How Two People Created Gaming’s Most Complex Simulation System

Dwarf Fortress already has more than two decades of development history and if it’s up to its two creators, Tarn and Zach Adams, they’ll continue working on their management simulation game for the rest of their lives. To get a clear understanding of Dwarf Fortress’ origins, we have to go all the way back to the eighties, back to the brothers’ childhood. Born in Kitsap County, Washington, Tarn aka Toady One and Zach aka ThreeToe grew up around computers from a very young age. Their father, Dan, worked in wastewater treatment, writing software to crunch data and run sewage plants. He supplied the family home with the latest computers and even taught his sons the rudimentary basics of coding, knowing programming and computers would become even more important in the future. “I was set in front of computers by my parents when I didn’t even know what was going on. I remember some kind of ASCII demon or alien dancing, and they set up the dance moves by pressing the keys. We also had a Vic 20 and an Atari.”

When Tarn was six, his father taught him how to use a ‘FOR loop’ in BASIC, to make something move across the screen. The brothers eventually started playing a game called “The Temple of Loth” and they noticed it offered full access to the source code. Seeing exactly how the game was made, it inspired them to develop their very first game. This ultimately led to them developing about 400 projects in BASIC, ranging from Tarn’s ASCII monster designer tool to a John Woo-inspired text-based assassin game. From a young age, they fell in love with Dungeons & Dragons and J.R.R. Tolkien. The two also downloaded many indie roguelike games from early bulletin boards, which were basically download servers that predated the internet. One of those roguelikes was Hack and dates back to 1984. Hack, along with Rogue, Larn and Ragnarok, all games the brothers also played, helped to popularize roguelikes at the time, and in turn greatly influenced Dwarf Fortress.

“We liked that you could choke to death on your food or fall down a stairway and fall on something poisonous you were carrying and poison yourself. That lineage was big, largely inspirational for Dwarf Fortress, especially the bones file part where you die and then you have in your subsequent playthrough, something about your previous character, more than a score or whatever, incorporated in the game. You can trace that directly to Dwarf Fortress persistent worlds. That’s where that comes from, this idea that you can kind of churn this system and build a history into it.”

Around the time Tarn and Zach were 12 and 14 respectively, they started developing their very first fantasy game together called Dragslay. It was also written in BASIC and Tarn described it as basically a D&D game where the player runs around and kills dragons and other creatures. Curious details and game variables already started to emerge here, like your character having an IQ score that got reduced if you were hit too many times in the head. It was meant to be just another project the brothers would soon forget about. However, a couple of years later (1993), when Tarn was in high school and taught himself how to code in C, Dragslay was revisited. Programming the game in C this time meant he and Zach could make it bigger and add things like a world map, a more complex combat system and keeping track of goblin populations. They kept working on it for about five years, but it would eventually evolve and play an essential part in the creation of Dwarf Fortress.

Programming has always been Tarn’s escape. He wasn’t the type of kid that would play sports or talk to others, instead he was very withdrawn and only occasionally had a friend to joke around with. Even then, he didn’t care about socializing and would spend almost all of his time behind a computer. There’s one close friend he made in high school that Tarn still sporadically talks to today. During their childhood, the brothers had to move around a lot because of their dad’s work, but eventually found their way back to Kitsap County by the time Tarn and Zach were 18 and 20. This experience brought them closer together and Tarn feels lucky to call Zach his best friend.

At this point in his life, Tarn didn’t see a monetary future in his and Zach’s video game projects. Not being interested in Computer Science as an academic pursuit, he picked math instead and went to the University of Washington. Since Zach was already studying Ancient History in the same university, they decided to share a small apartment in Seattle. The summer before Tarn went to grad school, he and his brother rebooted Dragslay. They changed the title to Slaves to Armok: God of Blood, named after the god from Dragslay, who was originally named after a variable called ‘Arm_ok’ that counts the number of arms you have left. For now, it remained a 2D project in a somewhat-isometric view and featured roughly the same combat system as the first public alpha version of Dwarf Fortress, as well as persistent enemy groups, meaning they could show up in your next playthrough.

