Factorio

I tend to like it when a game creates a specific challenge for me to overcome.

Well, sandbox mode does have one specific goal: assemble and launch a rocket. (To get help? To survey the area? Dunno. But launching the rocket is the goal.) When you finish that, the game is officially over, but you can keep playing with it if you want.

To get to that point, there are dozens of subgoals you'll need to accomplish, and then just zillions of micro-goals as you're putting your factory/ies together.

I find it absorbing, and I'm looking forward to playing around with the new version. I don't play it through Steam, so I'm not sure how much time I've put in, but I bet it's at least 60 hours.

I did try the demo and enjoyed what I saw of it. I wonder how much the game is going to keep changing and if I 'win' now will I want to play again later. Pretty sure I will get this at some point, just a matter of now or later. I wonder if RimWorld is more 'done'.

I played mostly in Sandbox and I did have fun restarting when I 'win' because there's always room to optimize or to try a new flow. Its pretty frequent to get nowhere near winning and restarting to try different things.

To me, the changes over time have been too slow. I think the game is worthwhile for its cost but it's not evolving in a way that has brought me back since that first 40 hours or so that I spent.

I go back and forth between Rimworld and Factorio. Like them both, but Rimworld sometimes turns me off a bit because I start to feel like I'm playing the Sims. So-and-So is SAD today, weep! They're both good games that pulled me in quickly and good values but if I had to put them in order I'd say Factorio before Rimworld.

I've spent maybe 5 hours playing. I finished the first tutorial campaign and went to start the next campaign. There was so much stuff already built I got intimidated and went to the sandbox. The interface can be a little clunky, mostly due to how things are organized. It's also difficult for me to decide what to research because I don't know what I need, nor do I want to read some how to play wiki or watch a video.

Building some connected mines -> furnaces -> machines can be pretty satisfying, but there is frustration not knowing what to do. I imagine I should continue with the campaign because it is probably trying to introduce some things in a logical manner.

I have to say that building things in SpaceChem felt more rewarding. SpaceChem can get pretty complex without bombarding the player with 20 things to build and mine just to get some things done. So, I'm mixed on Factorio. It may be something that really requires some time to get into, but there are so many things to play that I already own that I may not have the patience to really give it the time it needs.

One of the early things to shoot for is getting coal power plants online. Once you've got electricity, the game changes dramatically; not only do you gain access to many new things, it's also far more efficient to burn coal to make electricity than it is to burn coal directly in the relatively few devices that can handle it.

The optimum setup, for power plants, is 1 offshore pump, leading to 14 boilers, leading to 10 steam engines. If you've got a good spot, you can use conveyor belts to deliver down a center 'aisle', and have short inserters to load a first row of boilers on each side of the conveyor. Then you can build more lines of boilers, with long inserter arms, which take coal out of the first row of boilers and insert it into the second row. I think you can end up about 5 rows deep this way before you run into supply problems.

In other words, the coal goes from the conveyor belt to the first row, gets grabbed out of the first row and taken to the second, and then to the third, and so on, possibly getting burned up in the process if one of the engines needs a coal right then. Eventually it stacks up in the outermost boiler, and then the backup moves toward the belt, until all the boilers are full.

This lets you drive a hundred total steam engines from one conveyor belt, and you can expand slowly, adding new lines of engines as you need them. Note, however, that this produces a lot of pollution, so you'll definitely want lots of turrets on the outer edges. Note also that you need to keep a close eye on those belts and your coal supply, because restarting your factory if the power plant goes offline can be super painful. (as in, game-ending, as you're swarmed by creatures your turrets aren't shooting.)

Eventually you can go fully solar, but solar plants are fairly inefficient in terms of both resources to build them and land area used by the panels, meaning you have to guard more territory.... but you'll be pissing off the creepy crawlies less.

This also points you at a second goal.... trying to get to laser turrets. These are very effective against the critters, but require a fair bit of tech to get operational, and oodles of power, so getting accumulators (aka batteries) up at the same time can be a very good idea. But they don't need ammo, so you don't have to worry about keeping them loaded.

I didn't get to the point where I needed multiple steam engines yet for power, but I had a miniature version of what you are saying - coal down a conveyor belt being fed into the furnaces. It sure beat running coal manually to the places that needed it!

