Hi everyone. I know the new version isn't ready yet but I thought I'd post a new update to keep everyone informed on what's happening. Hopefully the version won't be delayed much longer!
Warning: This is going to be a long one...
Update on the Delay
Firstly, I'm sorry the new update still isn't ready. I had high hopes it would've been finished weeks ago and this delay is really frustrating. Nobody wants to see progress on the game more than the people making it, especially when the company is bleeding cash. This is not a conscious choice to sabotage our own success by working slowly.
Instead, the delay is the usual nasty problem of us trying to add something cool to the game that players wanted, and which seemed pretty straightforward to code - a naturally forming path system under the feet of your citizens - only to find out it's a complete horror show to implement as we reached the halfway stage.
But by that point, we'd spent so much work (and money) on the feature we didn't want to give up. We knew there would be a way to finish it, if only we kept banging away at the problem. And suddenly 6 months went by. Days can be a blur when you spend them staring at a computer screen.
The reason this feature has been such a pain to add is because it connects to the AI navigation system, which is probably the most complex aspect of the game, and has given us numerous headaches in the past. Until this version we told our citizens to pick off-road routes, and they did so by calculating a line across the terrain and adapting on the fly as necessary. Now we want them to (sometimes) ignore that system and pick paths that another citizen formed, a path that possibly didn't exist 5 minutes ago. But not always. Sometimes following a long path doesn't make sense when you could just cut across a field, so we needed to add in decision making factors and give each route a "cost" which the citizen weighs up.
And such tricky navigation issues come after all the graphical issues we've already (mostly) solved. For example, paths that form themselves underfoot also need to automatically form good looking intersections that still look good as more and more paths connect to them. And they need to look good on flat and mountainous terrain, and in snow, forest or desert biomes. Oh, and sometimes the player will want to slap down a huge building right in the middle of that complicated intersection just as 10 citizens were approaching it with urgent deliveries while being chased by enforcers. And all this has to work flawlessly.
It hasn't been easy, and solving all these problems is the job of just one man, our main coder Nick, a guy making his first ever proper game (like the rest of us) and who’s working alone at home with no one to support or advise him on the technical issues. Don't get me wrong, I think Nick is a genius-level coder, and building something like AS as your first full game is amazing. I can give him literally any feature request and provided he has enough time, he will somehow work out how to do it and make it run efficiently no matter how complex it is. And he does all this on pity pay (my term), and with serious health issues that are no joke. He never complains or loses his temper with us (perhaps he does with Unity) and accepts tasks from a first-time game designer who can't even work a calculator, let alone understand programming.
Which is why this version - which I personally expected to be out in April - still isn’t ready. Knowing our difficulties doesn’t change reality and give our players new content, but I hope it shows you the state we’re in.
To be frank, I wish I hadn’t put this feature in the game and I'm still unsure if we should keep working on it. I don't think it's worth the damage it’s done to our finances (or review score) over the past 6 months. The trickle of money we earn from AS is more important than ever due to day job losses from the coronavirus pandemic, but that precious money is currently flowing into a feature that is purely cosmetic and adds probably less than 30 minutes to an average play time (the extra time coming from people gawping at the path system and wanting to expand their town further). My mysterious calculator says we'd have to sell about 3000 copies to break even on this feature. And we sold 200 copies last month. Hmm… I think I'm learning something important about videogame production.
We're in a problem of our own making. This is not me begging for sympathy or complaining to you, our wonderful players. This is me publicly sharing my own idiocy and learning painful lessons as a developer.
Which I guess is what a dev blog is for!
Blame the Banished Guy!
All this hassle and delay with the path system has forced me to look at the deeper problems in our company, problems that have blighted our team since day 1.
What is our biggest problem as a team?
Why are we slow?
I (jokingly) blame the guy who made Banished.
