6 ms·

I'm far beyond the point of having any shame about the fact that I have about 2500 hours in this game. This game only starts when you launch your first rocket.

I'm far beyond the point of having any shame about the fact that I have about 2500 hours in this game. This game only starts when you launch your first rocket. Building huge Megabases is where the real fun lies for me, and that's probably true for most players. Trying for a target such as consumption of 10K science packs per minute, sustained. The size of that kind of factory is beyond anything, but it is awesome. Unfortunately, I'm discovering that my DDR3 era PC can't cope with such large bases, I need faster gear...

sandworm101 6y ago
2500? Noob. You aren't serious about factorio until you are measuring your case for a threadripper mobo.
- improof 6y ago
  racist!!
- leetcrew 6y ago
  maybe I'm taking this post too seriously, but I thought the consensus was that factorio is mostly bottlenecked on memory bandwidth?
  - Dylan16807 6y ago
    In that case still threadripper for more memory channels, but specifically a low-core version? But in that case you can do even better than building your own. The OEM-only WX line has 8-channel memory, and the cheapest one is a reasonable 12 cores.
    - leetcrew 6y ago
      good point! kind of absurd, but then again I'm the type of person to buy a 9700k and overclock it to 5GHz just to get more frames in csgo.
psyc 6y ago
Only 1500 here. Although I enjoy the early game very much, I agree that making a base as big as possible is where it’s at. One thing I notice is that most players talk in terms of SPM. I find this curious because I find it hard to get excited about the infinite science sinks, so I think in terms of launches per minute. I guess those numbers don’t sound as impressive?
- svrb 6y ago
  I think there's a bit of history involved; there's also quite a bit of added complexity from the different science packs vs just launching rockets (though I've barely planed any unmodded since the current science pack recipes came out).
- leetcrew 6y ago
  SPM and launches per minute are just different ways of saying the same thing unless you are launching so many rockets that space science is no longer the bottleneck for research. I haven't gone super far into the endgame, but in my experience this is pretty uncommon. also personally, I find rocket launches to be too abstract to care about as a goal. I get where you're coming from with the infinite science, but I find it a lot more satisfying when every new tier tangibly improves the efficiency of my factory.
- nwallin 6y ago
  SPM depends on LPM, but LPM doesn't depend on SPM if you just throw away the white science. So SPM is a more holistic measurement.
  - psyc 6y ago
    Ok, this actually makes sense to me. I forgot that the rocket factory is only a subset of total production.
- louwrentius 6y ago
  I see what you mean, but the SPM number is very easy to track, those rocket launches, much less so. Maybe with a mod, there was one but it died with the new Factorio upgrades a long time ago.
urda 6y ago
I enjoy Factorio for the trains, so I go for the train focused map and disable the critters. Why disable the critters? I'm just in it for the building and it's one less thing I want to fuss with after a long day.
- cooljacob204 6y ago
  Yeah trains are a huge part of the fun for me. I usually get mods that make resources more dense the farther away they are.
  - WGH_ 6y ago
    Isn't it in the vanilla game, though?
    - salzig 6y ago
      It is, for trainworlds.
      - jlogsdon 6y ago
        Resource patches get richer the further you go out always in vanilla, not just rail world. Richness is a function of distance from your spawn point.
- louwrentius 6y ago
  Yes, the train network is a very important part of the overall fun. I love trains and OpenTTD (although I don't play that anymore). I'm really waiting for an update where train stations will have queue's. This prevents 10+ trains racing en masse to that one single active/open station. It was in the works, hope it's in 1.1
nwallin 6y ago
> Unfortunately, I'm discovering that my DDR3 era PC can't cope with such large bases, I need faster gear... This is really unfortunate. "Fortunately" it seems to be a result of the simulation engine using only a single core- there is the opportunity for the developers to get a 2-8x speedup from switching to a multithreaded simulation engine. I imagine there's fairly loose coupling between, for instance, fluid calculations, biter pathing, and on-belt entities, which tend to require dedicated strategies for mitigating their outsized performance impacts. (to include using console commands to simply delete the biters) Those sound like fairly low hanging fruit for splitting off into three threads if there's no plan to completely multi-threadify the sim engine.
- wokkel 6y ago
  The game also allows multiplayer (I've seen 500+ players on a server). That requires some really strong consistency guarantees that you loose by adding threads.
  - nullc 6y ago
    Multi-threaded programming is in no way incompatible with utterly deterministic bit identical execution with sequential code. Factorio is already written in a deterministic way. The developers main argument is that they cannot get much speedup from concurrency because their software is primarily memory bandwidth bottlenecked and at least on consumer hardware using more cores doesn't substantially gain you access to more memory bandwidth.
- louwrentius 6y ago
  The developers have explained their reasons for the current constraints regarding CPU multi-treading and AFAIK it's not as easy as this. If you want to know more, they were very open about the development process so they will have a blogpost about this.
gww 6y ago
I sort of hope that eventually there will be official assembly plant, miner, smelter upgrades that can help reduce UPS issues. I always feel like once you can progress to tier 2 things (blue assembler, red belts etc) you can get to end-game upgrades quite quickly.
- louwrentius 6y ago
  Yes, at some point, building a megabase is just stamping a blueprint of a green circuits factory until green circuits are no longer a bottleneck, and things like that. There is no vertical upgrade path beyond the rocket. And that is too bad.
nullc 6y ago
Hm. I guess different people just like different things. I've kind of lost interest after building out all the non-infinite sciences. Scaling for the sake of scaling just doesn't appeal to me a lot. (To be clear to people who haven't played it: Finishing out all the sciences is something that cannot happen until after launching a rocket which is the nominal 'win the game' condition. I thought factorio was tremendous fun, and it ate up all my free time for a week until I got a rocket launched.) A lot of what I found interesting was these emergent challenges like train throughput being poor because of chests becoming unequally full and blocking the train with slow unloading... which can be resolved by using the logic control. I really liked setting up advanced oil processing to not block on overproduction. Stuff like that. In some sense I think the logistic bots kind of break the game-- they eliminate most logistic challenges, but on the other hand they're also the only way to both reasonable scale up capacity and also to handle minutia and worry about larger scale things. I guess I would have liked to see a parallel track of other technology which is harder to use that you could work on exploiting for those of us who don't find merely scaling up for the scale of scaling up that interesting. E.g. I think it would be neat if there were advanced chemical processes that required writing PID controllers to effectively drive, or resources scattered on the landscape that you needed to program autonomous tanks to go gather. Different enemies that require different strategies to defend against, etc. Maybe some of this stuff exists in mods, I haven't really looked at those.
- louwrentius 6y ago
  I see where you are coming from. It's just what motivates you in a game. The infinite scaling doesn't matter. It's about building out the factory itself, if that doesn't motivate you, it's just not for you. Regarding PID controllers, I've done that for real-life applications (hobby) so no desire to do that in a game. https://github.com/louwrentius/storagefancontrol https://github.com/louwrentius/storagefancontrol
  - nullc 6y ago
    > The infinite scaling doesn't matter. It's about building out the factory itself, if that doesn't motivate you, it's just not for you. Oh, I liked building a factory, in order to unlock then figure out the new mechanisms or defend against the bugs. If I didn't I wouldn't have played through completing all the science. But for me the game ended at that point because after that it was just more of the same with little more to discover. I can see how some people enjoy cranking it to 11, but that part isn't for me. Maybe I'm missing some of the challenge, but I feel like I could mostly scale to arbitrary size (I think my game currently keeps one rocket port with prod modules saturated) now by simply copy/pasting factory chunks and letting the bots replicate them. I guess I would find automatically furnishing long range construction with construction supplies kinda interesting to do... but I dunno why I'd want to try. :)

