Alien Nation: Going ‘Nowhere’ With Duangle

By Duncan Harris on April 2nd, 2014 at 9:40 pm.

This is the latest in the series of articles about the art technology of games, in collaboration with the particularly handsome Dead End Thrills.

A mindbending, mesh-distending life sim set in some vast alien abyss, Nowhere is awesome in the way Spore once was before it mutated into an existential Happy Meal. No such danger with this game, though, the husband-and-wife team of Leonard and Sylvia Ritter [together known as Duangle] now deep into a project that’s content, much like its shape-shifting ‘Nowherians’, to evolve naturally. Quite how it manages to map Maslow’s hierarchy of needs onto a universe full of amoeba people is just one of countless irresistible unknowns.

It’s a weird ecosystem, though, this modern crowdfunding. Almost every aspect of Nowhere’s development is exposed to its potential players – not least its developers who have to adapt to the demand for public alpha builds; competitive promo art; and their unique brand of trippy, nervy video newsletters.

DET: Indies choosing the crowdfunding and Steam Greenlight routes now have to sell themselves as much as their games, often vigorously. How hard is it for the ‘quiet types’ to do, say, a video pitch?

Leonard Ritter: Yes, that would be us!

Sylvia Ritter: They’re pretty stressful to prepare but we wanted to have at least some fun doing it, and I think we did, shooting the video. Another issue with it is to decide whether to devote time to promotion or to development, as we are only two people. We originally agreed to do a video for every alpha – so the next one will also get a video.

LR: I think doing these is a mixed bag of emotions that ranges from forcing yourself to do it, to it being somewhat of a guilty pleasure. I did a music act before this project, and from this I know that people want to see us having fun. So we try to do that.

DET: Does it make you better developers?

SR: The videos make things more personal; people see that there are actual humans behind the project. Paying for these human flesh suits paid off big time! In terms of being a better developer, it makes us better promoters; for actual development, ‘interacting with the community’ is irreplaceable. We run an IRC channel on freenode where most of the conversation takes place, and we have a few regulars there who we regularly ask for help and opinions. One of our fans who is a webdev set up a complete forum for us – that was quite unexpected.

DET: An open source forum, to boot. Any significance to that?

LR: Open source is pretty important to us. We try to keep an almost 100% open source workflow, which is super hard. The engine that’s developed alongside the game is also open, and the game itself runs from source. So if anyone feels like messing around with the code and modding stuff, it’s almost all there. Some parts are precompiled for performance, but it’s 90 per cent Python. The first experiences I had with game development were with modding Quake in the 90′s, so that part is pretty important to us; I’ve only ever seen good things come from that, and it’s the worst when a game is completely closed up but could definitely use a few changes.

DET: What kinds of changes are you anticipating?

LR: The game UI currently lacks customisation options; one player went into the sources and adjusted the keyboard configuration to his French keyboard. Of course I’m not expecting anyone to do my job for me, but we want the game to become pretty nailed down and fixed. And before we get into an argument with a player who doesn’t agree with the ‘director’s cut’, we’d rather just hand over the keys.

DET: Back to the whole crowdfunding thing, development can seem somewhat back-to-front in that environment: your biggest marketing push comes early, if not at the beginning. That much complicate things.

LR: To be honest the dream we had about this was that we would introduce our fundraising campaign, and the attention would be just enough to completely seclude ourselves for two years and only work with the backers. But God Had Other Plans For Us.

SR: The added pressure we got through that was pretty good though, because we were forced to improve our presentation. We can keep tweaking and see what works and what doesn’t.

LR: With my developer hat on I would grumble and say that any distraction is one too much. But we changed the logotype just once, and I think this new one is perfect. It’s interesting because it’s true bottom-up development: all parts inform each other. The logo informs and influences the art style, for example. Through the logo I got a perfect sense of what the beginning of the game would be like. The previous logotype was good but a little too frail and brittle to work well at small sizes. People didn’t get the NOWHERE / NOW HERE ambiguity, either, so there was a need to make it more clear.

DET: So, the changes in logo reflect big changes in the game?

