How you even begin to introduce an interview on Dwarf Fortress is a complete mystery to me, as will become quite apparent in the next few seconds. It’s a sprawling roguelike fortress-building dwarf simulator, which you’ve probably heard of even if you’ve never played. If you have played it, then you’ve probably either lost several hundreds of hours of your life, or you stopped playing in the first five minutes. This is a game made by just two people, it’s been in development for 13 years, and it’s scheduled to continue that way for another two decades, with irony nowhere to be seen.
So rather than try to explain this intricately simulated super-complex world generation adventure/construction/exploration game, I thought I’d let the man himself have a crack – game creator Tarn ‘Toady’ Adams.
PC Invasion: How would you describe Dwarf Fortress to someone who’s never played it before?
Toady: It’s basically a city building game that has aspirations towards being a sort of fantasy world generator. It wants to simulate everything about the world and you just kind of dive into one given world that persists over time and you take part in a small story there and bring in another small fortress or a small character and do it again. The things that you did before will impact the future and so forth.
I mean, we’re only part way through it so the extent to which we’ve succeeded varies wildly depending on the exact mechanism you’re talking about.
PCI: One of the defining features of Dwarf Fortress is the sheer detail of the world that’s generated. How did you go about putting that together?
Toady: It’s been a work in progress for 20 years in various forms. We used to play Starflight, which had a fractally generated planet map with randomly generated life forms on it, and the old roguelikes with generated dungeon maps with all the items and creatures placed randomly, so it was a paradigm that me and my brother were already used to and we were like, ‘Why isn’t more done with that?’
We wanted randomly generated maps like the Starflight maps but you’re wandering around like an adventurer, so we started writing a game which lasted five or six years, that was called ‘Dragslay‘ because we had 8 letters back then to say dragon slayer. I think I started the first randomly generated map when I was 13 or something. I made this kind of altitude map with forests on it and some towns and you’d go around adventuring in the dungeons and these groups of goblin type creatures would come out of the dungeons and sack one of the towns, and it was all randomly placed. If they killed your adventurer they’d put your skull outside their cave for the next time, so when you visited it there’d be a stack of skulls there and they’d yell the name of your previous adventurer. So a lot of this stuff you see in Dwarf Fortress here and there we were doing from then.
Then we had ‘Slaves to Armok: God of Blood‘, which we released on the web in 2000. Not completed, of course, *very* not completed. It had 3D models and you could walk on your hands or one hand and one foot, chop off different body parts etc.
Meanwhile, we were working on these side projects, and we started one that was going to be Dwarf Fortress. You make a little colony that you mine out, more or less like Dig Dug, but you have just a few dwarfs wandering about working. It would die somehow because the game would just get harder and harder, and then you’d bring an adventurer to go in and explore it and find your production logs and diaries and just get a high score, it would post your high score, and that was going to be our, almost, little arcade game, right?
The main thing was that it was in text graphics, so it was really fast to add features to it compared to the 3D monstrosity we were working on, and it started to steal features, future features and current features, from the Slaves to Armok game. Since we had like 100 people on the website at the time we said, ‘This is Slaves to Armok… part 2! Here it is! And it’s called, “Dwarf Fortress“!’
PCI: There’s so much going on with the game now, how do you decide what to work on next?
Toady: We have these aspirational goals. We might want to have something like boats, or magic, or a criminal justice system, or something cool to go in the game, but there’s an order that things need to go in. You need to understand property before you really get into crime deeply. There’s also a cart before the horse type of thing, do you want trade before boats or do you want boats before trade? We also think about what you can make use of now to keep the game fun. We always try and think about the experience of what the game is like, at least, mechanically. We think about how the mechanics will play out and how the exposition works, which is a really complicated matter in the current things I’m working on with poetry and music and so forth.
Like if you randomly generate a musical instrument, do you give it a name? Or do you give a description like, ‘twelve string long-necked lute.’ Then every time you pick up a twelve string long-necked lute you have to call it a twelve string long-necked lute. Or should it just be using a Dwarven word like urist, from the translation file. But then if you just say like, ‘I see an urist’, how do you deal with that problem? You have to have little tooltips popup or whatever, the kind of things that would be uncharacteristic of Dwarf Fortress. (laughs)
The choice as to what to work on is mainly pointing at some goal and trying to have some fun along the way. For instance our current list is the Dwarven taverns and inns, which just sounded fun, right? Being able to start a tavern and inn. Now, what makes up a tavern and inn? Well we want bar fights, we want gambling, and what makes a dwarf go to a tavern and so on? We need to think about needs, so we’re adding little temples and taverns at the same time. People have spiritual needs but they also need to socialise, some people just need to live a debauched lifestyle or whatever. Those needs are going to be stacked on top of all the other needs they have, but hopefully not in such an overwhelming fashion that you can’t play the game anymore.