This version of the fantasy game was very short-lived however as Tarn and Zach started having more ambitious ideas. They kept the title, but made the switch to 3D graphics. Over the course of development the brothers added many specific details and options to the 3D character models, like the possibility to remove the enemy’s skin with certain spells. They even went as far as to develop an option where the player could zoom in and inspect how curly the character’s leg hairs were, or discover the melting point of various materials. This version of Armok is the first piece of the Dwarf Fortress puzzle. In December 2000, the brothers created a website for their indie studio Bay 12 Games and used it to distribute Armok as freeware. Additionally, the site also lets you access the official forum where fans can connect with each other.

Even though Armok was their main focus for the next few years, Tarn occasionally took time off to work on side projects whenever he had a good idea. They were usually small in scale, needing only a couple of days of work to be completed. The most popular examples are probably Liberal Crime Squad and WWI Medic. In 2002, Tarn began working on a side project called Mutant Miner and this is the second piece of the Dwarf Fortress puzzle. It featured turn-based gameplay and was loosely inspired by Miner VGA and Dig Dug. Players dig underneath buildings to search for minerals and fight monsters. The goal was to discover radioactive goop, bring it back to one of the buildings at the surface and use the goop to grow extra arms and other mutations to speed up digging and fight off enemies. 

Tarn eventually tried adding extra miners, but since it was turn-based, the game started to experience lots of lag. This gave him the idea to switch to real-time gameplay to prevent all the slowdown and swap the one mutant miner for a group of dwarves. Furthermore, he came up with a more interesting gameplay loop. It still involved managing a mining operation, or a fortress if you will, but once the player “loses”, an adventurer comes in and is able to explore the ruins of your fortress. This section of the game played out like a more traditional roguelike where the player has to collect goblets, journals and other trinkets that were generated during the previous phase and bring the treasure back to the edge of the map as a way of building your high score. He called up Zach, who moved back in with his parents after graduating, to flesh out the idea over the next three days. It was this idea that would pay off big time 20 years later.

The brothers started development in October 2002 and estimated it would take about two months of work, just in time to release it during the Christmas holidays. Little did they know it was only just the start of perhaps one of the longest development journeys in video game history. Thanks to the simple ASCII graphics, they were able to implement features at a very fast rate, to the point it was starting to steal too much time away from their main project, Slaves to Armok. Therefore, development was temporarily halted in November so they could focus on Armok again and various other projects.

Fast forward to 2004 when Tarn and Zach’s enthusiasm for Armok was all but gone. The 3D graphics were slowing them down too much, causing more and more new features constantly having to be delayed. So, the brothers made the decision to fully switch gears to the dwarf game. Since early prototypes worked on much of Armok’s code, Zach and Tarn considered it to be a sequel to the 3D game. For kicks, it was even officially named ‘Slaves to Armok: God of Blood II: Dwarf Fortress’. After this change of plans was made public on the forum, along with a trailer and screenshots of Dwarf Fortress, Bay 12’s community started to grow more exponentially. People were intrigued by all the possibilities, especially knowing that the brothers wouldn’t be bogged down by 3D graphics anymore.

Dwarf Fortress’ landscapes are made entirely using the character set CP437, which was also used for the original IBM PC. It’s more commonly referred to as ASCII, a character encoding standard consisting of 128 characters, or code points if you will. As a result, trees look like clubs or spades, blue tilde symbols display water, and dwarves move through your world as smiley faces. While many players prefer this almost matrix looking aesthetic, most fans use mods that replace the character set with actual graphics, allowing for a more accessible gaming experience. Fans often used to request updates like more modern in-game systems, in-game tutorials, 3D graphics, a less confusing keyboard shortcut system, and better mouse support. Aside from the necessary and quick bug fixes, Tarn had stated that these type of suggestions would not be his focus.

“It felt like adding graphics would ruin our ability to work on the game, given our failure in the earlier 3-D attempt with Slaves to Armok: God of Blood. The DF modders have done a good job adding graphics though, and it hasn’t impacted development negatively, but that’s a bit different from doing it in-house. 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?” Instead of giving in to the many fan requests, the brothers followed their gut feeling and solely focused on adding and expanding features.

Tarn and Zach made for the perfect team during Dwarf Fortress’ development. While Tarn was programming away, Zach worked on the narrative and descriptive details that give Dwarf Fortress its atmosphere, informed by his background in ancient history. For example, goblins hang the skin of their conquered enemies from towers, inspired by the writings on Assyrian kings that Zach recommended. When Tarn runs into issues, the brothers typically hash them out together.