LR: We originally planned to make the game very organic, but 3D technology is something you don’t master, you reconcile with it. To make things with plasticity, you need a certain triangle count; if you can’t do it with the meshes, you need to do it in the textures. But with the Oculus Rift these tricks become obvious. So the style has become more crystalline in the process, and I hope we’re hitting just the right balance of both organic and crystalline design in the end.

DET: The plasticity you mention defines the world of Nowhere. How can technologies like HDR and ambient occlusion help you navigate such a space?

LR: Originally I wanted a little less from the graphics engine, but we learned that, especially in rather abstract environments, those visual cues for plasticity are all the more important. I overestimated the ability of stereoscopy to deliver enough information on its own; it only works for small objects very close to you, but not at all for distant shapes. Oculus even suggest in their guidelines to offer an option to turn off stereoscopy completely to reduce motion sickness, so that means we had to fall back to more traditional ways to relate depth, size, distance… which is where post FX like atmospheric scattering, AO and shadow mapping come in.

The HDR is more of a convenience: it’s easier for me to work with full range, and without it there’s no good-looking way to convey that a light source is glowing or bright. There’s also an additional perk: the filmic tone mapping has a very artful way to get the best out of the flattest colour definition. It’s almost like it does half of the colouring work.

So yes, without a good understanding of distance, navigation is difficult. Something else we’ve been working on is finding interesting ways to navigate through a gravity-free environment; something that supports sightseeing a little better and is more true to the actual experience of space, where momentum is conserved and the moving object doesn’t have that much influence over where it wants to go or when it wants to stop.

DET: Very few zero-gravity games capture that same sense of excited bewilderment that Descent had. Why is that?

LR: I played Descent back in the day. The controls were a little complicated and navigating the maps was quite disorienting. I’m not sure they have to be. The tunnels also had something claustrophobic to them. It seems to be a weird tightrope walk with space games: the space can’t be too narrow but it can’t be too wide, either. Most space games aiming for a realistic representation of space present you with a black desert. I hope we can sidestep all of these issues with our fantastic setting.

DET: The Rift feels like the de facto crowdfunding stretch goal at the moment. You seem to appreciate how distinctive its needs are.

LR: Can I say now that we wanted that device years before it was announced? I’ve been dreaming of a working VR device for ages. I bought one of these VFX headsets from my first salary, as payment by instalments. A huge disappointment. The opportunities I see for VR aren’t so much in replication of reality; to me, that means setting yourself up for disappointment. But when it’s about exploring alien territory and relating impossible experiences that require a high degree of presence, then the device is an enabler, not a hindrance. As an artist, you get a canvas larger than a Cinemascope screen. The next Rift addresses the current issues we have with it, which is the locational tracking and the low resolution.

DET: How do you feel about the whole Facebook/Oculus thing?

LR: WHAT THE HELL. No, I’m joking. But I think that was the first gut reaction.

SR: We’re still going to buy the new devkit.

LR: There are a bunch of great engineers at Oculus who I keep having conversations with on Twitter, and who helped us with our work, namely Tom Forsyth and Jon Olick. As long as they and John Carmack work for Oculus, we’re confident that it’s going to turn out great. Palmer Luckey can’t go on air saying, ‘Look, they’re not using us, we are using them,’ but that is how I interpret the deal. So, we’re cautious, but not alarmed. I think I’ve also tweeted, ‘I really need that device, so I’m going to pretend nothing happened.’

DET: Nowhere is full of grand concepts that only really come through in blog posts, yet screenshots and artwork – dare it be said, ‘trendy’ art styles – often decide crowdfunding’s winners and losers. Is that frustrating?

LR: We’re still figuring out how to communicate it. It’s true: visuals sell best. It is kind of frustrating when your game is explicitly about the exploration of procedural content, and any kind of concept art feels to us like a lie, really, because you can’t fully know what the world will really be like once it’s seeded. But we had a bunch of mails from people who wrote that they get it and they think it’s fantastic what we’re trying to do. The project is way more about the curation of interesting rules than a nailed-down one-time experience. Which brings me to the tragedy of Minecraft.

