How 'Factorio' seduced Silicon Valley and me

I agree: a well-designed educational course can feel like a video game in some ways, in that you're learning at a high, consistent rate.

Omg aoe2 is so addictive! Some things you might enjoy:

- YouTuber “spirit of the law”

- watching some build order guides. Fast feudal and fast castle build orders are super useful

- watching commentated games of the Viper

On the other hand, video games enable a wide breadth of intellectual experiences.

Being a simulated guitar hero and simulated pro skater is more enriching than the likely baseline of having zero experience with either.

And, video games can help in discovering real world passions — the number of guitarists who found their inspirational spark through Guitar Hero is likely significant. Same for Factorio or Minecraft -> programming.

If I could have got a steady hand with tweezers, I could easily have been a surgeon. Real patients probably don’t have noses that light up, which reduces distractions too!

Maybe you would have tortured yourself and still wouldn't be any good.

So? There are a lot of not fun aspects of programming; Factorio is what it's like if it was all just the fun parts.

i find tech hobbies fun because so may of the hard skills are transferable so i'm never starting from scratch, aka the hardest parts of a new hobby.

however, this is the trap i always fall into - i have these vague targets like "learn vue" and spend the whole weekend trying to figure out how to install node on a windows machine and run a basic test

Making Bluetooth 10% more energy efficient in the next few revisions is a terrible example because you absolutely cannot just tackle that by yourself. Not because it's technically hard, but because it's actually a people problem in disguise!

If you just walk up to the mailing list with the complete designs, documents, experimental results, schematics, trade offs, feasibility studies, you know what you're going to get? People saying "whoa, hey, great work but let's talk about this. I see here you've made assumption X about implementation area Y and that actually conflicts with the direction that we had in mind for the upcoming release, so let's talk more. To start, we'd like to see if we can explore option Z, thoughts?"

That ain't fun. It's rewarding but it ain't fun. Not like sitting down and messing with Legos in your own house is fun, or building a silly factory in Factorio is fun.

Trust me when I play minecraft, the moment I realize I don't have inventory sorting installed I quit the game.

Also, are there video games with ML-powered opponents or NPCs that'll act more like humans instead of the old-fashioned decision tree AIs? Seems about time. I know there are research projects that tried this for things like Starcraft 2.

Also saw Facebook's experimental Diplomacy AI. Wish that were usable the time we lost a player.

There are some LLM-powered services that claim to be copilot for games.

From what I can tell, it seems like generally they're doing RAG over the game's wiki, and then reading it to you in an attractive-lady voice.

where is our hacker news copilot when we need it?

I'm looking over the sentences in the comment you replied to and I can't figure out where you would have gotten lost, can you quote the partial sentence that confused you?

> It's the trucking game for truckers

Exactly yes, I wouldn't say it's only for engineers, it is perfect for smart people who chose a different path in life and need to feed this part of their brain with a fun simplified simulation of what it's like to be an engineer.

Of course, engineers would love this game, but we already engage in this type of activity all day much more intensely precisely because we love it, we already know how to do more interesting and valuable things with the same effort.

Similarly, I really lost the appetite for hard decision-making strategy games when I founded my startup. I was already having to take plenty of hard decisions, thank you very much, the real thing is much more interesting and more than enough.

Dave the Diver just turns into a slog is all. It's fantastic at the start when everything is novel, and all the systems hold promise. But the systems themselves turn out to not have good gameplay loops, so right around the underwater village it starts to kinda suck, which ironically is right around when all the systems are mature enough that to play well you have a giant list of "todo" items across a variety of systems, many of which you may find boring, but are obliged to interact with anyway, sort of like real life.

The advantage a game like factorio has over DtD is that you can kind of abandon huge sections and "start fresh" while still collecting rent from the work you did earlier to fund your new excursions. I guess DtD does this to a degree, but there's a lot less freedom of intellectual movement.

This is why I like Against the Storm. It is a novel (if familiar) challenge every time, and every time you "beat the odds" the game forces you to move on to the next harder challenge, "you beat this puzzle, you need a harder on."

> Once I get successful in it, I'm like "but I should be making money in the real world, not this world" as I do have side projects. the game does turn into a slog, if you approach it that way

Life is a slog if approached that way. Spending your time chasing more money instead of enjoying life is the quickest way to have a miserable existence.

Yes. Don’t you? Why not?

Not sure what "choose one" means here. I liked Factorio and Satisfactory, but didn't like Mindustry.

Well the reason I enjoy Factorio in the first place is because I love automating things, not necessarily the building part. There's an endless amount of things I can automate in my life or for others, and the effort I put into building those automatons, whether is personal or professional, continues to provide value after I'm finished. I have 1k hours in Factorio, while those hours did provide value to me, that value is pretty finite. That's my logic at least, I just feel guilty playing it looking at it through that lense. I got like 50 private repositories that aren't finished, or are interesting and fun that could provide just as much satisfaction and possibly continue to provide value in my life even after I'm finished with them. Right now I'm working on a little project to buzz myself into the front door of my building via an app on my phone that triggers the buzzer in my apartment that is a weird analog system that's 60 + years old (but funnily enough is called Auth Master). Not complicated or intense, but fun and interesting and will provide a standard of living improvement for me for years to come (assuming I don't move) after I finish!

