The best part of the vid is from 5:35 onward, where it shows pretty decent detail of the city.

It's interesting to see the way "perfectionist" players approach their games. The details of the optimization are often surprising. Even the specific choice of factors to optimize for are interesting. In SimCity, one player shoots for maximum population; another might go for the quickest time to reach a particular population level; yet another might seek to maximize industrial production. I like seeing how they work within the rules to reach those different goals.

Yes. SimCity is not really a `game', but more of a sandbox, where you set your own goals. Of course you can set your own goals in more traditional games as well, but they tend to come with a pre-defined one.

If someone had fun, how could this time be wasted?

It's interesting, that the kind of people, that see this as wasted time, because they think this isn't changing the world, that it isn't ethical to use your own time this way, that these people will always criticize themself and others regardless what they do, because there's always something more meaningful to do.

At the end these people won't change anything in the world, but just bring harm to themself and others.

In some ways it's equivalent to solving any other seemingly esoteric theoretical problem that people seem to like solving (and some people get PhDs for doing so!). He took a formal system (a particular simplified model of city simulation), and produced a constructive proof of some properties of the system. Of course, the SimCity simulation parameters are unrealistic in a lot of ways, but it doesn't seem inherently any more absurd than the people who get PhDs for game-theory theses that also rely on absurd parameters in their problem setup.

I'd be interested in seeing more of the analysis that went into designing this.

I'd be interested in seeing the maker give a lecture on that. Well that and getting a copy of the city and removing all the firehouses.

I'm a little disturbed at his population distribution in the final city. Scary how his optimal city has no one over age 60.

> Scary how his optimal city has no one over age 60.

Logan's Run or Soylent Green?

Well, that last city was optimized for population count, not longevity.

or culture.

or life quality.

bunch of commies in this thread!

i wonder if there's a scriptable game like simcity (maybe freeciv? i dunno) ... that could make for a good testbed for certain optimization algorithms

I'm not sure, but you just made me miss the day of scripting BBS Door games such as Trade Wars 2002.

I do not understand why such a large piece of a person's life would be spent doing this. Even model railroads have some tangible significance -- it is possible to share them with others in such a way that the others enjoy them and you bring happiness into somebody else's life. Perhaps this is because I find value in helping others. Some may find value in simply being pre-occupied as they wait for death. To each his own. I do not understand a lot of things.

What I found totally amazing wasn't spending years trying to optimize solutions to random problems others have created, but how this guy made a super-awesome video explaining how he did it. Very impressive. Loved the music and the camera movements. This guy should start charging to do demo videos for software products.

> I do not understand why such a large piece of a person's life would be spent doing this.

You can share your digital artefacts with people all over the world. Isn't that much better by your metric than a model railroad?

It gets tough to debate life values, because we all come from a different place, but I'll take the question as just for me. I can't answer for this other guy. His reasons are his own.

The problem with spending so much time optimizing an artificial mathematical model (game) is that less work and time is required to create the model than to optimize it. Therefore (to me), it's the losers than spend time working on other people's problems that never move up to the next level: being creative. It's just a mechanical process. (and yes, I'm aware that some creativity is involved in the optimization)

My value system is in creating something for other people. So creativity and value to others, are the metric I use when allocating my time. I guess if such a electronic work could benefit people -- by making them happy for a bit or providing them with information they needed or something, it would be worthwhile. I mean heck, under my value system you can stretch a lot of things to make them work. The question (for me) is whether or not those 3 years could have been better spent.

Yes, at the end it's all gone anyway. That doesn't mean piss away your life. If anything, it means you only have a certain amount of time to do things you find valuable.

My opinion only. To each his own.

Mathematicians mostly study the surprising consequences of what others create; and it's often easier to take action than know the consequences, as the latter requires perfect understanding. Poorly understood automation is the Frankenstein fear, a real danger, as each victim of Murphy's Law will rue.