Most devs I know treat that game like a freak accident, like it shouldn’t have happened, and so it’s unlikely to happen again. That’s probably what they infer from all the cloning that wasn’t equally successful. But I think that trying to emulate this exact game would be missing the point. First of all, Minecraft showed us what players really want. It has one important underlying design rule which isn’t too explicit, which is that every item, object, thing, and block in the world has some sort of a function and isn’t just there to be pretty. Without cloning Minecraft, one could inform his design by this rule and come out with a better game.

DET: There’s a pretty clear distinction between the black light style of the early alphas and the lighter, more textured look of your WIP stuff. Has one subsumed the other?

LR: The black light was a bit of the project that was clear from the start. But it became obvious that it was going to be tiring if it’s just that, so we wanted a bit of a contrast. Ideally it would be like the spaceship from the finale of The Abyss – sorry for the spoiler – where underwater the ship is all dark and glowing mystically, but once it emerges from the water it has more iridescent pastel tones. It felt like that would be a good day/night cycle. and would provide a good backdrop to convey two distinct feelings: one of knowing what’s going on and having a good sense of things, and one of being kept in the dark, being confused, seeing things.

It’s not too great yet but we’re getting there. I think ‘not knowing what’s going on but feeling magically attracted to it’ is one of the feelings we want to explore in the game. A few weeks back we watched all the seasons of Lost for the first time, and it really felt like this was somehow connected to what we were doing. I never had to wait for the next episode, so we could just watch it in one go and accept whatever what was happening, and it was a really rewarding experience. The series has all kinds of troubles but that’s what happens when you’re ambitious.

DET: Leonard, you were in the demo group Farbrausch, right?

LR: Technically I still am because you can’t really quit Farbrausch. You can announce it on the mailing list but no one will remove you from the crew list. Some have tried.

DET: Did you work on projects like .kkrieger and .detuned?

LR: I wasn’t directly involved with .theprodukkt and .kkrieger, but I was around when they were made. The music in .detuned is my work, and I’ve been with them when it was assembled. I also made the first demo with the .kkrieger engine after the game was out, so there’s that. I made two demos for Farbrausch: one was Theta, the other was Masagin [see below].

DET: And what kind of transition does a scener make into the world of actual game development?

LR: The guys ran a game company in Hamburg, 49 Games, that I ended up working for. It did mostly sports-related games for a German TV channel, RTL. Since then I think it got acquired by Bigpoint, and now they’re doing something else. It was weird: all our games felt like they were trying to reduce player interaction to a minimum. The rationale behind that was that the less freedom the player has, the less we spend on testing and tracking corner cases – almost like there was a suffocating risk aversion at place. I used to joke that if the devs could get their way, the games would just be like the demos.

The big advantage of the demoscene is that you really get to learn the craft of 3D programming there, and what kinds of tricks there are to make things look good, and have algorithms do the work. Every demo is a blank slate.

DET: Would you say you have a can-do attitude? That nothing is impossible?

LR: It’s definitely pointless to claim something is impossible before you have sufficient proof that it is. You’ve got to earn that right to claim something is impossible. I often tend to think that I can’t get what I want, and in many cases that turned out to be wrong, much to our surprise. On the other hand we both try to be very careful about wanting what we think is humanly possible.

For example, the thing about the Nowherians having a life on their own: that originated from a concept from around 2007 – only back then the idea was to populate a small village. You would jump into different people on the same day, learn about their hopes and dreams, and try to make their day a little better. On the same day there was an election, so the change in mood would be reflected in the voting booth. The grand arc was that you would try to get the people away from a destructive vote towards a constructive vote. But the detail was about doing a lot of different jobs as a lot of different people, getting to know characters, and having a way in which you could feel the consequences of your actions through yourself. To turn the recklessness that players exhibit in games like GTA against them.

The issue with that was the sheer work involved with building an awful lot of set pieces. We threw the idea around for a week and then I shelved it.

DET: An ‘abstract playground for human experiences’ could describe both Nowhere and Eskil Steenberg’s Love, the latter being somewhat impenetrable due to its interface and lack of tutorials. How accessible are you making Nowhere?