Dwarf Fortress was making good progress and Tarn was perhaps making even better progress in school. When he graduated he was named best math major. Afterwards, he applied to 17 Ph.D. programs, got into 15 and eventually chose one at Stanford University. In 2005, Tarn earned a doctorate in geometric measure theory and this landed him a postdoctoral research position at Texas A&M University. It’s a career path that most people can only dream of, yet Tarn would quickly discover that the future he envisioned for himself since his undergraduate days, wasn’t what he really wanted. Juggling between his postdoc research work and Dwarf Fortress proved much more difficult than he imagined. Combined with the competitive nature and high pressure of academia, Tarn’s mental health began to suffer and he eventually grew depressed.

In order to do his postdoc work properly, he felt he had to give up his game development hobby. This however was unimaginable to Tarn and right then and there he decided it was either going to be making games for a living or doing nothing. He never loved math, but he loved video games, that much he knew. Tarn describes it as a very stark moment in his life. In the summer of 2006, after a year of working at Texas A&M, Tarn told his department head he wanted to resign. After an emotional conversation, the university offered to relieve him from his postdoc work and give him a teaching position for a single year instead. This new position would earn him a salary of 50.000 dollars. Hearing this was a huge relief and Tarn gladly accepted the offer. The extra money would give him a fighting chance of making things work as an indie developer.

In the meantime the brothers continued working on their passion project as often as possible. At this point, Dwarf Fortress had evolved far beyond its initial concept and was starting to resemble how people know and play it today, with a few key differences. The entire map was still randomly generated, but it was fully 2D, meaning there was no way to dig underground or climb walls or buildings. Instead, at the start of each new game, players were given a wagon with seven dwarfs and had to explore from the left to the right of the screen. You would always encounter a river, then a chasm, followed by magma. Along the way, players were still able to build and craft items while dealing with certain dangerous encounters. There was even a sort of endgame present in the form of discovering and mining adamantine ore at the far right of the map. Once you started mining adamantine however, there was always a chance your fortress would be lost soon and the player would receive a game over screen.

The decision to choose dwarves as the games’ main race was an easy one, as it seemed like an obvious choice to Tarn and Zach for a fantasy game about mining. Initially, their personalities consisted of pretty basic elements like food preferences, skills, relationships, current happiness, total stress and so on. However, as development progressed, the brothers wanted each dwarf to have their own identity, much like a character in a book for example. Therefore, more specific attributes were thrown into the mix such as imagination, artistic interests, curiosity, depression, trust, achievement-striving, just to name a few. In total there’s about 50 personality facets, 30 intellectual values, a bunch of specific needs and 120 different emotions. These are then divided into four categories: “creature”, “cultural”,” family” and “during play” and all of them combined make up the personality of a single dwarf. This translates to much more interesting gameplay moments where dwarves can hold grudges, are shy, take in pets and go crazy if a goblin kills those pets, become fond of specific items and even hoard them. Of course, this is all determined before you even start playing and Tarn also wanted to make sure dwarves can evolve and change if a certain impactful event happened to them.

Believe it or not, the dwarves’ physical attributes are perhaps just as detailed as their personalities. Dwarf Fortress simulates muscles, blood vessels, nerves and more. Because of this, combat can get pretty violent and instead of a traditional HP bar, dwarves can lose limbs and even teeth, something that’s directly inspired by the tabletop RPG Cyberpunk 2020. This and other detailed combat systems were also partially inspired by history books like The Iliad. The two developers have also stated how games like Starflight, Ultima, Seven Cities of Gold, Moria and many more roguelikes inspired them while creating Dwarf Fortress.

Tarn expected that after his year at Texas A&M was over, he was going to have to live on his savings and eventually be forced to get a side job in order to support himself. However, on August 8th 2006, the game’s first alpha version was released. A few months earlier, on April 17th 2006 to be exact, a forum member wanted to donate $75 for Tarn’s birthday and asked if the website could get a Paypal button. Once it was up, the brothers, for the first time ever, earned money from one of their games. Later on, they also implemented a subscriber system to make it easier for fans to donate a monthly sum. In the next five months, they made around 300 dollars, which was enough to cover the Bay 12 website’s 20 dollar hosting costs.