sandworm101 6y ago

2500? Noob. You aren't serious about factorio until you are measuring your case for a threadripper mobo.

improof 6y ago
racist!!
leetcrew 6y ago
maybe I'm taking this post too seriously, but I thought the consensus was that factorio is mostly bottlenecked on memory bandwidth?
- Dylan16807 6y ago
  In that case still threadripper for more memory channels, but specifically a low-core version? But in that case you can do even better than building your own. The OEM-only WX line has 8-channel memory, and the cheapest one is a reasonable 12 cores.
  - leetcrew 6y ago
    good point! kind of absurd, but then again I'm the type of person to buy a 9700k and overclock it to 5GHz just to get more frames in csgo.

psyc 6y ago

Only 1500 here. Although I enjoy the early game very much, I agree that making a base as big as possible is where it’s at. One thing I notice is that most players talk in terms of SPM. I find this curious because I find it hard to get excited about the infinite science sinks, so I think in terms of launches per minute. I guess those numbers don’t sound as impressive?

svrb 6y ago
I think there's a bit of history involved; there's also quite a bit of added complexity from the different science packs vs just launching rockets (though I've barely planed any unmodded since the current science pack recipes came out).
leetcrew 6y ago
SPM and launches per minute are just different ways of saying the same thing unless you are launching so many rockets that space science is no longer the bottleneck for research. I haven't gone super far into the endgame, but in my experience this is pretty uncommon. also personally, I find rocket launches to be too abstract to care about as a goal. I get where you're coming from with the infinite science, but I find it a lot more satisfying when every new tier tangibly improves the efficiency of my factory.
nwallin 6y ago
SPM depends on LPM, but LPM doesn't depend on SPM if you just throw away the white science. So SPM is a more holistic measurement.
- psyc 6y ago
  Ok, this actually makes sense to me. I forgot that the rocket factory is only a subset of total production.
louwrentius 6y ago
I see what you mean, but the SPM number is very easy to track, those rocket launches, much less so. Maybe with a mod, there was one but it died with the new Factorio upgrades a long time ago.

urda 6y ago

I enjoy Factorio for the trains, so I go for the train focused map and disable the critters. Why disable the critters? I'm just in it for the building and it's one less thing I want to fuss with after a long day.