Yeah, electricity is a huge, huge deal. It's the big transition to real automation, the kind you don't have to babysit anymore.

My epiphany was...

Spoiler:

finding out the inserters could use both sides of the belt...feeding coal into one side and ore into the other lane and both would be supplied to the furnace with one inserter.

Only took 3-4 hours to realize that, but I'm glad I found out through my own little epiphany. Solving those little logistic challenges feels so good, no matter how small.

You were right, Malor. I could do worse with $20. So good! (Awww, crap! How did it get to be 2:30am?!? )

Running everything off that one conveyor belt gives a nice sense of accomplishment, and then you get to learn what 'single point of failure' means. Factorio, failing is fun!

I keep bouncing off this game. I'm still waiting for there to be some point to all this busywork. The process of setting up a factory from stone age tech to robotic automation doesn't have any payoff other than a placeholder pat on the back for launching a satellite.

Tamren wrote:

I keep bouncing off this game. I'm still waiting for there to be some point to all this busywork. The process of setting up a factory from stone age tech to robotic automation doesn't have any payoff other than a placeholder pat on the back for launching a satellite.

It kind of reminds me of Minecraft before there was a dragon in The End. Playing the game was just Minecrafting for Minecrafting's sake.

Factorio is a bit like that at the moment, it's mostly driven by your own goals. Hey, can I automate launching a rocket every 5 minutes? Can I create a giant train network? Can I use robots to auto-build a complete base from practically nothing? Etc.

I've seen all sorts of wonderful mobs like Factorissimo, Bob's Mods and the like. But even if you have pocket-dimension factories and 20 different levels of tech progression and automated hunter killer robots and whatnot. You still come back to the fact that the game ends with a placeholder.

I know they have some neat stuff on the horizon like space platforms but all of the recent patches have been optimizations and bug fixes with no real added content.

Tamren wrote:

I know they have some neat stuff on the horizon like space platforms but all of the recent patches have been optimizations and bug fixes with no real added content.

I haven't gone back to Factorio since the end of May and I have to admit that I didn't even make it to the rocket. I kinda got lost in micro management, optimising tiny details and losing focus. I really didn't care too much about the final goal... Still, I'm waiting a while until I get back to it. This is my first early access game and especially those who played Kerbal in early access are always talking about how awesome it was to see changes whenever they fired up a new game. So, I'm looking forward to experiencing that with Factorio as well... I have to look into the aforementioned space platforms. It doesn't mean that they want to expand the factory building on space level, does it?

You still come back to the fact that the game ends with a placeholder.

So are you wanting more of a story, as in, 'you're trying to be rescued'? In essence, it's a game about building an economy, and it seems like almost any ending point would be arbitrary. "Okay, your economy is done now."

If you're building your economy to accomplish a specific goal, that's pretty much what the rocket is.

Maybe it's that the end point is so black/white? Perhaps a more grayscale goal would be preferable? In other words, maybe you need to build an economy that can field 10,000 tanks, and it's a matter of time/materials invested in factory expansion versus how many tanks you're putting out.

Or maybe there just need to be more guns/butter decisions, to make it more interesting?

I just find the interface bizarre and unpleasant... At least, from six months ago. Dunno if they have changed it yet...

Yeah, the interface is pretty weird. I was able to adapt to it okay, but it was work. The way research interrupts everything is particularly annoying.

It hasn't changed recently, and I'm not aware of any plans to change it anytime soon.

edit: well, they made some quality-of-life improvements in the last build, mostly indicators and things to make placement easier, but they didn't change the actual interface, just some of the ancillary graphics.

Malor wrote:

So are you wanting more of a story, as in, 'you're trying to be rescued'? In essence, it's a game about building an economy, and it seems like almost any ending point would be arbitrary. "Okay, your economy is done now."

I thought about this for a while and I think I know what the game is missing. There is zero random element to this game right now. Oh sure the resources spawn in random locations and random amounts, but it's always the same resources. You're only going to ever build one factory, so all it really determines is how far your train tracks and power lines have to stretch in order to reach a mining outpost. There isn't much of an antagonist either, biters are no more interesting than a fire ant infestation. They'll leave you alone if you don't piss them off, but you are forced to fight them and loot their pink pearls because of the tech tree. You can build an impenetrable fortress and reduce them to noise pollution and nothing more. Or you can just kill em with fire so they are never a problem again.