Once the alpha version was available to download, the project quickly made its way on different websites and blogs, causing word to spread. Unsurprisingly, donations kept increasing over the course of the next year. In 2007, after his last year at Texas A&M, Tarn finally started working full-time on Dwarf Fortress, a dream come true for him. Though he admitted it wasn’t always easy. “If the laptop broke during a given six-month stretch that would’ve been it for the whole full-time thing and we were just lucky about when our computers broke. I had my bank account skim zero a couple times. I never did the whole, like, mortgage… I don’t own anything to mortgage, first of all, we haven’t gone into debt, but we were at nothing several times.”

While things were tough financially, Tarn was content. Paypal donations had reached a point where he was able to support himself on a small living wage. Most of it was spent on basic living expenses and the rest went to Zach for his help on the game. Thanks to all the contributions from fans, the project was self-sustaining, making it unnecessary to charge for the game. Aside from helping the brothers financially, the growing amount of support was also a clear sign they were on the right track. With a bigger fanbase, it became more important to properly maintain the site’s forum, as it was their only form of marketing and the only way to talk to fans. After all, this was well before the existence of social media. Zach and Tarn put in a lot of effort, every single day, to make sure the forum’s not plagued by trolls and conversations stay on topic. At a certain point, the two developers started sending personalized rewards to fans that donated money. These were either stories written by Zach or a piece of crayon art drawn by Zach and colored by Tarn, adding even more to the personal nature of the brothers’ passion project.

With his new job as full-time indie developer, Tarn left Texas and moved back into his parents’ house before finding his own place closeby in Silverdale. While he was living his dream, his older brother Zach couldn’t say the same. After graduating, Zach got a job at an Amazon warehouse and later at a Naval shipyard. He was dealing with some personal problems so Tarn wanted to be there for him. It was also a good opportunity to collaborate more easily on Dwarf Fortress again. Zach has said that once Dwarf Fortress started to take off, it saved his life.

The brother’s maintain a list of about 2,600 ambitious features they plan to implement for version 1.0. Tarn admitted however that not all of these features might make the cut, since that would probably mean he won’t be able to finish a “completed” version of Dwarf Fortress in his lifetime. Zach and Tarn always have five to ten new features on the horizon to implement next, choosing the ones that make the most sense to add after the previous one, with interconnecting themes and logic helping their decision-making. “If you want to just say, ‘I’m doing Dwarf Fortress political intrigue now,’ but we don’t have notions of law, property, customs, status, economics of any kind, and don’t really understand personal interrelationships and territories and things well enough to do that any justice. So it gets kicked down the road.”

Tarn draws bubble graphs, Venn diagrams, and other ways to think of all the features and their connections, leading to stacks of pages, post-its, and text files filled with ideas. He likes the notion of being able to hold the game in his hands. Working on multiple varied features helps Tarn to maintain focus and stay motivated. In contrast to more regular game development where coders have to perfect their work before release, Tarn and Zach have the freedom to experiment and completely shift focus when necessary. Since they’re always working on multiple features at once, the brothers simply rotate into the next topic whenever they feel like they’re hitting a wall. This gives them time to come up with fresh ideas for the previous topic.

Most of the time, Dwarf Fortress is all Tarn can think about and a normal day in his life goes as follows: he wakes up around 3 p.m. and proceeds to work on Dwarf Fortress until he goes back to bed around 6 a.m. 100-hour work weeks are common for Tarn, though it doesn’t necessarily feels like work to him. “I don’t have, like, a life or whatever. Why would you take time off when you can write Dwarf Fortress? That’s not bad. The 100-hour weeks are not all necessarily on Dwarf Fortress. It’s, like, 80:20 or something, and the 20 is side projects and stuff. Oftentimes the side projects that I spend time on, 3-4 hours in the evening. You sit in front of the TV and not watch it, with a laptop on your lap, writing a myth generator or something, right? That’s kind of how it goes.”