5 and a half years ago, back when we were starry eyed noobies who wanted to try and make our first ever game, we decided to make a 3D city builder (as opposed to a 16-bit style 2D game like most sane startups). If one person made Banished, we wondered, how hard could it be for 3 beginners to make a similar game? But I didn’t realise at the time that the guy who made Banished had 10 years industry experience. Nor did I think twice about designing a game even more technically ambitious than Banished. In Banished you could only look down, you could only build in a flat forest, and you could only build in a grid system. Wouldn't it be cool if we did away with all those technical limitations on our first ever game?! I'm sure first-time coder Nick can make all that while simultaneously pumping out new buildings and gameplay features to keep the fans happy.
Well, here we are. 5 and a half years later. Still wrestling and suffering from the implications of these early naive decisions.
If we hadn't recruited a second part-time coder (Adam) by simply begging on Reddit, all our working hours would be consumed with fighting the technical foundations of Atomic Society. That’s not how you make an Early Access smash hit. But like all the stickiest problems, we were just good enough to scrape through the hard work rather than hitting a brick wall. If we'd hit a brick wall it might have actually forced us face harsh truths. Instead, we discovered if we banged our heads against any problem long enough it went away, at least for a while.
But I don’t recommend headbutting as a long-term career.
On top of struggling with over-ambition, one of the worst parts was not even knowing the path we'd chosen was supposed to be hard. We work in relative isolation. I personally don’t have any other developers to talk with. Working in a vacuum can all too easily lead to blaming your co-workers, or blaming yourself, when in fact nobody's to blame. It's just that you're in an extremely difficult situation.
Which is why I was deeply relieved when I recently stumbled across a dev blog written by the man who actually made Banished…
Perhaps you wondered why there hasn’t been a Banished 2. At least one of the reasons is he's having the exact same struggles we are. He's trying to make a game that is a lot more like Atomic Society on a technical level. Large portions of his dev blog could’ve been written word for word by me.
This man made Banished. He is certainly not behind schedule because he lacks talent or experience. But no matter who you are, certain game features are painfully time consuming and difficult to implement. And unfortunately, we decided to make our first game with a bunch of these features.
To any aspiring indie dev reading this, beware! Here be monsters. Especially the most vicious monster of 3D pathfinding. AI pathfinding is literally a full time job on bigger teams. Stick to games where the AI is on a Super Mario level of complexity. We spent much of the last version redoing our pathfinding system to eliminate bugs and now we're spending even more time trying to integrate a new feature into the pathfinding system.
And it's bleeding us dry.
Brick Wall Time At Last?
The real reason our team is slow is not just because we tried to expand on Banished. We also add features that have no financial benefit, and we proudly stick with them even if they prove to be incredibly difficult to do. I’m astounded by my own naivety.
Right up to the present version we assigned tasks by me asking Nick or Adam "is this feature doable", which basically means "can it be done in 3 months"? If the answer was yes, we just went for it. I never stopped to think of the value of said feature.
Let's assume you're in the privileged position of being able to pay your development staff (British) minimum wage. 3 people working full-time for 3 months on a single feature means that little feature will cost nearly £14,000 in wages, and that's assuming it doesn't run late. 14k decisions should not be taken lightly, it's the price of a small car. But we'd been working for free for so long before coming to Steam that we'd developed bad habits, like implementing whatever felt good at the time. And if a thing took ages to make, we had nothing to lose but time and sanity. That kind of attitude doesn't lead to a sustainable salary.
These recurring mistakes placed us in a vicious circle. We’re running out of money fighting problems we shouldn’t be fighting in the first place, and because we’re distracted and exhausted from fighting such problems, we lose momentum on updates and serious bugs slip into the game, which leads to bad reviews, more delays and lower sales.
I like to be as open as possible on these blogs because most people presumably aren’t hate-reading them, and maybe they'll help someone out, so I don't mind saying that in 6 months our little company will be flat broke, broke to the point of literally going (back) to the welfare office. In January we had 12 months runway (Silicon Valley speech for “time to poverty”) and we blew half of that on this path system. I know we’re talented enough to eventually make it through anything - even going bust - but there has to be more to life than survival.