cooljacob204 6y ago
Yeah trains are a huge part of the fun for me. I usually get mods that make resources more dense the farther away they are.
- WGH_ 6y ago
  Isn't it in the vanilla game, though?
  - salzig 6y ago
    It is, for trainworlds.
    - jlogsdon 6y ago
      Resource patches get richer the further you go out always in vanilla, not just rail world. Richness is a function of distance from your spawn point.
louwrentius 6y ago
Yes, the train network is a very important part of the overall fun. I love trains and OpenTTD (although I don't play that anymore). I'm really waiting for an update where train stations will have queue's. This prevents 10+ trains racing en masse to that one single active/open station. It was in the works, hope it's in 1.1

nwallin 6y ago

> Unfortunately, I'm discovering that my DDR3 era PC can't cope with such large bases, I need faster gear... This is really unfortunate. "Fortunately" it seems to be a result of the simulation engine using only a single core- there is the opportunity for the developers to get a 2-8x speedup from switching to a multithreaded simulation engine. I imagine there's fairly loose coupling between, for instance, fluid calculations, biter pathing, and on-belt entities, which tend to require dedicated strategies for mitigating their outsized performance impacts. (to include using console commands to simply delete the biters) Those sound like fairly low hanging fruit for splitting off into three threads if there's no plan to completely multi-threadify the sim engine.

wokkel 6y ago
The game also allows multiplayer (I've seen 500+ players on a server). That requires some really strong consistency guarantees that you loose by adding threads.
- nullc 6y ago
  Multi-threaded programming is in no way incompatible with utterly deterministic bit identical execution with sequential code. Factorio is already written in a deterministic way. The developers main argument is that they cannot get much speedup from concurrency because their software is primarily memory bandwidth bottlenecked and at least on consumer hardware using more cores doesn't substantially gain you access to more memory bandwidth.
louwrentius 6y ago
The developers have explained their reasons for the current constraints regarding CPU multi-treading and AFAIK it's not as easy as this. If you want to know more, they were very open about the development process so they will have a blogpost about this.

gww 6y ago

I sort of hope that eventually there will be official assembly plant, miner, smelter upgrades that can help reduce UPS issues. I always feel like once you can progress to tier 2 things (blue assembler, red belts etc) you can get to end-game upgrades quite quickly.

louwrentius 6y ago
Yes, at some point, building a megabase is just stamping a blueprint of a green circuits factory until green circuits are no longer a bottleneck, and things like that. There is no vertical upgrade path beyond the rocket. And that is too bad.

nullc 6y ago

Hm. I guess different people just like different things. I've kind of lost interest after building out all the non-infinite sciences. Scaling for the sake of scaling just doesn't appeal to me a lot. (To be clear to people who haven't played it: Finishing out all the sciences is something that cannot happen until after launching a rocket which is the nominal 'win the game' condition. I thought factorio was tremendous fun, and it ate up all my free time for a week until I got a rocket launched.) A lot of what I found interesting was these emergent challenges like train throughput being poor because of chests becoming unequally full and blocking the train with slow unloading... which can be resolved by using the logic control. I really liked setting up advanced oil processing to not block on overproduction. Stuff like that. In some sense I think the logistic bots kind of break the game-- they eliminate most logistic challenges, but on the other hand they're also the only way to both reasonable scale up capacity and also to handle minutia and worry about larger scale things. I guess I would have liked to see a parallel track of other technology which is harder to use that you could work on exploiting for those of us who don't find merely scaling up for the scale of scaling up that interesting. E.g. I think it would be neat if there were advanced chemical processes that required writing PID controllers to effectively drive, or resources scattered on the landscape that you needed to program autonomous tanks to go gather. Different enemies that require different strategies to defend against, etc. Maybe some of this stuff exists in mods, I haven't really looked at those.

louwrentius 6y ago
I see where you are coming from. It's just what motivates you in a game. The infinite scaling doesn't matter. It's about building out the factory itself, if that doesn't motivate you, it's just not for you. Regarding PID controllers, I've done that for real-life applications (hobby) so no desire to do that in a game. https://github.com/louwrentius/storagefancontrol https://github.com/louwrentius/storagefancontrol
- nullc 6y ago
  > The infinite scaling doesn't matter. It's about building out the factory itself, if that doesn't motivate you, it's just not for you. Oh, I liked building a factory, in order to unlock then figure out the new mechanisms or defend against the bugs. If I didn't I wouldn't have played through completing all the science. But for me the game ended at that point because after that it was just more of the same with little more to discover. I can see how some people enjoy cranking it to 11, but that part isn't for me. Maybe I'm missing some of the challenge, but I feel like I could mostly scale to arbitrary size (I think my game currently keeps one rocket port with prod modules saturated) now by simply copy/pasting factory chunks and letting the bots replicate them. I guess I would find automatically furnishing long range construction with construction supplies kinda interesting to do... but I dunno why I'd want to try. :)