Contrast that to games like Rimworld or Dwarf Fortress that have a tech tree but everything else is completely random. What location you choose to embark on shapes your entire experience in Dwarf Fortress, and to a slightly lesser extent Rimworld as well. It's perfectly possible to embark on the beach beside the ocean, or on the very lip of a volcano. No two games are alike even if you embark on the same general terrain. You also have to worry about the thoughts and needs of your people, in Factorio you don't even need to think about starvation.

I don't think they can do much to change the situation given how Factorio is currently set up. You can't just make a map without oil the way you can play Dwarf Fortress without steel because it would break the Factorio tech tree. You can't even play without the biters because you need artifacts to make high tech items.

Finally finished launching my rocket. So daunting when I first started but once you get the pieces going it's just a matter of time. My system definitely could be improved. I think it takes like two hours for me to finish making the rocket. via automation Still if I wanted to boost the speed it would just be copying what I have. Did this on peaceful. Seems way to tough otherwise since you have to expand so much. Kind of wish there was some more power options in the game. Steam and solar don't seem to be enough.

Played with portable roboport and logicistic bots near the end. They help a lot. Plus using schematics with them and expansion is fast. Would love to use them more but robots are expensive.

Only think I sort of want to do completed all research but I guess I won't spend that much time. If I play again it'll either be multiplayer or with mods or both. Great fun overall.

Self-replicating factory that builds out around itself in a spiral of large cells. Pretty amazing. Requires two mods on top of Vanilla Factorio to work. One that allows automated blueprinting, and another for automating naming of train stations and moving them around. It's amazing.

Just stunning. I think I have to come back to this game. Been craving a game with good building and its been long enough that I've forgotten 'optimal' setups.

polypusher wrote:

Just stunning. I think I have to come back to this game. Been craving a game with good building and its been long enough that I've forgotten 'optimal' setups.

I just got back into it a bit ago with a new peaceful game on a seed that was posted on a reddit forum. Just automated a smallish steel smelting setup and should move on to oil soon. Green and Red science are completely automated and produced at respectable levels for now.

Version 0.15 is due to be released on the experimental branch on Tuesday. This is an enormous update, which will be adding a great many things. A partial list includes:

  • The research overhaul. Blue science is becoming easier to craft, and the alien science packs are going away, to be replaced by four new types of science pack. This should both smooth out the difficulty curve and increase the complexity of the endgame.
  • Infinite research. The research chains that provide upgrades will no longer end. Instead, they will increase exponentially in cost with each additional level, forever. They are also adding new upgrade chains, such as to mining productivity.
  • Nuclear power. This includes the addition of uranium as a new resource. Apparently it will require distributing sulfuric acid to the miners in order to obtain it, but in exchange it can generate an enormous amount of power. They've also completely reworked how boilers work, so be prepared to re-think all of your steam power plant designs.
  • The combat overhaul. The game has needed a combat rebalance for a while, so I'm curious to see what they end up doing here.
  • High resolution graphics. They've been talking about this one for a long time, and we're finally going to start seeing it in 0.15. They will be phasing in the new graphics over the course of 0.15 being on the experimental branch.
  • Tanker cars. Now you can transport fluids on trains without all that tedious barreling and unbarreling.
  • A programmable speaker. Now your circuit networks can make sounds! This is clearly the most important addition.

And there are undoubtedly other things I'm forgetting or don't know about. I know what I'm doing next week!

So excited! I might actually launch a rocket!

ooooh. i had been thinking about getting back into this recently, but i might hold off a little for this to hit!

Ditto! I'm very interested in the science pack additions. I guess I'm very interested in all of it...who am I kidding...any excuse to play!

Man, I thought I had a pretty large-scale factory, with the ability to launch a rocket at about 2/3 of the speed of the assembler. But the one that V453000 links to in his post (imgur album) utterly dwarfs mine. My current extraction rate is about 15K iron and 13K copper/minute, but his setup looks like it's ten times larger. He mentions having a hundred thousand solar panels! (I'm at about 14K).