I think I made a whole game about that...
What To Do?
In summary, we're insanely ambitious (emphasis on the "insane") when we should be ultra cautious because we're poor beginners. Technically simple games sell just fine. We choose tasks with little consideration of their financial implications and whether they will meaningfully increase player engagement (and therefore sales). And we're stuck making a game that requires 6 months of work to implement cosmetic features. These sound like obvious mistakes, and they are, but until now we were in a survival mentality of "just get it done".
At least now I know why Atomic Society has been such a slog. The next question is what to do about it...
The simplest but scariest solution is to totally scrap the dreaded path system that’s taken 6 months, fix the worst bugs in the game, and let Atomic Society be what it is: a small but at least stable and cheap curiosity on the Steam store. Most of our bad reviews are about bugs rather than content, which is understandable. We'd fix them and move onto a (much, much simpler) new project with all our hard earned knowledge before poverty really kicks in around Christmas and perhaps live to fight another day. This is probably the smart choice.
But can I really ask Nick to scrap all his work from the past half year, just chalk it up to a learning experience, and give up on a feature that is probably 90% complete? If I wait a week or two it might all come together…
And more importantly, am I willing to walk away from my beloved Atomic Society?
I recently went through our task list, getting rid of ideas that would never get done (employing a little of that cost analysis stuff) and just erasing plain bad ideas that were invented 4 years ago. After this long clean up, I was still left with 219 features that could make great additions to the game. Ideas ranged from mundane stuff like achievements and foreign language translations to schools and a full on end-game story, involving you taking on the raiders once and for all and being reunited with the people from the bunker. If we put these ideas into the game, we’d have a game to be proud of, right? We'd be an Early Access critical "success".
Maybe. If we’re willing to work for free for another 5 years, and sadly the world has already moved on from this game.
Despite romantic notions of the plucky entrepreneur, nobody really works for free - unless they’re immune from basic survival needs. Somebody has to pay the bills and put food on the table. If it isn’t you, it’s your wife or family or loan shark, etc. We self-funded Atomic Society before Early Access through day jobs, primarily through our artist's (Nani) day job. Her salary allowed the rest of us to focus more on AS. She was the agriculture to our early civilisation. But her day job doesn’t exist anymore thanks to coronavirus.
This all feels like I'm considering a breakup... with a videogame. Atomic Society and our team had a whirlwind romance, we could be so good together, but it's been almost 6 years, and every day feels like a battle. Maybe we could make it work, but maybe the smartest thing is to leave while we're still friends.
I don’t know what the answer is yet. We’ll have to talk as a team, but I think we're getting close to some big, life-changing decisions.
Anyway, that's why the version is running late! :)
Anything Else To Talk About?
The last six months have not just involved 4 people stressing over a path system and deep introspection into one's own business limitations. A few other things are happening...
The Save Bug
Thanks to our players reporting it, we are definitely aware there’s a bug in the game that can sometimes cause saving to freeze at about 50% which is something we want to fix urgently. This is the worst kind of bug but annoyingly it isn’t a bug we’ve been able to reproduce at our end. We’re trying to catch it, but if anybody suffers from this bug, it would be lovely if you could quit the game when it happens and email us a output_log file.
Here’s how to find the file after you've quit the game:
We will fix this bug and your file could make all the difference.
New Building: Terraforming Station
While Nick has been busy with the path system (understatement alert), Adam has been busy beefing up the patch notes with some new content. We’ve started on a terraforming/mining/digger station (haven’t nailed down the final name).
I’ve heard that players would like more control over the layout of the landscapes and this building will let you do just that. After building it, you can mark out areas for workers to flatten and they’ll go over and crush the landscape, kind of like when you place a building. This should help players expand their town, especially on the more mountainous maps. But naturally, I’m a little concerned about the bugs this might create, as anything that changes the landscape can affect the AI and the dreaded pathfinding won't like that... But we’ll see. Hopefully this isn’t a task hiding a handful of complications. We're trying it out anyway and it's coming together.