It's the pure science vs. engineering chestnut, that old favourite of science fiction: the pursuit of understanding for its own sake vs. making something useful (the trope has abstract knowledge as crucial in the end). But in reality, one isn't better than the other, both are needed, and it is, as you say, a matter of personal value.

Here's a related comment (you might like the cartoon too): http://www.reddit.com/r/pics/comments/bw3le/caveman_science_...

> The problem with spending so much time optimizing an artificial mathematical model (game) is that less work and time is required to create the model than to optimize it. Therefore (to me), it's the losers than spend time working on other people's problems that never move up to the next level: being creative. It's just a mechanical process. (and yes, I'm aware that some creativity is involved in the optimization)

I can see where you are coming from. But I do not entirely share your view. E.g. inventing the game of Go is much simpler than mastering it. But I would not call go players losers.

> My value system is in creating something for other people. So creativity and value to others, are the metric I use when allocating my time.

Why make an exception? Just use: "My value system is in creating something for people." You are worthwhile, too.

I've toyed with reverse Solipsism for some time now. Must be too much philosophy.

What did Steve Martin say once? He didn't have a lot of philosophy in college, just enough to screw him up for the rest of his life. :)

I don't seek validation in others -- but I do seek validation in my own assurance that I am creating something that will prove of value to others. By that metric, I question if I would be able to be so certain about that if I had spent those 3 years doing what he did.

Game-playing, drinking, singing, hiking, flying, etc are all wonderful activities -- but they are wonderful because they help me optimize the other things I am doing. When I start focusing on the pleasure they give me in themselves, I lose track of the larger goal. Perhaps even forgetting I have a larger goal. Not a good thing.

Reminds me of a Peanuts comic strip:

Lucy: I'm intrigued by this view you have on the purpose of life, Charlie Brown. You say we're put on this earth to make others happy? ... What are the others put here for?

As an aside, and because SimCity doesn't get mentioned much here, if you're on OS X and want to play Simcity 4, I spent a lot of time getting it to work "just so." The trick is.. Windows XP + Parallels 5 (I'm a big VMware Fusion fan but Parallels 5 boosts SC4 performance wise somehow) + software rendering (hardware rendering is a FAIL/unstable on both Parallels and VMware with SC4).. and I can run everything on high no probs on my iMac. Works a treat.

I'm using Ubuntu but this should still apply to OS X. SimCity 4 works nearly perfectly (both software & hardware rendering) under Wine. There used to be a problem with the installer but based on test results on the Wine AppDB that works now too. Also you can set a custom resolution using "-CustomResolution:enabled" and using eg. "-r1440x900x32" which helps make the game more playable on widescreen monitors. I used to run SimCity under VirtualBox but with larger populations it's much faster under wine.

I play SC4 myself on my netbook. It's a great time killer on a long plane ride.

As someone who's played SimCity for many years, I'm super impressed by this. Although he didn't use cheats, SC3000 was the best cuz it had a cheat where you could place a specific building and have it be there permanently--without any worry of abandonment. SimCity 4 is much harder because there are a lot more variables to deal with, including the fact that you can run multiple cities in a region.

Huge SimCity nerd back in the day here - his city design is impossible in SimCity 4; the new game requires all lots be connected to some kind of road. The secret to his success is that there are no roads, every single tile is connected to a subway station that's impeccably designed. There's no traffic congestion of any sort.

Personally, I think SC4 teaches people a bad lesson - that cars are an absolute necessity to urban planning.

Cars might not be, but roads certainly are.

How do you expect to even build the place without a road? Much less deliver furniture and other things?

It's not so much that roads are necessary, but that every single lot must be directly adjacent to one - it's entirely reasonable in real life to have large complexes that are pedestrian primarily, connected via some kind of mass transit, where road density is greatly reduced.

This is impossible in SC4, since every single building must be directly adjacent to a road.