That's just... holy cow, that's huge.

The changes in 0.15 look outstanding. This has been a long cycle, but it appears to have been well worth the wait.

0.15 is out! If you play the Steam version, just opt-in to the 0.15.x beta.

Note: I edited this very heavily, and I still don't feel like it's a very good post, but I'll leave it up anyway. It's kind of wall-of-text... skip down to the boldface for the important part.

So I've been playing with 0.15, and am mostly pretty happy with it. I've been getting micro-stuttering, which is a first for Factorio; turning on vsync at least hides whatever the problem is. I didn't feel like I got much traction with my bug report, either, so I'm not real hopeful the problem will get fixed until it gets much worse.

I spent a long time figuring out the "Kovarex Enrichment" process for uranium, which is this really weird problem born of a bad UI; he wants to require an initial pile of 40 U-235 before you can start manufacturing new 235 from the common 238. So the only way he could make that work, in the existing code, was to have it input 40 U-235, and 5 U-238, and then output 41 and 2. This is a really weird setup,, because you're having to unload and reload the exact same stuff into the exact same centrifuge, while identifying and extracting the surplus. It took me quite awhile to work out how to do this safely, in a way that guaranteed I'd never run short of my 'seed crop', as it were. I've subsequently browsed the forums, and see more elegant solutions than mine, but I'm pretty happy with how mine worked out.

This took a long time to scale up to the 8 centrifuges, probably several hours. Generating the initial 40 235 was the slow part; every time you process 10 ore into 1 uranium bar, you have a 99.3% chance of getting U-238, and 0.7% chance of U-235. So this means, on average, you'll need to generate about 5700 bars of "junk" to get the 40 you want. Once you've got that initial seed, things accelerate dramatically. The base machine can produce 1 new 235 per minute; with two Speed 3 boosters, it's every thirty seconds. This means it takes about 20 minutes to seed the next centrifuge, 10 for the third, 6 minutes and 40 seconds for the fourth, and so on. I seeded up to 8, and then sat back and let it churn away; the sushi-belt design I used ends up holding about 120 U-235 per reactor... the 40 that are presently being processed, and then 80 in reserve. I could probably engineer around that with some effort, but generating enough 235 to completely fill all the inputs only took about another 40 minutes, and I didn't have to babysit.

As an upside to all this, I also learned a fair bit about circuits that I hadn't before. Once I had a good-sized seed running, I started peeling out surplus into an 'out' chest; I arbitrarily decided on having 200 U-235 there, and a store of 18 times that much 238 in one nearby. I could have gone for more 235, of course, but maintaining the proper ratio would need more than one chest of 238, and I didn't feel like fiddling with it.

So, right now, I've got four really basic controlling circuits running. All input from ore processing goes into a chest line; the first circuit shuts off the power to the ore processing bank at 10K total units of U-238. (there's nothing specific about this figure, it just seemed reasonable.) The last chest in the line has two 'out' inserters, one that feeds the Kovarex cycle, and one that spills off 238 to make fuel rods.

The Kovarex sushi belt is fed by, and feeds back into, a steel chest. There are two 'in' inserters and one 'out' inserter to feed the Kovarex reactors... the two 'ins' make sure that any surplus goes into the chest, rather than staying out on the belt more than it needs to. The one out runs all the time, constantly spitting out lines of U-235 for the Kovarex reactors to grab; nearly always, this is ignored, and loops back into the source chest again.

There's also a little, fourth arm set on that chest, which feeds the final 'out'. I arbitrarily decided on having 200 U-235 in my 'out' pile at any given time. A circuit reads that box status; anytime it's below 200, the fourth inserter on the U-235 loop chest starts pulling them out, one at a time, until the final out box hits 200 again. Simultaneously, the main out for the U-238 chest line fires, and starts feeding U-238 into the Kovarex input. It's not especially accurate, and I believe U-235 will slowly build up in the system, but this isn't really a problem. Even if the 235 sushi box and belt get totally choked, it won't matter, because that's my final output -- it's okay for it to be choked. And my 238 input will never choke, because it shuts itself down anytime it hits 10... and it's got like 40K of empty storage. (it's a long line, I wasn't sure how much I was going to need.)