LR: Yeah. I played Love. It’s frustrating that the game doesn’t really care that you get it. The little bit of tutorial he added are full screen text pages that are rather difficult to read. But several people told me the game is great once you get over that. My experience with it was that I felt like just walking around and looking at stuff, but there were already enemies shooting at me, so suddenly I was mostly busy with running away, and that was kind of frustrating.

A long while back I read a fun article where someone reviewed life as if it were a video game. It’s of course the usual stuff: the graphics are great, and you can do everything but most of the time you have no idea what you should do next. He complained that the tutorial phase was a tad long, but at least you get two experienced players from the beginning who walk you through the whole process. And that’s actually not a bad idea. Children are sheltered right from the beginning. They can explore without caring much about dangers because their parents take care of that. And it’s really up to them to decide when to leave home and grow up, which means leaving the tutorial phase.

It’s not necessary to codify these stages explicitly so that you know whenever you’re going from one stage to the next. But it is important to have these kind of facilities you can turn to, and to start off with a sense of being protected, which is what gives you the confidence to go exploring. And I would think of that as the least intrusive version of a tutorial.

DET: Do you have to be mindful of the tools you give your players, and any familiarities of the interface, when player behaviour is so important? Are you trying to avoid them lapsing into certain habits?

LR: Habits are welcome, they inform the AI. In fact, that was one of the starting points for the AI concept, which is that we’re not as impulsive as we think. We always ease into habits, and when we’re thrown out of our element we’re usually pretty bad at coping until that new one has turned into a habitual cycle. You can build Markov chains from repetitive behaviour that happens in a certain order, and from that make predictions for what action is most likely to happen after the one before. Markov chains tend to synthesise results that are somewhat understandable, and involuntarily humorous, but always interesting.

DET: How about the way the Nowherians express themselves? That must be as tricky as it is crucial to player engagement.

LR: Working out the interface to get the right kind of expressiveness and simplicity is really hard. On the one hand, you want to maximise your options; on the other hand, the interface has to make sense within the game’s world, and be more helpful than in the way. How it works in the game is that you’re growing your utilities or tools or limbs or organs on top of your body’s surface. The mapping of utility to key or button is part of the game interface itself; only very few inputs are fixed, everything else is up to your preference.

The AI is symmetrically designed: that is, it ideally does not have more or less abilities than you have; everything the AI can do, you should be able to do and vice versa. So, other Nowherians follow the same rules. You can learn something about a Nowherian just by looking at the kind of limbs or organs he has grown for himself. As all Nowherians are metamorphs, it’s an information society, not a resource society; you learn how to grow limbs by looking at what other Nowherians are doing. If you imitate their behaviour you can do the same thing. Once you know how to grow a certain limb, that knowledge stays with you.

Communication with AI is something I’m still not happy with. At the moment the concept is that you learn a pidgin language of 27 symbols which you can combine to form more abstract sentences. But there’s also a need for a more emotional communication, by expressing mood through sound and sight. The emotional, animalistic communication is what comes next, and the language system will appear later when it’s more clear how much it’s really needed.

DET: How do the two of you complement each other when working together? Do you challenge each other?

LR: There’s at least one loud argument per week. Originally we divided the work so that Sylvia would do concept art and handle the business end, and I would be responsible for design and programming. I’m feeling a little guilty at the moment because she’s given me so much art to chew on in the beginning, and I’m still busy nailing down the tech to make it all a reality. She’s currently waiting for me to get that part done so she can do more creative work again; can’t be more than another two weeks, I think, then we have a solid way to interactively build geometry.

I can be really dug into my work, so far that it’s nearly impossible to ask me anything while I’m involved with a problem. Without her the whole endeavour wouldn’t work, financially. She’s taking care that we meet our deadlines and keep delivering. We’re both already challenging ourselves pretty hard. So, our role to each other ended up being the one who says: ‘You’ve done enough. This is good enough.’

DET: Finally, then, what can backers expect to see of Nowhere over the coming months?