During an interview with New York Times magazine, Tarn admitted he’s also not that interested in having kids. He wants to stay focused on his passion project and he feels kids might be too much of a distraction. Furthermore, as Tarn got older, his interest in finding a romantic partner dwindled to the point he’s just not anticipating it anymore. For many years, his cat Scamps did provide plenty of companionship, but he sadly passed away in June 2022. Tarn immortalized his beloved pet in the game.

As mentioned before, Dwarf Fortress used to be a fully 2D game, only utilizing an x- and y- axis. After a while, Tarn and Zach felt this was too limiting and wanted to add 3D architecture and 3D fluid simulations. Unfortunately, they had no idea how to implement something like that within a world of text graphics and weren’t willing to dive into such a big challenge. At least until they encountered a certain design problem. In short, players are able to build bridges to cross rivers, as well as build flood gates to control water flow. However, when the brothers added aqueducts, things started to get very very messy and complicated because of the lack of 3D architecture. So, after about five months after releasing the initial alpha version, they began experimenting with adding a z-axis to the game.

One thing the brothers immediately discovered is that world generation had to be completely overhauled so that rivers and chasms would now be giant open lakes and caverns, filled with interesting things to find and collect. According to two developers, the main downside to this, is that it now felt like less of a game overall. The clear objective of going from left to right to discover interesting things to expand your fortress with, had now transformed into something where players could dig and advance in any direction, not knowing what they would stumble onto. To Tarn and Zach, the upside of having total freedom to build more intricate structures and having a more interesting world overall with endless possibilities, far outweighed the downsides.

Needless to say, this major update proved challenging to program for Tarn, who called it one of the most mind-numbing things he has ever done. Not only did he now have to take details such as fluid mechanics and cave-ins into account, but he had to go back to all the function calls that relied on the x and y-axis, and see how a z-axis could fit in. This new and improved 3D version of Dwarf Fortress, even though on the surface it visually still looks the same, was publicly released in 2008.

By now, the gameplay loop looked more like this: you generate a world, build a fort, you eventually lose, the fort gets added to the world, you start a new fort or play adventure mode, lose, it gets added to the world again, and so on. Losing is very much part of the experience and Tarn even describes it as the entire engine that runs the game. It’s why Dwarf Fortress’ motto is “losing is fun!”. The important thing to the brothers was that when a player loses, their impact on the world would persist and be visible in the next playthrough. An example would be encountering the ruins of a previous fort or bumping into an old friend or enemy you made during your first time playing adventure mode.

Although the art style is simply made up of numbers, letters and symbols, generating a world requires quite a lot of CPU power. Even modern computers struggle a bit during the process, thanks to Tarn’s extensive programming that results in highly detailed landscapes. He and Zach specified everything to the extreme, from geologically accurate ore deposits to river erosion capacities, climate and economic patterns, narrative and historical texts constructed in real time, systems that model animal migration, the list goes on. These elements also change and evolve as the game progresses, and the player’s fortress and the surrounding world builds up its own history, potentially spanning over hundreds of years. The story of a fortress’s rise and fall isn’t scripted beforehand, but generated in the moment by all the dizzying amount of variables. Sometimes this can make for funny situations and accidents, that even startle Zach and Tarn themselves.

“We didn’t know that carp were going to eat dwarves. But we’d written them as carnivorous and roughly the same size as dwarves, so that just happened, and it was great.” Another fan-favorite mishap was the drunk cat bug. After Tarn had added more poisonous effects when sprucing up venomous creatures, dead cats started piling up in people’s games, especially around places with alcohol. They would start to vomit all over the floor, causing them to ultimately die. He couldn’t figure out what was causing this, and even enlisted help to find out what was going wrong. As it turned out, an old footprint code would cause alcohol to get on cats’ paws from walking on spilled beer or wine. They had the ability to lick themselves clean and after doing so, caused them to ingest the alcohol on their paws. Unfortunately for the cats, one number was off in the so called “ingest-while-cleaning code”, meaning the small creatures would ingest large amounts of alcohol, sending them through all the symptoms of alcohol poisoning, which was added along with all the other new poisonous effects.