For me it's me PhD thesis. It's a drug, and there's always something to do either on the code or writing or reading. I'm even a bit disappointed by Space Age.

Yes, and?

> You don’t need to make software for mass appeal.

Hear, hear. I would tell you all about the tools my household runs on but it is as you say. Still gives me a lot of satisfaction to make stuff that makes life better, and for someone more adventurous than me it might also be something you can use to try out new technologies or build something for on your CV. The only downside is that internet connectivity being down means, e.g., you need to remember to load the grocery list before leaving the house that has the server in it and better hope your phone doesn't decide for you that the page needs to be unloaded on the way!

> You don’t need to make software for mass appeal.

Some people, like myself, enjoy writing code but need a purpose to write code. If I don’t have a reason, or problem to solve, I can’t just sit down and start coding.

>This makes me sad.

I gotta say, I think I agree.

Maybe it's just rose-tinted-glasses, but I remember a time when software was split between "IBM-Corpo" culture, and zany SV/MIT/Caltech culture where people threw things at the wall and proceeded when stuff stuck.

It kind of saddens me that it feels like it's now only IBM-Corpo, and everyone feels the need to be ever-productive and adhere to strict rules and schema.

tl;dr : I remember when the fun Factorio game was qbasic.exe and no one blinked about it. We all had fun.

(p.s. I love factorio now too)

> You don’t need to make software for mass appeal.

In Factorio, the things you make don't need to have *any* appeal. It's just fun.

If you have more fun making random wonky software, that's great for you and I'm happy for you. But I personally find Factorio to be more fun than software when I'm trying to enjoy my days off. (at least, some of the time. Sometimes I have fun making software too but only if I have that 'itch'. With Factorio, it's fun even if I don't have an itch)

An app can be like a home cooked meal

https://www.robinsloan.com/notes/home-cooked-app/

> Git is weird and bizarre

The core of Git is incredibly simple and elegant (content addressable database). It's an amazing approach to how to represent the data that needs representing.

The UI is the challenging bit, and that's because Linus just slapped something on to demonstrate it, expecting that a proper UI would be built over the plumbing he'd created.

Used? Certainly. Well loved? Oh boy do I disagree.

I think it's an extremely powerful tool, and it's worth knowing how to wield it well, but that power comes at the expense of user friendliness, especially for junior devs who don't have an intuition of git internals and how commands map to them.

Nods in Emacs.

This behavior is known as "play". Many of us find it stimulating. If it doesn't work for you you have my sympathy.

> I can make whatever software I want, for whatever purpose

Such freedom is extremely rare when you get paid for writing or maintaining software.

Not everything is about impact.

I think a lot about this in the context of video game minmaxers/netdeckers who just copy a researched, optimal behaviour. It's completely alien to me but some people really get a kick out of just being an implementer, following rote instructions exactly to the letter.

yes once you watch those videos it takes away the problem solving element as well.

It can be but my point is when you face those types of challenges in you work and solve them you realize how much better and satisfying those challenges are in real life compared to Factorio which takes away the fun part of playing it.

> Development has fun parts and it has tedious parts. The tedious parts are what usually makes for a good product. Factorio removes that part; because it is not about producing a product, it's about having fun.

Factorio also has tons of tedious parts. It really isn't all that different in that regard in my experience.

There are many games these days that are designed to have a certain element of tediousness (grind) in them, because this gives you the sense of being productive without actually producing anything.

> That's up to you, but for me it'd be a curse, not being able to do thing just for fun anymore, because of always feeling a pressure to be productive.

I don't think they mean to say they always feel a pressure to be productive. If it costs you exactly the same amount of energy and you get the same amount of joy from it, then why would you not choose an activity that produces something of enduring value?

You're discounting the joy you get from actually solving a problem for yourself and other people which can provide serious long term satisfaction and monetary rewards.

For diversion, I try disconnecting from screens and doing something healthy that involves the use of my physical body. I find walking, hiking, swimming and running very enjoyable. I'm also a big fan of tennis, racquetball and table tennis. I still play video games occasionally when there's not an option to engage with the physical world for diversion but I really prefer chess to most modern video games.

Programming can still beat most games in terms of joy. I just miss that burst of satisfaction from completing something in programming because it’s never really done just good enough for production.

Factorio and other open ended optimization games scratch the same itch as programming without any payoff. Speed running feels different because execution is so important optimal changes over time, and you get a finite endpoint.

You're discounting what's actually being said here, and said by many people.

I don't play Factorio for the same basic reason as so many others in this thread: it's too much like programming, and I want to use that energy actually programming. Something I enjoy very much, something which has paid every expense I've had for the last fifteen years, and something which directly produces value for other people.