Here's Nani's work in progress model for it.
Trees Now Have A Function!
When we released Atomic Society, I was weirdly adamant I didn’t want the game to involve cutting down trees. It sounds foolish, but I was so burnt out on crafting/Minecraft games at the time I just didn’t want to make a game where you start by hacking down trees. I even went so far as to research why cutting down trees might be a bad idea after a nuclear war (they’d be radiation sponges from the toxic rain). This wasn’t just my anti-lumberjack prejudice, there were a few technical issues, like balancing the maps that don’t have many trees, and persuading the pathfinding to cope with disappearing trees (hi pathfinding). Altogether it just didn’t seem worth the effort. But I’ve always wanted to do something about this criticism. Therefore the upcoming version will at least let you gain something from trees.
The Scavenger Hut now has a radius circle around it (work in progress visually), and the more trees in that radius when you build it, the more bonus lumber you’ll gain on a periodic basis. This increases the strategy in building as you may want to expand towards natural resources and rely less on salvaging. We’ll balance the game so this doesn’t make it too easy, but it's ready to ship.
The next version comes with a new track from our composer Dawid Dahl. It’s a special guitar track that will play when you finish all the goals, including the new goal to raise a family/heir.
Working with Dawid on the soundtrack was one of the best creative moments of the whole development process for me, hearing his demo tracks for a game I was designing felt magical. Whatever happens with AS, I'm really happy with our soundtrack.
And there’ll be a bit more of it in the next version.
Personal: Do I Like Games Again?
To get into the weeds again, around November last year I lost the desire to play games entirely. It wasn’t depression, I’d guess it was burnout from viewing games through a professional lens for too long. I’m someone who wants to make games passionately and scrutinises the slightest detail of almost every game ever made – which is not really how you fall in love with games. It can actually turn you against them. I felt “done” with games.
This isn’t the first time I’ve felt such a loss. It happens every so often, but this was the longest time such apathy had stuck around while I was making AS. I got slack at making dev blogs, and lost some of my fire as the unofficial director of our little team, which was contagious on the others, or so my ego likes to think. It also caused inevitable “is this game dead?” forum posts.
Sometimes I think games are kind of pointless, and making them is folly. In a sense I’m right. Games don’t matter. The coronavirus outbreak didn’t change my mind about that when 700+ people were dying every day in my country. I felt trapped. I wanted a career where I don’t have to worry about reviews, missed deadlines, and being paid.
And then as summer bloomed, all my gaming passion came back. For the first time in 6 months, I genuinely wanted to play games again. The burning passion for games that’s been with me since childhood resurfaced. It’s weird, like a tap that can shut off and on. If I had to guess, I’d say I’d matured a bit and started to see games as my best shot at a career no matter what I thought about them - which led to me figuring out the problems with our slow development. Games are not just artistic expression, they're a livelihood.
I’m almost 38, but I’m a part-time janitor. If I want a career, it’s probably make a (popular) game or scrub floors. Scrubbing floors is actually pretty chill, but I like to think I have some talent for designing games too. I don't think it's just coincidence at least 3 big studios took a shot at making a post apocalyptic city builder after us (Frostpunk, Endzone, Surviving the Aftermath), and some even marketed themselves on the law changing aspect. Or maybe it is a coincidence. Either way, I now really want a second swing of the bat and to employ my newfound design knowledge and understanding of game development. I feel ambitious again. We've learnt too much to give up. AS was just the apprenticeship.
And ultimately playing a good game is just so much fun…
I've obviously gone on for way too long. This is probably the longest blog I’ve ever written.
To wrap it up, we will come together and decide whether to keep struggling with the path system or just push out what we've got, like the family system and other things above. Or perhaps the path system will suddenly come together at the last minute. Either way, I will try to keep people updated more frequently on how things are going and what the future of the game is.
I hope somebody benefited from reading my avalanche of words, and that honesty about the project counts for something. Thanks for reading and I hope you're enjoying the apocalypse!