In February 2014 Xavi "SuperMalParit" Canal, of tiny indie developer SMP, told the Towns community he had had enough. He would no longer work on the game that over 200,000 people had paid for. For most of its players, it was final confirmation of what they already suspected: Towns had been deserted.
That, most thought, was that. But then there was a sliver of hope: in the background Xavi hired a programmer, Florian "Moebius" Frankenberger, to take over development of the game and continue updating it. The community raised its collective eyebrow. Would players finally see the Towns they thought they were getting when they handed over their £20?
That hope lasted little over two months. This week Towns once again hit the headlines when Frankenberger announced he, too, had had enough. Players cried foul. Towns would remain unfinished. Promises were broken. Customers felt misled. Some demanded refunds. Others called Towns a scam. But what really happened?
The story of the fall of Towns is the story of a small and young development team who made costly mistakes. It's the story of how real life affects work. It's the story of how angry forum posts can cut deep when they're left unanswered. And it is a story we may hear repeated in the years to come as more developers and incomplete games flock to the increasingly open Steam platform.
Towns, a modest city building game inspired by Dwarf Fortress, was one of the first 10 games to successfully pass through Valve's fledgling Steam Greenlight system. Valve's idea was that the community would help it decide which games should release on its digital platform. At the time - the autumn of 2012 - it needed help. Steam had exploded, with a rapidly rising userbase and an enormous number of developers desperate to sell to it.
Towns whizzed through the Greenlight approval process. The Steam community thought it had potential, and so the game secured enough votes to earn the right to be sold on the Holy Grail of PC digital distribution. Looking back, its creators admit they got carried away.
"We always felt good about our game," Towns lead developer Xavi Canal tells Eurogamer. The 36-year-old husband and father of two young children hails from Barcelona, Spain. English isn't his first language, he says, so he prefers to chat over email.
"But one of our mistakes was the too early release on Steam. Ben warned me about that but I finally decided to release it."
Ben "Burningpel" Palgi was Xavi's partner in the development of Towns. Hailing from Israel, the younger of the two developers worked from home getting Towns off the ground. He tells me his side of the story from half the world away, again over email, because English isn't his first language.
"We always felt good about our game. But one of our mistakes was the too early release on Steam. Ben warned me about that but I finally decided to release it."
Towns lead developer Xavi Canal
Towns released before Steam Early Access had even been announced, and so this idea of games releasing incomplete, so popular now, hadn't entered the gamer consciousness. Then, we expected games we bought to be finished, as bug free as possible, and properly tested.
As many players discovered, Towns was not complete. It was still in alpha, really. Indeed it had already gained a positive buzz from alpha backers since development began in 2011.
The main complaint was that existing features were not complete. The tutorial, for example, was essentially a wall of text that served as a placeholder. But there were also missing features. The buried town feature, which meant towns would would show up in other players' worlds, wouldn't be added until later. Trading wasn't at launch either.
This was the disconnect: the game's description on Steam said nothing about Towns still being in development, nothing about it being an alpha. Paying customers felt misled when they played it.
"Those missing features were not why people felt upset," Palgi says. "They felt upset because the features they did experience were not up to their accepted standards."
Prior to launch the pair argued, and it sounds like those arguments were quite heated.
"The short Greenlight campaign was a tremendous boost to our confidence and we were literally overwhelmed with the reaction to Towns and with the fact it got selected in the first batch, but, this and the reactions of our alpha backers have caused a rather peculiar effect," Palgi says. "It made us believe the game was good enough and ready to be sold straight away.
"I had long talks with Xavi about the game state before launch and how I felt it should be postponed at least until the inclusion of some features and the expansion of existing ones. But the circumstances, and what I can only describe as a blind spot, made him feel the game was finished and in a good state to be released through Steam."
Towns was fully playable, as Canal points out now, but required major patches before it could be considered anything close to finished.
"I can say no," Canal says, nearly two years later, "the game was not finished on the release because it still missed some major features."
"Because Xavi felt the game was in a good enough state, and could be considered as finished on its own, he didn't see it as selling an incomplete game," Palgi says, "so there was no question about advertising it as complete.
"The game was released with the thinking that it was ready and finished and patches applied afterwards were for expanding an already complete game, and not as part of an ongoing development of an incomplete game."
Looking back now, it's clear Canal got carried away, fuelled by encouragement from alpha backers and the Steam Greenlight campaign. Here were two young and inexperienced developers who were on the cusp of selling a game from the biggest PC game shop in the world. Towns was reported on by hundreds of game websites because it was in that list of the first 10 games on Greenlight. SMP was on the hype train, and it clouded their judgement. Or, as Palgi puts it: "This was a case of an inability to criticise your own work, rather than knowingly releasing an unfinished product."
Towns launched on Steam on 7th November 2012, and soon after the complaints began pouring in, on the Towns forum and on the swelling Steam forum. Canal ended up having an email back and forth with Valve and a week after release Towns' description on its Steam product page was changed to warn people the game was incomplete. SMP released a demo and posted on the Steam forum to recommend people try it before buying the full game. "Of course that was not enough," Canal says now.
