the upshoot is that it would be a crazy amount of rewriting/work and it ain't happening any time soon, no before 1.0 at least. These posts usually rapidly devolve into internet experts telling the devs why they're wrong. Multithreading isn't a magic bullet. There are limits to how much a task can be broken down into parallel processes, mostly based on many calculations having to use the same data and all affecting the ultimate shape of that data (different threads are not good at sharing data without a lot of extra work, which creates overhead that at some point can outweigh any gains you make from multithreading) and some tasks are just inherently linear because subtask B requires data that is only available when subtask A is complete, for example. Slashrat fucked around with this message at 19:45 on May 13, 2019 |
|
# ? May 13, 2019 19:28 |
|
Galvanik posted:My base beautification project continues. The actual building layouts aren't anything special, but the tile art is getting more complicated, and in my opinon, pretty. Wow. How did you do that? You should export some blueprint strings so I can copy it. Slashrat posted:Multithreading isn't a magic bullet. There are limits to how much a task can be broken down into parallel processes, mostly based on many calculations having to use the same data and all affecting the ultimate shape of that data (different threads are not good at sharing data without a lot of extra work, which creates overhead that at some point can outweigh any gains you make from multithreading) and some tasks are just inherently linear because subtask B requires data that is only available when subtask A is complete, for example. 1 thread per logistic bot. It's not like they interact with anything else LLSix fucked around with this message at 19:54 on May 13, 2019 |
# ? May 13, 2019 19:51 |
|
LLSix posted:1 thread per logistic bot. It's not like they interact with anything else But the bots do interact with me and my locational data. Lemmie tell you, no predictive algorithm could account for my pathing without real time updates every frame.
|
# ? May 13, 2019 21:33 |
|
KirbyKhan posted:But the bots do interact with me and my locational data. Lemmie tell you, no predictive algorithm could account for my pathing without real time updates every frame. How often I get hit by my own trains sees to this.
|
# ? May 13, 2019 21:38 |
|
chairface posted:How often I get hit by my own trains sees to this. A mod idea: provides Futurama style werecars: every thirty nights, your vehicles turn against you and try to run you over. If they succeed, your model is randomly turned into a car.
|
# ? May 13, 2019 21:49 |
|
So what can I do with over 2400 blue assemblers? Deconstruct my blue assembly assembly station? got it. Pretty sure I never need to build any more blue assemblers ever again. Ass_Burgerer fucked around with this message at 03:38 on May 14, 2019 |
# ? May 14, 2019 03:35 |
|
Turn them into yellow assemblers!
|
# ? May 14, 2019 03:46 |
|
That's only 1200 yellow assemblers!
|
# ? May 14, 2019 03:47 |
|
Ass_Burgerer posted:
Feed ‘em into your green assembler line. Don’t scrap the blue line, that way lies regret.
|
# ? May 14, 2019 04:13 |
|
Half-wit posted:Turn them into yellow assemblers! FISHMANPET posted:That's only 1200 yellow assemblers! Yeah thats a good start to a megabase
|
# ? May 14, 2019 04:34 |
|
Now you just need to make enough smelters to feed them all
|
# ? May 14, 2019 04:35 |
|
Ass_Burgerer posted:
I deal with this when upgrading from lower tier items. Place a requester chest for blue assemblers with an inserter that puts them directly into your yellow production and let your robot network take care of the logistics. You then wire a circuit condition so that no new blue assemblers are created until the chest is empty. This leaves existing production pipelines idle and ready to kick in at a moments notice. I guess it depends on how long you intend to play but you'll need more eventually. I'm at about 2k active yellow assemblers with another 1k in the logistics network ready to be deployed as soon as a blueprint is placed.