In-game incidents like this provide insight into the inner workings of the brothers’ captivating brains, that fuel the mind-boggling amount of detail and variables in Dwarf Fortress. All these elements construct one of the most complex simulation systems in modern games, making it just as much a story generator as a fortress management game, just like Tarn and Zach envisioned it. Fun fact: while the brothers mostly use the internet to fact check things like the weight of gold, there was one time they couldn’t find the density of saguaro cactus wood. To help them out, a Dwarf Fortress fan ordered the cactus wood, determined the exact density using liquid displacement tests and informed the indie developers about the results. So not only is it now in the game, it’s also easily searchable on the internet and all because of Dwarf Fortress and its community. 

The brothers grew up playing computer games with notebooks in hand and logging their journeys in detail, as well as competing with each other to get the highest score. Therefore, they aimed to create an experience that would log your adventure in captivating detail, while also providing a way to earn a high score. This desire was the spark for the game’s Legends mode. It lets players explore the records of their world in the form of extensive production logs and diaries that the player can read through, ranging from retelling epic battles to a poem written by one of your dwarves. All of these elements add to the player’s score. A direct major influence for this mode was the game over screen in the video game Hack. As you can see, it details how the player died and preserves that information for future playthroughs. Before working on Dwarf Fortress, the brothers had also experimented with something similar in an earlier side project called Fantasy Slaughter. This game featured cards with environments the player had to choose between, which eventually would lead to you facing an opponent. After the player died, they could consult a huge score list with statistics, like how many dragons you killed and how many items you collected.

In Dwarf Fortress’ Legends mode, they took it to a whole other level by allowing players to create grand narrative arcs in their own fantasy worlds. What the brothers had not anticipated was how fans started producing entire animated videos based on in-game stories. During the first year of the game’s existence, Something Awful forum members passed around a single fortress, creating an infamous story called Boatmurdered. One player would play and save its game, then send the file to a next player, and so on, meanwhile documenting the fortress’ downfall. Boatmurdered was featured across several gaming sites and made the front page of MetaFilter, a popular blog, which alerted many people to the game’s existence.

It’s no coincidence that stories like this happen all the time in Dwarf Fortress. A lot of the game’s detailed variables are a result of Zach’s creative mind. See, in order to come up with what elements go in the game, Zach writes a good amount of stories first. Then, the brothers isolate all the elements that make the narrative interesting and even possible in the first place. These then get programmed into the game by Tarn, making it possible for Dwarf Fortress to generate similar narratives as the one’s Zach wrote. What’s more, the brothers even share these breakdowns of Zach’s stories on the Dwarf Fortress website.

Dwarf Fortress was made with a combination of self-taught C and C++, using a free version of Microsoft Visual Studio as the development environment. Tarn basically cobbled together his own engine using OpenGL and SDL, which helped to port the game to macOS and Linux. Tarn prefers using his own engine over something that already exists, like Unity for instance, though he does admit it can be a real pain if he needs to accomplish anything other than text graphics. He uses the audio tool FMOD for sound design.

The game has been in development for so long that Tarn occasionally runs into bits of his code he hasn’t seen in years, specifically when he’s implementing new features. These bits of old code can cause unforeseen problems, requiring Tarn to always double check if everything is working as intended. For instance, if teleportation were added to the game, it would influence economy, diplomacy, espionage, and many other systems, making it extremely complex to implement.

People from the Dwarf Fortress community have offered to go through all the lines of code for him to clean it up where necessary. However, Tarn prefers to do all the alterations to the code himself and have full control over what goes in. Another reason for not letting fans help him, is that Tarn worries he would end up more as a project manager, which is something he’d never want as he loves programming too much. Additionally, he and Zach have very deliberately decided to not turn Dwarf Fortress into an open source project, knowing there’s people and companies out there that would copy the code and sell it in the form of a different game.

While having no desire to be project managers, the brothers do their best to keep the community informed on their progress. Aside from regular updates on their website, Tarn also answers 20 to 80 questions every month and does his best to look at all the player suggestions. “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.”

The community also plays an important part in bug tracking and, as mentioned before, the creation of mods. Fans provide mods that make the game more aesthetically pleasing, add more items and creatures, fix vital bugs and more. All of them are backed with the full support of Tarn and Zach. The community also maintains an extensive wiki, and funnily enough, the two developers even consult the wiki themselves when working on the game. Fans also organize meet-ups where they wear homemade Dwarf Fortress shirts and share their love for the game. Tarn and Zach occasionally even join these meet-ups to connect with the community.