I also pick up a little handheld most evenings and play games on it. Lately, Puzzle Bobble and Galaxian. Scratches a different itch.

The dichotomy between 'play Factorio' and 'program computer' does not reduce to the dichotomy between 'have fun' and 'joylessly pursue productive endeavors'. Clearly for some people Factorio doesn't produce the former dichotomy in the first place, and more power to 'em. For others it does, and the conclusion that this means we don't like fun is simply invalid.

I don't think they're saying that per-se, I think they're saying: it's perfectly ok to both "have fun" AND "not build something meaningful" at the same time. I don't think you need to read in that having fun necessarily excludes doing something meaningful.

While we're on the topic of assuming intent: that invented word "problematic" - problematic to whom? Something being a problem is almost always a subjective stance, not a universal truth as that word implies.

The core gameplay loop is still very much the same. More complex recipe is not really a more complex game. It just means more of the same planning and calculation but with more variables - which is to say more tedium not more fun.

Looks like you tried the game a long time ago. There is totally a way to automate building with the Blueprint Designer which basically allows you save blueprints of full buildings (or building parts or whatever you want) and to build them in one click.

In a stylistic way, yes. Not in the « what a beautiful multi floor truck station I built there, oh and btw, take this stair then on the mezzanine, take the door at the left to access the commands of the station. » way.

It’s just different games. Factorio is excellent as building production lines while Satisfactory is excellent as making feel you « in » your creations with 3D, a lot of architectural options, a really immersive sound design (quitting your noisy machine room to a corridor and hearing the sound go down when the door closes).

It’s different. But there is one important common thing with Factorio though : they are both made with a lot of love and with a lot of attention to every detail. In fact, both games could be an exemple of design and ergonomics for even professional softwares.

I mean, theoretically...

I read about a half hour every day. That means I probably clocked some 6 thousand hours in my lifetime. Imagine all the time I could’ve invested in playing the rat race more, in those wasted 3 years of books :(

Maybe; I limit my gaming this way - to 40 minute increments, and then go do something else.

If anything I think it makes me prioritizes games where 40 minutes is a satisfying adventure and skip games where 40 minutes is just time-wasting.

Factorio doesn't go on sale. The author's reasoning being it's disrespectful to those who paid full price for it.

Edit: official forum link stating thus: https://forums.factorio.com/viewtopic.php?p=159626#p159626

People look at me funny when I say this, but it's true.

I work in performance - a space where we're thinking about threading, parallelism and the like a lot - and I often say "I want to hire who play with trains". What I mean is "I want people who play Factorio", because the concepts and problems are very very similar. But fewer people know Factorio, so I say trains instead.

I think I know why it's enjoyable even though it's so close to work, too. It's the _feedback_. Factorio shows you visually where you screwed up, and what's moving slowly. In actual work the time and frustration is usually in finding it.

> In fact during my flight training my experience with simulators imparted a few low-level bad habits I had to work through when it came to operating the controls.

When I first started flying my instructor specifically recommended against flight sims for that very reason.

Now that I have my PPL I only use them to practice approaches for new airports that I’ve never visited before and to familiarize myself with the cockpit layout for planes I haven’t flown before. It helps build some minor muscle memory so I can focus on the harder parts of flying.

Yes, factorio is not coding, though it follows some principle of programming, such as input-process-output, workflow, load balancers, modular design, etc. Many players say that the factories often resembles PCB.

but that's the extend of it, you don't have to do heavier programming practice like algorithm, complex state management, input parsing and validation (unless sushi), or anything involving combinators until finishing the 1.0 version.

Factorio is not a puzzle game. You can get puzzle like gameplay if, for example, you are on a map with heavily restricted space, or are trying to micro-optimize something. But that is analogouse to code-golf or microoptimizations in programming, which are fairly puzzle like.

For typical gameplay, the only puzzles you run into are the ones you build yourself into. The name of game is to design your factory in a modular and extensible way so that you do not build yourself into a puzzle.

"not in any way similar to coding" is a very strong statement

many people feel it feels similar to coding, in the way you have to slowly refactor designs + work involves a recursive breakdown of tasks + tracing/debugging of issues

Many people are quoted in the article claiming it is in fact useful for training/educating software engineers.

You don't find coding similar to solving puzzles in any way?

I can think of a bunch of similarities just off the top of my head: throughput management is probably the most critical skill in factorio, and obviously comes up all over software engineering. Managing production belt chains is not unlike managing your program's call stack. Ensuring the right resource reaches a production facility is like managing input types.

What about Zachtronics games like TIS-100 and Shenzhen/IO and Exapunks that are literally about coding (albeit in a simplified assembly language).

Theoretically, I guess telling HN people they should be out saving lives could save lives.

To his credit, Peter Singer actually is a crazy ascetic.

I mean, you work to make money, and you can donate that money to all kinds of charities that do save lives.