After that, we’re going into artifacts and magic after that and then into the framework of society and stuff so that we can do fortress starting options and so we understand laws and religion a little bit better and then you can embark as a pilgrimage or something like that.
Then we’ll have to decide, do we want to do criminal justice, or boats, or the economy? It’s not decided what direction things are going to go in. As long as the game is overall playable-ish and roughly balanced between different parts of the game, it’s fine. There’s a big difference between an unplayable alpha and something like this.
PCI: Do you have a general idea of what you’re working on even years ahead?
Toady: Yeah, we had a list, it started almost a decade ago (laughs), it was of 100 things that would make version one of the game. We count up one centi-version number for each one that we finish and we’ll get to 1.00 and that’ll be it.
That lasted for a while… (laughs) we got maybe to 31 that way.. then it changed a bit. Some really cost more than others.
So we broke it up into a system now, I think it’s got about 2700 points and I’ve just got this giant file. Every couple of years I write a new program, or have a new giant piece of paper on a desk, or a flashcard system; partly just to keep it fresh, I guess. We check things off, that reports some points back and increases the version number – it’s all systematised – and there’s not a lot of wiggle room left in that system but we’re pretty satisfied with what’s there. The main thing is to say what’s in and what’s out and how far something needs to go before it’s considered version one-ish.
We’re not keen on having playable monsters or playing other races for now. Modders do that just fine. We can just wait before we concern ourselves too much with that, we’ll just continue with what we’ve got. What’s in for this time, and probably for the next release is having some permanent additions to your fortress. Like if a human bard comes in and impresses everyone in the tavern as a visitor then you might hire them on and they stay in your fortress.
There are some things to come to terms with there, because right now that causes a bug where they’re considered kind of half pet half visitor because it used to consider everyone who wasn’t a dwarf that belonged to the fort permanently as an animal, because that’s how it was. I remember one of the world gen civilizations ended up with like an eleven queen for the dwarves, because of some population movement or whatever, but when that queen finally ended up at the fortress they were treated like an animal, and could be put in a cage or something (laughs), I can’t quite remember the details.
Some things still fluctuate as to how far we’re going to go. We don’t know quite how far we’re going to go with magic systems. It’s a big point, we don’t really have a lot of it in Dwarf Fortress, but it’s also, for a fantasy setting, one of the crucial points.
PCI: That’s actually going to be the dwarves playing around with magic, is it?
Toady: Yeah, that’s the main part that the player will see. There’ll probably be a lot of tangential stuff that comes out of that.
It’s a tricky one, right? Because you don’t want to have people running around shooting D&D fireballs at each other necessarily. But there’s also popular offshoots involving runes and that kind of thing, do we get into that? That’s the kind of thing that’s fun to do spontaneously. Are we going to suddenly jump in and do an alchemy system? Or some kind of magic scrolls, books or whatever system and throw that in there? Who knows? It doesn’t matter, as long as that’s something interesting to the game.
PCI: The UI for Dwarf Fortress is certainly not the strongest part of the game, is that something you’re going to work on?
Toady: Part of it is thinking about what the point is. How many people are we trying to serve? In what way? Is it better to be a niche thing for a thousand people? Or is it better to serve a million with something that’s going to be shallower because you have to work on getting appearances up? There’s a balance there that we can’t say we’ve found. We’re not professional business type people that do professional business type things and understand this stuff.
We do read when there’s an easy suggestion for something. Currently someone has mocked up a whole way just to reorganise the hotkeys on the main page. (currently in the game) One of the middle keys is ‘d’ for ‘designations’, but that’s actually how you make your *whole* fortress by mining… You REALLY want to give that a better name and put it at the top (laughs).
PCI: For new people, the game has a very steep learning curve. Right from the start with the embark players are flooded with things to think about, what skills to give their dwarves and what to take and so on. Do you see that as part of the game?
Toady: Not necessarily. Something we’re working on is embark scenarios, for example. Would you like to be a military outpost? We can pick a dangerous spot for you here, here’s five spots, what do you think? These are the different traits. This one is near a goblin fortress, this one is in a dangerous undead forest, and these are the advantages and disadvantages. While at the same time, people who are used to the giant lists (which we currently have) will just have more lists and more choices and they’ll be able to choose what religion would they like to have for their pilgrimage, and what specific goal do they have, if any…
PCI: Dwarf Fortress has quite a devoted community and you work pretty closely with them. How do you find that?