Back in 2011, Tarn, now 33 years old, was offered a job at a big game studio, though he never disclosed the name of the company. He rejected the offer, saying he prefers working with just his brother and not being part of the competitive gaming industry. Another company offered Tarn and Zach a 300.000 dollar deal to license the name Dwarf Fortress, but they felt this amount would not compare to the long-term donations they would receive over time.

The following year, important features like abilities, syndromes, detailed cities, tombs and more traps were added to the game, as were vampires, werewolves, necromancers and undead. As with every new version update, donations increased and eventually stabilized again, making it viable for Bay 12 to keep relying on fan’s donations. Every year around Christmas, donations would also momentarily increase. In 2013, donations averaged about 4000 dollars a month. That amount is of course before taxes and is split between the two developers. Naturally, the game’s active player base kept growing as well and Dwarf Fortress was now being downloaded about 100,000 times a month. Despite that, in 2015, Paypal donations started dropping below 3000 dollars. This motivated Bay 12 Games to set up a Patreon account in an attempt to increase funding again. Thankfully, this worked and allowed Bay 12 to keep Dwarf Fortress free of charge.

After a decade since the original release, Bay 12 released version 0.42.06, symbolizing they were 42% into finishing version 1.0. Right after, the brothers began working on a myth generator and magic system, which would usher in a whole new way for players to create and experience their own stories. “I haven’t done anything remotely like creation myth generation before. It’s exciting. It’s like starting a whole new game. And it really is a separate project. And then it just becomes one of the little cogs in the Dwarf Fortress machine. They’re all separate games in a way, but then the interconnections start to grow and you have a common framework, but they’re all very different from each other.”

Procedural myth generation would allow Dwarf Fortress to simulate gods, demons, heaven or hell and anything in between. The ambitious magic system on the other hand will be one of the most difficult and impactful changes to the game, something that’s going to take a long time to implement. Tarn and Zach have gone on record saying they aren’t afraid to just take the game, pitch it on the ground, and put it back together if necessary to achieve their vision.

The brothers love the idea of implementing magical abilities that would extract a heavy cost as the game progresses, an idea that’s taken straight from the Conan The Barbarian movie. They want to do more than just create a set of spells depending on your level and work with the possibility of making deals with spirits and otherworldly beings, and having to maintain a social relationship with them. Dwarves could chop down the wrong tree, angering a tree spirit in the process, and then would be able to interact with them, with many magical implications as a result. Additionally, Bay 12 plans on adding tools for players to edit maps and myths themselves. Essentially, this would mean you could play god, shaping the world and its deities as you see fit. Zach gave the example of being able to recreate the world of Westoros and live your wildest Game of Thrones fantasies.

Sadly, finishing the magic system and myth generation would have to wait. In March 2019, indie developer and publisher Kitfox Games, announced it was partnering with Bay 12. Though Tarn and Zach had always strived to stay away from the more mainstream gaming industry machine in order to maintain their freedom as developers, a family health crisis made the brothers reconsider. “This is primarily because the healthcare system just really sucks, and I got sick, and it ended up costing a lot of money, and my wife’s company was able to pay for it with their insurance. But if the same thing happened to my brother, we’d be wiped out, Bay 12 would not exist. And so, that was the original reason why we’re doing this. It’s also exposing the game to a wider audience that we haven’t reached before, so that’s pretty cool.”

The illness Zach had to get treated was skin cancer. After removing a tumor on his nose, it grew back within a matter of months and resulted in Zach needing partial reconstructive surgery for his face. Unfortunately, cancer runs in his and Tarn’s family as their parents have been dealing with the disease themselves for years. On top of that, Zach has been on expensive medication for many years to treat his bipolar disorder. The brothers don’t like sharing health related news with their fans, seeing they don’t want people starting a fundraiser, but Zach opened up about it during a Vice interview, shortly after the Kitfox announcement. With Tarn and Zach now being in their forties, it’s no surprise they’re getting increasingly worried about their health and the potential cost of treating more forms of cancer.