Then, a few months later, in March 2013, Valve launched Early Access. It was where Towns should have been from the beginning, but it came too late. "I would have waited for Early Access to kick in and take that route," Palgi admits, in hindsight. "In fact, if we had done that, we would have developed in a far greater positive environment, and perhaps the notion that people pay us so we could actually put that back into developing the game would have been stronger in some of our minds, resulting in a better game, even if it took us a long time to do so."
"I would have waited for Early Access to kick in and take that route."
Towns co-developer Ben Palgi
The release of Towns on Steam should have been cause for celebration in the Canal household, but for Xavi it was a time of great sadness. It was around this time that a close family member was diagnosed with a potentially fatal disease, one that derailed development at a time when players were demanding it most vehemently.
"I remember a lot of nights coding in the hospital with my laptop," Canal says. "Also, with all the rage on the forums the pressure was too high for me. At one point I decided to get away from it all for a few months.
"Of course that was even worse for our reputation and when I came back there was no way for me to stop that snowball."
Palgi remembers the pair retreating into their shell when the feedback became too much. "The rage was so strong that it took us a few weeks if not months to recover from it. It was extremely hard working under the stress and our reaction was to mistakenly shut ourselves out from all the drama surrounding the bad launch to try to work on the game in an 'industrial quietness'.
"This, I believe, caused even more disappointment because it made some of the players think we just took off and ran with the money."
Money is an interesting word when it comes to Towns. Despite the negative reaction, it sold wonderfully well, and for a time was in the coveted Steam top-sellers list. But exactly how well?
Canal tells Eurogamer Towns has sold nearly 250,000 copies, "a lot of them because of discounts". His company, SMP, made around $2 million in gross revenue from the game.
What happened to all of the money?
"This is one of the things people used to rage at us about," Canal admits. "The 'you took the money and ran'. We are really far away from that. By doing some simple math you will see that gross revenue is far away from net revenue. The typical 30 per cent vendor splits [that's the cut Steam takes], the bank fees, about 50 per cent taxes and the split of the money between the team members reduces the amount by a lot. Also, by living in Europe you have to convert the USD to EUR and pay the VAT taxes."
Many players have wondered, quite reasonably, why SMP didn't use the money made from sales of Towns to complete the game, perhaps hiring an additional programmer. "The question of what happened to the money is a genuinely good question to me and you," Palgi says, "but with having the notion that the game was finished and realised, it becomes irrelevant.
"If you cannot recoup further investment through maintaining sales, and you think the game is finished and in a satisfactory state, you will not just spend money on it to improve something you think is already good enough to begin with.
"Obviously, good developers would invest that money back into their product regardless of its ROI potential because they wish to give most to their costumers and to fully realise their own game potential, but in the light of the stress and burnt out feelings, this was sadly not the state of mind regarding Towns."
So what did Canal spend his cut, when all was said and done, on?
"Basically the revenue was spent on personal bills," he says.
It's at this point that Dungeon Keeper-style real-time strategy game Dwelvers rears its head. SMP invested in Dwelvers - indeed Palgi now works on it - but to what extent?
"When we did that, we were still working on Towns," Canal says. "And since I already saw it as finished, I didn't think it needed more money. So, the money for Dwelvers wasn't coming instead of spending money on Towns, it was unrelated and was done because we thought the developer of Dwelvers was a talented man who should be able to work on his game."
"When we funded him, we were still working on Towns," Palgi elaborates, "and it's true that it was money that could have gone to Towns, but at the time, this wasn't the mentality."
For his part, Canal points out that SMP released six major patches and many minor ones for Towns post launch. "We can't say we did nothing since the release."
"I was, and still am, burned out."
In February 2014 Canal announced he was leaving development of Towns. For him it had all become too much. Looking back now, it's obvious that stress of caring for his family and dealing with the Towns community got on top of him.
"I was, and still am, burned out," he says.
"I was working 24/7 on the game... all the stuff around forums, vendors, email support... even on vacations. Add the personal issues and all the rage on the forums, it was pretty stressful."
But there's another reason Canal called it quits: the nature of Towns itself.
"Our game is that kind of games where you can always add more stuff to it. No matter how many things you added, there is always more things to add. From content to features. So, I just decided to stop that after the v14 build."
Palgi tells Eurogamer he was surprised by Canal's decision, but in hindsight should have seen it coming. "While the stress he was under was something I was fully aware of, I assumed he would get over it and we would continue as before," he says.
"What did shock me was that he decided to end the development just when we decided on a candidate [Frankenberger] to replace him as a programmer. We had sent out a contract. This was a long process of choosing between a dozen candidates.
"Hearing it, I left the team right on the spot and gave up all the shares I had in the game."
Palgi officially left SMP on 1st January 2014. Canal later hired Frankenberger. "Again, from his perspective, he was hiring someone to work on an already finished, and while not perfect, good game."