Also, because of the city size limit increases between SC3000 and SC4, the pathfinding algorithm used for transit calculations is really barebones - your sims will only ever travel towards their destination. Even there was a direct subway line to their destination right behind them, they won't take it. This makes transit in the game a complete pain, since the model the game follows is counter-intuitive to the max.

The trick of course is, with a grid road network, this pathfinding works really well. This only compounds the problem - since a subway system, no matter how well designed, has a hell of a time competing with a road network - and in no time at all you have citizens screaming about congested roads, and a perfectly good subway system that's sitting at 5% capacity.

There are ways to "trick" the game - mostly involving designing your road networks in such a way that you create disconnected islands between demand zones. In essence, you're creating a ludicrously broken road network nobody will ever use just to get people onto mass transit. But it's worth it (in the game, and IMHO maybe real life also) - the capacity and density of rail mass transit blows the crap out of driving, and allows you to get people much further, much more quickly. Air pollution becomes a complete non-problem

Sounds more like a problem with the game rather than a lesson about cities having cars.

Yes. If you have a point you should just write it, because I don't understand it from that question.

An apartment building is effectively a large grouping of houses that are not directly connected to a road. People carry couches, or TVs. Maybe apartments are not like this where you live? Where I am, there will be a couple of square blocks of buildings that are all considered one complex...inside of that complex is a parking lot, and you walk to your building, then your apartment.

It wasn't meant to be snarky.

You mean an apartment complex, not an apartment building (at least that's how it's called here).

A couple of square blocks? But still I'm sure there are roads in between the buildings. And you have to transport the couch in a car or truck at least to the general area don't you?

No one builds buildings that have no road access whatsoever, and the only travel is via subway or walking.

Sim City is just an abstraction. You can pretend there are tiny pedestrian paths between the buildings and the roads only represent major roads.

I haven't played SimCity since version 1, but in those days I thought it unrealistically penalised any city without a huge rail network.

But I think the idea of a city without roads is unrealistic, except in the case of Venice where canals take the place of roads for delivery and construction purposes.

I wonder how many of the theories could be applied to actual urban planning.

A grid of subway stations with a stop at every block? Sounds rather expensive.

Besides, all attempts to build "perfect" cities in real life have resulted in dull and lifeless places. A certain amount of chaos is necessary to keep cities liveable.

This is exactly Jane Jacobs thesis in the greatly influential book "The Death and Life of Great American Cities"

What Jacobs found is that communities thrive when there's a certain amount of chaos and layered complexity that arises from people interacting withe each other and the environment. That's why places such as Greenwich village and many old European cities seem so vibrant and full of life - they're not centrally planned, they've had time to mature and people have had the possibility to interact with the environment and each other in the chaotic neighborhood.

I think it's a pretty straightforward phenomenon: "chaos" = emergence and evolution. Prescriptive design = stagnation.

No intentionally-designed system has ever scaled to match the nuance of emergent complexity.

Agreed, simcity uses a static model. So the rules can be deduced. In reality we don't actually know what the rules are and they are much more dynamic -- i.e. the introduction of the internet, the car, the cell phone, etc. These have impacts on cities, and what people expect from them.

Nature (including what humans decide to do) is very good at iterative optimization. When people do something and it works, they tend to continue doing that.

The reason that overly-planned cities (I live in one: Phoenix) don't work is that it's impossible to out-think millions of people all working to optimize your city in parallel all the time.

Obviously you need some urban planning to assist with things that regular citizens don't understand (utilities), but for the most part, if you let people be, they will build a city on their own.

I think your conclusion (only do the minimum planning you absolutely can't live without) is too extreme in the other direction (compare Huston with Portland or Vancouver).

I don't know what the ideal level of urban planing and regulation is, but I think it's probably somewhere in the middle of the road.

what if the subways weren't subways we know of today, but rather conveyor belt like roads. might be feasible if in a more automated future.