LR: We’re going to release the next alpha this month, I think in about two weeks. It’s going to feature the new particle system, and the sculpting/meshing system has been completely redone and is now fast and robust enough to have players interact with it. Hopefully we’re also going to have two or three limbs you can grow. There’s also a new crafting UI which is more conjuring than crafting, as you draw shapes on the surface to initiate processes.

Parallel to that there’s the Steam launch. Right after that, the next step is to continue working on phenotypes and bring the AI back into the game. One drawback of the open alphas is that the game has to look presentable, so I’m having a really hard time with just putting placeholders and white boxes everywhere. That means the gameplay waits for the artwork.

Also, in order for the AI to be able to do anything and not just roam around like cows, some of the lower level mechanics have to work already. The next one is the energy resource system: acquiring and storing energy and routing it across the surface. But the plan is – and we’re still in good time – to have all important basic themes of the game in there by December 2014. Then it should be just scaling and adding to the gameplay’s infrastructure.

Nowhere’s crowdfunding total currently sits at around 20 per cent of its goal, all backers receiving regular alpha builds and a Steam key when it’s available. The homepage is here, and the often fascinating developer blog lives here.

, , , .

11 Comments »

  1. PopeRatzo says:

    …the demand for public alpha builds

    Who are these people demanding alpha builds? Are you 100% sure that there is such a demand beyond a small group of indie developers and kickstarter hucksters?

    • SominiTheCommenter says:

      I don’t even want beta builds. I just want finished games, please.

    • Muzman says:

      Well, clearly in this case they are talking about the demand for public alpha builds coming from the people who demand them, yes.

  2. kwyjibo says:

    It’s been Greenlit, but when’s it going on Steam’s Early Access? I’m sure it’d get a lot more traffic than their current website.

  3. Geebs says:

    Proverb: He who is certain that only good things can come from modding has never been to “lovers lab”.

    • Kitsunin says:

      Meh. I mean it’s weird. Extremely weird. But I’d argue that it’s good that it exists. Just for the sake of it.

  4. Dinjoralo says:

    Wow, I had no idea Leonard was involved with Farbrausch. That certainly makes me more confident in NOWHERE.

  5. shagen454 says:

    Looks like DMT!

  6. Lobotomist says:

    This looks great.

    Finally the game that is not afraid to push the envelope. Abandon the troupes of recognizable (real world , elves , history , novels , comics … ) And go for pure abstract new world that is created to fit the gameplay , and not the other way around.

    I am really curious how will this pan out for them.

  7. Josh W says:

    That markov chain idea is really clever. If I’m right, they are going to build the AI from player behavior by tracking how your actions lead to one another. “You’ve done this, so you’re probably going to do this next.” The tricky bit of this is including perception into the behavior framework.

    One easy way to do it is taking a leaf from game theory, and working on the basis of “behaviour + received behaviour”. This creates a matrix not a list, which makes the resulting description to build the markov chain 4d not 2d. Depending on how granular they want to go, you could reduce this significantly by working in terms of “behaviour reciprocated or not”, but that might still not produce the right dynamics.

    I really like the idea of coming upon yourself from a previous life and going “right, I don’t want to offer to trade with him, because he thinks that’s a kind of attack”.

    These two ideas together suggest to me a kind of radar tagging tied to emotional logic, where you can select creatures in your environment that you want to interact with in some way, and then that unlocks the possibility of certain actions when in their vicinity. So you could tag a creature as aggressive, then run away from it, with some cues (I’m thinking space fighter radar cues, but there could be others) showing you how near it is and in what direction.

    This would have functionality for the player, in that it would allow you to learn controls in stages, not accidentally attack people you meant to just grab on to, and would allow you to keep track of people when you’re swinging around off of the walls of someone else. But it also would mean that when controlled by the computer, these creatures would be signalling their emotions before making the appropriate actions, and would be acting on the basis of internally defined but possibly temporary relationships with different creatures in their vicinity.

    I could see that growing well above “behaviour received and given” to be honest, if you consider how the AI building algorithm would be interpreting the player’s choices of who to interact with in the first place, but I imagine that what would be quite a large amount of data on the player end could be chopped up into a much smaller program when it’s turned into an AI.