|
# ? May 14, 2019 05:06 |
|
LLSix posted:Wow. https://pastebin.com/94WvXD2Y It's a work in progress so the sizes and shapes of things aren't really formalized. As for how I do it I usually go into a biter free map with cheats (infinity chests and infinity power) on and just fuss with patterns until I find something that catches my eye, then work on expand that into a larger tileset. Or I'll mash together two existing things and see if I can link them in a pleasing way. The rectangular plaza on the right and really long one in the middle are examples of doing that. The large circular one was a proof of concept I did to see if I could use a spreadsheet to map out various angles and intersection lines of arbitrary shapes. Like, if I have a 60 diameter circle and want to make that into a 5 point star pattern, where exactly to the lines go in a cartesian plane. That sort of thing. This is the template the final stamp evolved out of. ps: That Khorne symbol stamp is a memorial marker for biter battles that are particularly bloody Galvanik fucked around with this message at 10:02 on May 14, 2019 |
# ? May 14, 2019 09:58 |
|
I have been meaning to post pictures of my megabase but posting equates to less playing so it's a low priority. Anyway, I finally got my off-site advanced circuit production online after spending several weeks of building the supporting infrastructure. I then realized that I had massively over built my smelting operations which allowed me to quickly build more off-site facilities. The biggest strain on my production is easily low density structures so that became my main weekend project. It's working pretty well so far but low density structures are absolutely terrible across the board. In production, I made 32(!) lanes of copper plate inputs and the output doesn't come close to saturating 8 belts. Next I realized that the stack size is 10 which makes it awful for transport. My standard 4 car train can carry 16k plates but only 1600 low density structures. Anyway, I now have 8 lanes of low density structures on my main bus which should be able to sustain 2 rocket silos, satellites, and science. The problem now is that I have 8 lanes of copper plates on the main bus that is hardly used because the primary consumers are off site. Even worse is that with plastic being made off site and solid fuel made from coal, oil production is totally out of wack with little petroleum consumption. The main bottleneck has become flying robot frames due to the lack of lubricant for electric engines. My next project was going to be greatly expanding coal liquifaction for solid fuel but instead I'm forced to make solid fuel the old fashioned way just so I have enough byproducts for lube. RVWinkle fucked around with this message at 18:27 on May 14, 2019 |
# ? May 14, 2019 18:24 |
|
I launched a rocket on my previous save, then decided to start over with the Space Exploration mod because the idea sounds really cool. Got really annoyed with how it forcibly modifies vanilla recipes without any option to leave them be. So, I wound up learning enough Factorio modding to make my own to put them back. I may wind up doing the same for the robot attrition stuff if the minimum rate is too much of a hassle and if I can figure out how. If anyone would like a copy I'll post a link. Also, Construction Drones is the cutest mod and I don't even care about anything else I love them with my heart: https://gfycat.com/farflunggranularbarnswallow metasynthetic fucked around with this message at 02:16 on May 15, 2019 |
# ? May 15, 2019 02:04 |
|
This is amazing art. I didn't know you could make stuff like that in Factorio.
|
# ? May 17, 2019 03:43 |
|
metasynthetic posted:I launched a rocket on my previous save, then decided to start over with the Space Exploration mod because the idea sounds really cool. Reminds me of Total Annihilation builders. Have to try them out.
|
# ? May 17, 2019 21:37 |
|
After four aborted attempts, I finally played long enough to launch a rocket. Behold my atrocity. I love how you can kinda see all the times I did not give a gently caress, like how I just hand-carted all my yellow science materials because I was sick of building belts. https://imgur.com/a/7OWUChO Can't wait to start again, already have a ton of ideas for things I want to do differently.
|
# ? May 19, 2019 11:15 |
Finally looking to move into 0.17, what are the must-have QoL mods that people recommend? I'm not looking to add the whole new tech and production chain mods, but, like, ones that make it so resources don't spawn intermingled, etc.
|
|
# ? May 19, 2019 22:52 |
|
Arrath posted:Finally looking to move into 0.17, what are the must-have QoL mods that people recommend? 0.17 completely re-works mapgen so you probably won't need any resource overhaul mods. I play with a few small tweaks that I really like:
|
# ? May 20, 2019 00:03 |
|
Arrath posted:Finally looking to move into 0.17, what are the must-have QoL mods that people recommend?