This is stuff of science fiction. I remembered Arthur C. Clarke for having written this, but apparently H. G. Wells was first: http://en.wikipedia.org/wiki/Moving_walkway#Science_fiction

Jules Verne did it before that, in the Floating Island if I remember correctly. Which is a parody of 19 century USA, by the way.

I recommend reading Wells' "When the Sleeper Wakes" - the description of "future" technology is fascinating (particularly the moving walkways)

Heck, it's feasible now, but the power consumption and maintenance costs would be fairly insane compared to the plain old cars-and-roads solution. (How often do you see an escalator broken down? Pretty damn often.)

You see more stopped escalators because the escalator is the equivalent of combining the train and the track into the same thing. Tracks don't require the level of maintenance that trains do. Trains can be taken off of the tracks for maintenance, so you don't see the entire system break down when maintenance is done.

I use the rail system in Portland for transportation, and I can say that they fairly regularly have issues with trains on the tracks, or taking down sections of track for maintenance over a weekend.

The vast majority of the stopped escalators I've seen are because some idiot thought it'd be fun to hit the stop button.

How were you able to verify this?

Perhaps he's the one pushing the button.

...as long as it's not built by the same people who implemented it at DFW. That place is hell to navigate.

I would say almost anything: There are hundreds of libraries build together to improve certain life-values (education levels?) in those maps.

The other day I was studying 'dashboard' systems for ERP software and the like and it occurred to me how SimCity and Civilization type UIs would fit the bill (if only they worked on reality).

Remeber the old Caesar series from Impressions/Sierra?

It was ostensibly a city-building sim set in ancient Rome, but gameplay was more like a supply-chain logistics simulator. About 3/4 of the time spent playing was in allocating raw materials to production facilities and designing distribution networks for finished goods.

If Impressions or Maxis decided to start writing enterprise software, they'd probably dominate the market within a couple of years.

> If Impressions or Maxis decided to start writing enterprise software, they'd probably dominate the market within a couple of years.

Yes, if the enterprise software makers would compete on their software. They seem to compete more on having better sales teams.

impressive, fascinating, yet a tad bit disturbing... kudos none-the-less

To be fair, you have the 6th highest karma on a discussion site on the Internet that, on a good day, 0.0001% of the world's population looks at. Not criticizing.. jus' sayin', is all ;-)

I know, I know, I'm only envious :-)

from commets to TFA:

To everyone belittling his 'waste' of 5 years of his life on a pointless endeavour, remember that in life, everything is essentially pointless, you are born, you live, you die. And thats that.

You may in life build the perfect Simcity city, or refurbished a muscle car, or discover electricity. But at the end of the day you end up dead, and its all irrelevant.

I'm sure that people like Martin Luther King Jr and others who have made a difference would beg to differ.

Your life doesn't have to be pointless.

Your two sentences are inconsistent.

The assertion was that life or existence itself is pointless, nothing to do with the esteem of individuals. You say "people like Martin Luther King Jr" would beg to differ about gtt's assertion, but you follow-up by saying "your life doesn't have to be pointless." So, a life can be pointless? So you disagree with MLK? Either that, or your entire comment doesn't really have anything to do with gtt's.

Regardless, the worth of an individual wasn't really what gtt was getting at.

So, SimCity guy felt compelled to be awesome at SimCity, besides that I don't know much else about him. But there are billions of people in the world, so there is room for someone like him and there is almost no reason to stifle diversity in personalities. After all, that's how MLK eventually came to be. Perhaps the same force that compels SimCity guy to build awesome Megacities is compelling him to do something else that most people would consider more worthwhile.

It's no worse than the three years of TV that millions of people have spent their time on for the last 3 years.

I've actually found play to be indispensable at being the best you can be at work. Because the end result really doesn't matter, play lets you experiment more than you would otherwise. This experimentation allows you to discover possibilities that you otherwise would have been too risky. Normally this applies to playing around in the same ___domain as your work (I make stupid/playful websites to practice my web programming skills, for example.)

