Why the creator of Dwarf Fortress is really excited about boats | Hacker News Reader

Why the creator of Dwarf Fortress is really excited about boats

pcgamer.com

67 pointsrhlala12 comments

I never really paid attention to individual dwarves when playing, so I had no idea they had so many personality traits: http://dwarffortresswiki.org/index.php/DF2014:Personality_tr...

I wonder if someone's written an online quiz to give you a code to insert at the bottom of your emails, like those old hacker code one-liners people used to put into their sigs. (Can't for the life of me remember what they're called now, and searching for "code email signature" is about as useful as you'd expect.)

You're thinking of "Geek Code": https://en.wikipedia.org/wiki/Geek_Code

That's the one!

Oh... When I saw that in passing I always thought that it was short hand for the psychology discipline - you learn something everyday.

Do you mean https://en.wikipedia.org/wiki/Just_another_Perl_hacker ?

A bit off topic, but regarding DF: If I wanted to make a much more simpler game where some "factions" would iterate basic geopolitics among themselves what kind of resource should I be looking for if I was aiming at basic consitency? I mean probably some AI book, but other than that or maybe exactly that, there is some blog or article that would exemplify the basics? Perhaps a paper? DF is full of simulations, super complex, but I wonder what is the simplest implementation to get surprised by the "emergence" of the game itself and not pre-established design.

You don't need AI. The starting simulations and agents don't need to be that complex. You make a toy first, and then see if it's of use. (I think Europa universalis and other games from that house may suit you?)

Matter of fact, if you have time, use the bay12 Way.

Write a story of what you'd like to see. Then break it down into interactions, atomic items, compound items, and then go from there.

Do note, that DF allows for hugely complex interactions and emergence partly because it has the advantage of a millions year of development time, unburdened by having to conform to market or business expectations.

So this allows for moderately deep systems to interact with shallow systems and yet more systems which are very deep.

The sheer number of systems results in all sorts of emergent behavior, without having to resort to depth of simulation.

So for example, you don't need a z axis when you start out. You may well want geography though, since that determines ecology and weather.

These inform resources, nutrition, growth, abundance and derive from there to economy.

Assuming you don't simulate evolution, and hard code races/civs - you have the basics to define conflict.

But unless you write a story down, it results in an antiseptic sim. A story or text helps you identify quirks of your style, which you want to see reflected in your sim (you may want people to always have a predilection for ponies, or prefer cats over dogs. You may hardcore pacifism as a civ trait which leads to other results )

This keeps some sort of emotional and narrative hook, which makes the simulation worth engaging with.

Give this video a shot - Simon Swain builds something like what you're looking for, but as an interstellar war simulation, but with a little fake economy underneath: https://youtu.be/0HJPilemNns?t=255

And the actual interstellar war bit is here: https://youtu.be/0HJPilemNns?t=1416

If you watch the whole video, you'll see that you can actually get a lot of complexity out of some really, really basic underlying rules (and a nifty visualization to boot!)

There is a DF-styled game called RimWorld currently in development. The creator is fairly active on the game's subreddit and has blogged frequently about development.

Not an exact target to what you're seeking, but hopefully it'll be useful to you.

It's like painting. You don't start with cats and flowers. You start with circles and straight lines. Don't think about how to make the game of your dreams, but consider if this is worth spending a half-day-job amount of money for the next 5 years on it. If not you can just stop there. If it is, start with something simple like tic-tac-toe.

given the complexity even quite simple cellular models can show (e.g. https://en.wikipedia.org/wiki/Wa-Tor for something a bit more "realistic" than the classics like Game of Life), something like that could be a starting point: generate a field with cells with neighbors, define some basic properties per cell, have a set of rules/policies to execute under certain circumstances.

I'd be super happy if Dwarf Fortress managed to make canal locks navigable by the AI.

Reply on news.ycombinator.com