Before signing a deal with Kitfox, other indie publishers also expressed interest in Dwarf Fortress, but Tarn and Zach needed someone that was willing to go beyond the standard tasks and also be in charge of things like hiring extra developers. That way, the brothers would be able to focus solely on development. Tanya X. Short, co-founder of Kitfox Games, and her team were more than willing to take on all those responsibilities. The first order of business was to create a commercial version that could be sold on Steam and Itch.io. Tarn and Zach sat down with the team at Kitfox and thought, “What would be the best possible commercial release?”

“So basically the whole process of commercializing the game meant we really wanted to think about all of the things that were making people bounce off of it. And it’s a long list. Because the text graphics are actually very bad for some people. And then the keyboard input was nonsense. There was no mouse input at all. No tutorials to speak of. People like to say, ‘I pop open Dwarf Fortress and then I also have the wiki on another computer running’. And that’s probably still true, but that’s not a good ad for the game. So we added tutorials, tool tips. We changed the flow of certain menus, completely rewrote other menus, and we even changed the AI behavior of the dwarves in certain cases. This was just a process of trying to maximize the number of players.”

Their plan was to release it in a couple of years so time played a factor in some of the design decisions. For instance, developing 3D graphics would be too time consuming so the team settled on a top down 2D pixel art aesthetic instead. Kitfox hired Mike “Mayday” Madej and Patrick Martin “Meph” Schroeder, mostly known for their tileset mod packs on Dwarf Fortress, to create pixel art of every single element. Musician and sound designer Dabu was hired to create a fully-produced premium soundtrack. Before the official release date was announced, the Steam page stated that ‘time is subjective’, a reference to the games’ uniquely long development cycle and a statement that the game would be released when it was truly ready. To put fans of the original at ease, they also stated the classic version of the game will still remain available for free and will receive the same updates as the premium version.

On December 6th, 2022, after about three years of dedicated teamwork, Dwarf Fortress’s premium edition was officially released. According to Tarn, the response to the commercial version has been mostly positive, although there’s a vocal minority on the official forum that claim the brothers sold their soul to the devil. Overall though, the community was very excited about all the changes and this was evident when the first sales numbers came in. Kitfox had estimated Dwarf Fortress would sell 160.000 copies in two months’ time, instead selling that amount within 24 hours of release. After a month, the game sold almost half a million copies, blowing away all pre-release sales estimates. The brothers’ health insurance problem, the main reason for doing all of this, was taken care of overnight. Tarn and Zach thanked their fans. “Dwarf Fortress is a wild success it seems, but it is all an illusion without you. None of this was possible without the support of the people that helped us out. Along these amazing 20 years, by far the best in my life. We win… and it’s your fault!”

To help manage the existing community and the influx of players discovering Dwarf Fortress for the first time thanks to the premium release, Kitfox hired content creator ‘SalfordSal’. More surprisingly, a second programmer from the Dwarf Fortress modding community, with the username ‘Putnam’, was also hired. The brothers describe it as a historic moment, since it was the first time someone other than Tarn and Zach has seen the Dwarf Fortress source code. It’s a big adjustment for Tarn. Even though Kitfox handles most of the hiring and marketing outreach, Tarn still has to manage people to a certain degree. As mentioned earlier, he never wanted to become project manager but he and Zach realize they have to adapt.

Ultimately, they’re happy more people are discovering and actually playing the game now. Their current goal is to keep the operation light enough so they can just keep going with what they have. In the past, Tarn has mentioned that even if version 1.0 would be finished one day, he and Zach would just continue improving the game for the rest of their lives. For now, the brothers have years worth of features left on their list to add to the game, with the enormously complex magic and myth system on the top of that list.

Though the brothers are now millionaires, they remain remarkably calm under these latest developments. While Zach has mentioned he and his wife plan on buying a new house, Tarn admitted he hasn’t had much time to think things over. “It’s just a ton of money, but it’s also for 20 years. So when you divide that by 20, you’re kind of back down into normal tech salary range. Which is still pretty high, obviously. I’ll have to rework… my, uh, whole life, figure out what the heck’s going on there. I don’t really have any ideas or plans right now. I’ve just been thinking about dwarf stuff. We don’t have plans to sell Dwarf Fortress 2 or whatever in five years, right? It’s just the same game. So this tail off this game, for however long it goes, that’s it. We have to be responsible with what we’ve got, now, for the rest of our lives really.”