This is just a video game, but there are still a lot of transferable skills that are practiced: concentration, planning, forethought, logical thinking, multi-variable problem solving, persistence, etc. These habits would serve him well on a programming job, for instance.

Actually given the stated time span, I think this would have been an occasional side project.

Though it might seem like it that isn't the case - people here are generally helpful and down to earth. The problem is that there's an ingrained fear here that the site will turn into reddit[1] with all of the oneliners, memes and silliness that defines it nowadays. HN is the last resort for intelligent discourse for many users. And they don't want to lose it.

As a result of this fear anything that resembles a comment that might have been made on Reddit will get downvoted.

[1] I have nothing against Reddit, on the contrary I personally write loads of silly comments there and enjoy the oneliners.

It's not fear that motivates me. It's simply desire.

I desire to read thoughtful, insightful commentary, and I don't want to get bogged down reading boring garbage. Therefore, I upvote comments that I find thoughtful or isnightful (even if I disagree with them) and I downvote anything I think detracts from that.

I don't have anything against snarky one-liners, memes, or silliness per se. They just have to be remarkably clever or insightful in order to meet my standards. The "nerds have no sex life" comment that triggered this subthread was a boring retread of a tired joke, and the whining about downvotes that followed it was even more boring.

Don't think you can't be silly or snarky or post memes on HN. Just understand, the bar is set a lot higher here -- your silly/snarky/meme-filled comment needs to provide a lot more value than on other sites.

Given the subject matter, I wouldn't pop this in the intelligent discourse pigeon hole, however your point is well taken.

I am not sure if it is my background in planning or my enjoyment of beating puzzles, but I think this subject matter is exactly on par with this site. Watching the videos, you realize someone perfectly reversed the algorithm that makes Sim City 3000 run. Through trial and error this guy beat the system. I think it takes a lot of intuition to get this completed.

Granted, from a purely logical viewpoint this city wouldn't work, but that's not the challenge. The challenge was to beat Sim City 3000 and he did that. Think of it as trying to get the highest score in Pac Man or Super Mario Brothers. There are competitions on that! And from one screenshot from the videos, it seems as though there are competitions for Sim City 3000 too.

Yes this city wouldn't work in real life. The subway system would be prohibitably expensive; getting all the power from neighboring cities wouldn't work; having groups of buildings to satisfy a citywide requirement wouldn't work. Yet he beat the game. That is all.

Funny jokes seem to get upvoted (see the first comment here http://news.ycombinator.com/item?id=1277622)

But the funny jokes that get up-voted are generally intelligent, too (I think).

At least we can say the bar is higher than reddit where half the stories are dominated by a big 50-level-deep thread of stupid puns.

I think that one was skewed by fanboyism.

No, not intelligent, or snobby. Big-souled.

Like Douglas Hofstadter said: "Small-souled men, beware!"

I (and a lot of us here) have no patience with constrictive thinking. In this case, why was sex brought in? Trying to artificially decrease a person's worth is petty.

I think you got me wrong, I'm certainly against criticizing someone's neat simcity video with a jab about getting laid (frankly, such pent up frustration might say more about the accuser).

But I also see a fair bit of "oh my beautiful mind is too good to read this pointless article" on occasion -- That's simply a perhaps-better-spelled version of the same small mindedness.

"oh my beautiful mind is too good to read this pointless article"

Perhaps it's just that time and attention are limited...

If that's the case, said genius should quit commenting on message boards and get back to work :)

I want to procrastinate in style.

Nah, to max out SimCity he probably only has to know about a few dozen variables and how they interact, to solve world hunger he probably has to know how thousands of variables interact with each other. Maxing out SimCity is one of the easier problems out there :)

But he could made one step, say reducing number of varibles involved from 1000000 to 999999. And someone else could reduce this number too...

Maybe he did that reduction by stepping out of the way?