"Because my own indie game is not for sale yet, I don't really have any experience in how the sales run."
Florian "Moebius" Frankenberger, who took over development of Towns after Xavi Canal quit
When Florian "Moebius" Frankenberger announced he was leaving Towns behind this week he became the focus of player anger and the wider scrutiny of the internet. But in truth he inherited a tricky situation. Explaining his decision to leave the project, Frankenberger said sales of the game were less than a third of the amount he had been led to believe would be the case when he took over development after Canal left. Frankenberger's deal meant he would receive 15 per cent of revenue after tax and Steam's traditional fee were removed. Clearly, this wasn't enough.
Does Frankenberger feel he was misled by Canal? No, he tells Eurogamer. "I honestly think it was not done intentionally. I know now that the numbers went down before I even signed the contract but I believe Xavi when he says he thought this was a normal fluctuation. After all this is my first indie game (except for my own) that I've been working on. And because my own indie game is not for sale yet, I don't really have any experience in how the sales run."
The entire episode raises important questions: are developers obligated to work on games released on Steam Early Access, or in alpha states, until paying customers consider them to be complete? If they are, what's the cut off point? When should they stop and move onto a new game? Who decides when and whether a game is officially finished? And what happens when developers decide to leave a project paying customers consider to be incomplete? Are they entitled to refunds?
"I don't think there can be a given standard of how a finished game should look like on paper," Palgi says.
"You can't break a game, look at its parts, count them and decide it's finished. You need to look at it as a whole and feel its experience rather than its appearance."
Don't blame Steam or Early Access, Palgi says. Game development is "in the hands of the developers themselves". "It's their game. It's their reputation. It's their responsibility.
"People should not judge entire platforms or business models because of few failed investments," he continues.
"The Early Access and Kickstarter models have already enabled us as gamers to enjoy far better and interesting games that we would not have had if those developers were still tied to publishers.
"I strongly believe in those models and have paid for several games in Early Access. Some of them were a disappointment, but blaming it on the payment model is like someone who stops believing in paying with a credit card because someone on eBay didn't deliver his product.
"I think it's silly to overlook the extreme benefits this system gives because of the obvious flaws it can have. Also, I am willing to bet that traditional projects get cancelled or delivered in bad states just as much Early Access or Kickstarter projects, if not more."
Frankenberger tells us he asked himself similar questions many times over the last few months. "And I think it is especially complicated with Towns as it was on Steam before Steam had the Early Access program.
"I think customers can expect the development of a game up until a certain state. I think this state is reached when the features that were used to promote the game have been added to the game and work properly. So the customers got what they paid for.
"So I can really understand the people that have been promised features that are not in the game right now."
"I am willing to bet that traditional projects get cancelled or delivered in bad states just as much Early Access or Kickstarter projects, if not more."
Towns is abandoned, but not cancelled, its developers insist. Indeed Canal says a "well known player" who works for "one of the bigger gaming companies" has already expressed an interest in the rights to Towns and its continued development. "In the upcoming days we will try to reach an agreement."
This was news to Palgi and Frankenberger, but both want to see Towns realise its potential. That's perhaps surprising, given what's happened, but there it is.
"I am as upset as everyone else, if not more," Palgi says. "For me this was an investment of two-and-a half years of hard work and sometimes sleepless nights. It was my first commercial title, which I knew would have a strong impact on my entire career as an indie dev.
"Xavi has stated on our forums that several studios have already approached him with the intent of buying the IP and continuing it themselves, so there's still hope this wonderful concept and fresh take on the genre might be realised. All in all, and even though the game was clearly not fully realised, I think it's still a good game that has interesting concepts within it, and that can provide content and playing hours that even more expensive triple-A titles can't provide."
Towns is no longer Palgi's concern, but the fallout from its collapse has stuck with him. He's working on Dwelvers, which, as mentioned, SMP once invested in.
"It's a totally different experience and while I am perfectly aware some might be reluctant to buy Dwelvers because of my, and SMP's, involvement in it, I am pretty sure in the end most people will realise Dwelvers is a good game with a talented, level headed and honest lead developer behind it."
Frankenberger has picked up development of the indie game he was working on before he started on Towns. It's out this autumn.
As for Canal, his future is less clear. "I'm trying to reach an agreement to let the development continue. After that, I don't know what will come. I think it depends of how all the Towns story ends."
The story of Towns is a story of regret and mistakes. Mistakes made by a young, inexperienced duo of developers who have suffered because of their naivety. Was Towns a scam? I don't think so. I think Towns was a game made by a couple of developers who messed up. The story of Towns is a warning: to gamers who pay for incomplete games, and to developers who make them. It is a cautionary tale.
The story of Towns ends with an apology, from Palgi, Frankenberger and, perhaps most importantly, lead developer Canal.
"I understand the criticism because things from the outside are always clear," he says. "There is a never-ending war if I try to explain my reasons to them, and that's not good for any of the parties.
"So, the only message I can say is 'sorry'."