|
# ? May 20, 2019 00:48 |
Cool thanks guys.
|
|
# ? May 20, 2019 01:56 |
|
I just started playing a new game with "Tinystart" and "Auto deconstruct" and they are both awesome. Even Distribution. Closest First Inbuilt Lighting Afraid of the Dark Squeak Through All are nice ones to take a look at. "FNEI" or "What is it really used for" seem to both be good for figuring out recipes if you are playing with alternative recipes of any kind (Bobs, Angels, Krastorio, etc.).
|
# ? May 20, 2019 02:20 |
|
So I clearly misunderstood how pollution affected biters. I set my new map to very slow expansion and the maximum starting area size. My pollution cloud is nowhere near any biters, but I discovered some really big nasty biters on the map with radar, so checked evolution and it's up to 0.687 already, with 1% due to time, 99% from pollution. So clearly the evolution due to pollution is completely unrelated to whether the pollution is actually touching the biters. Guess I need to ramp up military science and build defenses before my pollution cloud does extend to the biters, or I'm gonna get stomped by bugs!
|
# ? May 20, 2019 07:26 |
|
I recall The Locator mentioning this a while back but sustaining 1k science/min is really tough. I rebuilt my main base with beacons and I now peak at around 1.2k science but things start falling apart after about 15 minutes. My current bottleneck is Stone because at some point 8 lanes of dirt simply isn't enough.
|
# ? May 21, 2019 18:48 |
|
RVWinkle posted:I recall The Locator mentioning this a while back but sustaining 1k science/min is really tough. I rebuilt my main base with beacons and I now peak at around 1.2k science but things start falling apart after about 15 minutes. My current bottleneck is Stone because at some point 8 lanes of dirt simply isn't enough. I got that base stable at 970spm (1 silo) but it seems like I am doing nothing but replacing shrinking resource patches with fresh ones, but the time it takes to get the outpost set up and rails laid just means another patch of something is drying up so I have to do another one. The only way to get that base to scale any higher would be to completely remove the rail spaghetti and smelters currently supplying the beginning of the bus and replace it with a properly designed rail network supplying new remote smelter lines. My problem now is my ups has dropped into the high 40's and it's noticably slow now. Not sure if I will work on it much more. In the meantime I've started a modded game using Krastorio which seems to be a lighter version of Bob's/Angel's but all in one mod. Playing with Tinystart which really makes the very early game much more pleasant, but the raw ore requirements for plates are kind of brutal (3 ore and 3.2 seconds per plate). It takes 3 yellow belts of ore to get a single full belt of plates.
|
# ? May 21, 2019 20:47 |
|
Make a rail line to a really far out spot and you'll find very rich patches. Long trains to move the ore on their own network.
|
# ? May 21, 2019 21:41 |
|
KVPM basically requires something like Angel's Infinite Ores; the cycle of breaking down old outposts and building new ones further out remains but the core of every patch outside the Starting Zone is undepleteable ore tiles. My plan was always to use identically-named train stations that only switch on when they have a train load of ore on hand, as described by a lot of folks in this thread. 2-3 old stations per train and an idling area with a siding for each train; throughput numbers require that the receiving stations never be occupied for longer than the minimum amount of time to unload and pull away. It has been pretty awesome reading about folks actually reaching for what I basically did the math and groundwork for and then noped out of, thanks for y'all's stories ~~
|
# ? May 21, 2019 23:06 |
|
The Locator posted:I got that base stable at 970spm (1 silo) but it seems like I am doing nothing but replacing shrinking resource patches with fresh ones, but the time it takes to get the outpost set up and rails laid just means another patch of something is drying up so I have to do another one. ... Are you Tuplex?
|
# ? May 21, 2019 23:22 |
|
|
# ? May 25, 2024 13:22 |
|
KirbyKhan posted:... Are you Tuplex? I am not!
|
# ? May 22, 2019 00:09 |