The final circuit reads the status on a second box, adjacent to the U-235 out. That goes to an arithmetic unit, which divides the total U-238 by 18. If the adjusted U-238 total is less than the U-235 total, the second out on the big 10K bank of U-238 ore fires up, and starts spitting out ore until enough has showed up and been loaded into the final output. (This, in turn, will probably enable the ore processing centrifuges until the 10K source bank is refilled.)

The overall system shouldn't choke on anything, shouldn't overfill itself, keeps up nicely with 8 Kovarex units, and depowers the ore processing when not in use. This, indirectly, mostly powers down the mining as well. I still have the inherent drain of those devices, since they're suspended on full output, rather than being truly off, but, well.... power's not much of a problem, anymore.

I also set aside an emergency seed of 320 U-235, enough to completely reseed my entire production line, if my logic turns out to be faulty. That's my parachute, as it were, in case I screwed the pooch somewhere.

After hours doing this, I finally built a reactor. I was going to model my first attempt after Kovarex's post on how he did his, but I spotted a slightly cleaner design in their forum, and used that as a blueprint. (functionally, it's just about identical, but it's very nicely laid out.) In the process of building it, I finally figured out how to use a personal roboport, and also figured out how to get robots to chop trees and remove rocks. I never realized that robots could do this, and me with a couple hundred hours of play. Sigh. It's a "deconstruction planner", by the way... hit B for blueprints, click the Deconstruction Planner button, and it will give you a red icon. Drag that icon over anything you want removed, and construction robots will do it if they're in range. If you've got a personal roboport and some construction robots in your inventory, stand nearby, and watch things disappear like magic, but the main logistic network can do it too, as long as everything's in the green zone.

So I learned a ton of stuff, and..... realized that I unbelievably over-engineered my uranium solution. (boldfacing because wall of text, and this is the really important bit.) 1 U-235 and 20 U-238 make ten fuel cells. Each fuel cell will last one reactor about two minutes. That means that 1 minute of unboosted output from a Kovarex centrifuge will drive 20 minutes of a reactor, or about a 20:1 multiplier, Put two speed modules in it, and a single lonely Kovarex centrifuge can drive 40 nuclear reactors, all by its lonesome.

Me? I've got 8 boosted centrifuges. I should be able to support at least 320 reactors. Assuming the smallest reasonably efficient reactor size (4 units making 480 megawatts), that's at least 38 gigawatts of power generation, and likely a fair bit more, with a bigger reactor setup. (To put this in context, I was bootstrapping up from coal, skipping solar to go straight to nuclear; I used my 180MW of coal power to build something that nuttily large, without even realizing.)

So. A little advice from experience: don't overdo it. You don't need much volume from your uranium facilities!

edit: but, start uranium mining and ore processing as soon as you reasonably can, so those machines can be churning away to make your seed stock. At the very least, you'll need filter inserters so you can separate into separate chests for later on, when you eventually research Kovarex enrichment.

Because of the wall of text there, I figured I'd repeat this bit: start your uranium processing early. You'll need to make about 5700 uranium bars to get your seed stock to get launched on your nuclear cycle. Those 5700 bars can then last you an exceedingly long time and give you a huge amount of power, but they take a LONG time to dig out of the ground and get processed.

A 1:1 ratio on miners to centrifuges seems to be about right, for making the initial bars. The uranium patches I've tapped so far have all taken exactly eight side-by-side adjacent miners, which didn't quite fill 8 centrifuges... one was a little idle. At 10 seconds per bar, times 8 centrifuges, you can expect about 48 bars/minute. To generate the 5700 you'll likely need, this will take on the order of 2 hours, maybe a little longer because of that last semi-idle centrifuge. Once you've got your first Kovarex unit going, you can seed the next in 40 minutes unboosted, 20 fully boosted... so that single unit is about 3 times faster than your whole existing stack, six times faster if you speed boost it.

At that point, your need for U-238 drops to about 22/minute... 19 to make that minute's 10 fuel cells, and 3 to make the U-235. (double to 44/min if you're boosting). You can use the remainder for other purposes, like uranium rounds. Or you could scale up and make atomic bombs, I suppose, although I haven't tried one. (I'm in peaceful mode to learn 0.15.)