Toady: They’re…. they’re cool. (laughs)
They’re our playtesters, right? We have community bug tracker managers – people working very hard because there are a lot of bugs and it’s a thankless job – and someone ported the game to Linux and iOS for us. We also have questions every month – we’ve kind of gone into a rhythm where I answer anywhere from 20 to 80 questions every months, and I read the suggestion forum. There’s tens of thousands of threads on there now and I make sure to read the opening post of every single one, and the discussion if it’s something that needs a discussion. There’s 13,000 topics and I’ve read every one. Most suggestions can’t go in, obviously. They’re good, but, you know, we already have a plate that’s full. There’s only so much you can do… but it’s all cool and they give us a lot of feedback. It’s great having a group of people coming along for the ride.
PCI: The game actually really inspires people creatively, with stories and even artwork. Did you see that coming or was it a complete surprise?
Toady: We knew that we’d tell stories about games and so forth, that’s how we ended up writing some for the design for the games was to write stories at each other and then analyse the sentences of the stories and work out what the elements we needed to make these stories happen, but we had no idea what was going to happen when Dwarf Fortress made it just a little big bigger and we had people doing these giant movies and animations and stories of these succession forts they’d passed around save files and so forth. All of that was community generated stuff that we didn’t expect at all! And now it’s got to the point where we start expecting it.
A couple of days ago I just posted some output from the poetic form generator – ‘here’s 30 or 40 types of poetry that give a description of how you should do the metre and what the subject matter is’ etc. Within a couple of hours there were already example poems being written by people.
When that version is actually released and we have example poetic and musical and dance forums and so forth, there are going to be people integrating that into their actual game stories, if they’re so inclined, and it’s going to be cool. The thing I don’t know, and this is something maybe people will wager on, is are we going to see people re-enacting a dwarven dance description? That’ll be a new one! I have no idea what’s going on there… I know there’s a lot of musically talented people who’ll be able to knock off the music in zero time flat, but the dance I have no idea.
Part of that is going to be on me because I don’t know very much about dancing so I’m just going to read and see what I can cook up. I couldn’t find a common terminology or anything like I could with music and poetry, so it’s going to be difficult to fire up people’s imaginations the same way. I’m going to give it a shot, but it’s a tricky problem. Things are very specific. You have positions and moves and ballet, but there’s those intertwining circle dances with, ‘simple footwork’ and ‘graceful movements of the arms’ and so forth. It’s tricky, we’ll see what happens.
This whole thing is mostly just so dwarves can sing drinking songs. The poetry is going to link up with the dancing and the music and it’ll be able to describe what your dwarf is going when he’s moving left and to the right in the tavern and you’ll be like, ‘is this dwarf’s path-finding broken?’ and it’ll say he’s ‘dancing to the urist’ or whatever and you’ll say, ‘what’s that?’ and it’ll pop-up the five paragraphs that you always never knew you wanted about dwarven dance theory.
PCI: In the past you’ve said that version 1.0 of the game is about two decades off. Someone might be inclined to say that means the game will forever be in alpha. Is that a possibility?
Toady: We use the term ‘alpha’ in popup when you play the game, since it is kind of a rough game right and we want people to know that, and since it was free you shouldn’t complain about it! It’s cool… play the game… if you want. But perpetual alpha, that phrase means something else, right? That means you’re vapourware and you’re overpromising and you’re a bullshit artist. It can mean those things, but I would reject that idea.
We’ve been working on it for 13 years now, we’re 40% through with the goal list as of this latest release. 20 years still seems like a fair figure for 1.0. But, you’ve got the whole 90/10 rule, right? So maybe we should be saying 200 years instead… that’s probably to get it out of alpha. The last little push there for the 180 years of interface and graphics design for it or whatever. We’ll have some sort of technological singularity to help us out by then I suppose…
PCI: Is motivation at all an issue or are you still looking forward to getting in there and putting the next thing in the game?
Toady: Yeah! Absolutely. How could you not be? Like this poetry stuff, that’s been cool. Can I put this thing together? What’s the paragraph going to look like? I had no idea, everything is a surprise to me. This is why me and my brother were writing these procedural generation games from early on is because if you’re going to make games yourself you’re going to want to play them yourself, you want to be able to play them yourself so you can keep playing videogames. That’s how this one is going, we try to make things that surprise us. The output has been cool and I’m looking forward to see how it mixes in with the fortress.
Every next step is exciting, I can’t wait to get a boat moving around – big old boat, taking a bunch of tiles, multi-z-levels, a big old mast sticking up, people jumping around off the mast… and boy! Whales as one W are going to look small, do we have to look at multiple tile creatures again? All that kind of stuff.
If you have enough fortitude of character to want to give Dwarf Fortress a shot or you’re a returning veteran wanting to take a look at the new additions, what you probably want to download is something called PeridexisErrant’s Dwarf Fortress Starter Pack. This is the game with all the obligatory graphics packs and UI assistance tools. Watching a tutorial or twenty on YouTube probably wouldn’t go amiss either…