This might be a very busy time in the project, preparing a build of Bientôt l’été for IndieCade. But it’s also an easy time. Finally I have a sense of what this game will be. And I’m just working to implement everything needed for that to become a reality.

I have a long but very concrete to do list. So I’m just going down the list. Easy! It feels comfortable to know what I am doing, for a change. Not that there’s no new issues to deal with all the time. But it’s different from not knowing what you’re making at all and doubting everything.

I don’t think I like prototyping very much. The uncertainty that comes with it is unpleasant. I know it’s important to have a healthy measure of doubt when you’re creating. But in truth, I much prefer to just make something and risk utter failure, than to break my head over how to do it right. But then again, I have been known to do things that I didn’t like doing too. I’m kind of obsessed with getting results. The idea that the journey would be more important than the destination, is not something I adhere to much in my work.

So, yes, I know what Bientôt l’été is going to be. It’s a strange project. On the one hand, it’s a very modest and quaint little art game. There’s not much going on, neither in terms of story or in terms of play. And yet, on the other, it’s very rich. There’s a single player part and a online dual player part. There’s lots of text, all voice acted by two actors, and subtitled in three languages. There’s beautiful music. There’s two fully developed 3D interactive animated characters. You can choose your avatar. I still want to make an iPad extension to the game. And there’s a whole solar system, of course.

This looks like a list of disparate elements. But it all hangs together very well, I find. The fact that some aspects of previously unfinished projects have snuck in probably helps my confidence. It feels like we’re building something that we have been wanting to make for a while. I look forward to playing it!

I got up way too early this morning. Couldn’t sleep any more. Kept thinking about the work on Bientôt l’été. It’s a bit of a crunch period at the moment as we’re trying to get a presentable work-in-progress version of the game ready for IndieCade‘s late deadline on Monday.

I’m not the only one working hard. Walter Hus is delivering new music tracks at rocket speeds. Laura Raines Smith is animating the dressed characters out of Texas, while Auriea is modelling their undressed version here. Our very first intern Daniel Hellweg is working on the café table at the moment. And Theresa Schlag is taking care of the architecture. And even outside of Tale of Tales, somewhere in Russia, Universe programmer Neodrop is studying a troubled system from Bientôt l’été in an attempt to improve the performance of their wonderful visual programming tool.

This busily buzzing little hive adds quite a bit on my plate. I need to communicate with collaborators and evaluate the assets that come in, in between adding the final essential features to the programming and making the whole thing run acceptably. But they’re delivering such good stuff that I’m happy to interrupt my work for it. And after each such occasion, the game looks and sounds a little bit better.

Bientôt l’été is starting to feel very nice. Once in a while, I get a glimpse of what it can do. I’m beginning to see it, to understand its potential. Like vaguely deciphering the silhouette in the mist of a person I would like to know better. But I don’t have time to really sit down with it. And when I try, there’s always this or the other thing that urgently needs fixing. Hopefully the build for IndieCade will be playable enough to evaluate (and perhaps do some much needed playtesting).

Coding is easy. If you know what you want, you just type it in and the program is done. At least that’s how I felt today when transcribing my visual Universe graphs to Javascript.

But I also realized that being creative is not something I can do in code. Typing in all the logic felt like locking in it all down into solid concrete that can never change again. As if the code, fluent while typing, instantly solidified into an arcane amalgamation of symbols that lose their meaning to humans as soon as they leave the fingers and become the exclusive domain of the machine.

This is a fine method when you know what you want.

But knowing what you want is a bad idea when it comes to interactive creation. When advertising uses the phrase “the limit is your imagination” they mean that the possibilities are endless and anything is possible. But in my experience, when dealing with interactive pieces, my imagination is a very real limitation. If you only create the videogames that you can imagine, you’re not going to get far. You need to get dirty. You need to program. You need to make the machine come alive and collaborate with it, play with it. Interactive things should be made in an interactive way.

And that just doesn’t work in code for me. I need visual tools. I need blocks to position, links to physically connect. I cannot play in a word processor. Copy paste is not my idea of creative interaction. When I see all of my work, when I see the relation between the logic and the game on the screen, the effect of the systems that are running, then I start seeing new combinations, I get ideas that go much further than my imagination.

Not just in the sense of “wacky ideas”, but in the much more useful sense of ideas that can actually be built. When you listen carefully to your game, it will tell you what it wants to be. It’s a bad idea to ignore that. And this sort of communication works so much better in a language I can comprehend.

Thanks to Antares Universe, I am programming again. Universe is a visual programming add-on for Unity3D. I have been using it virtually exclusively to program our prototypes and games since several months. Thanks to Universe, I have been able to get my hands dirty again with actually programming, with actually expressing ideas for interaction and procedures in working logic. As opposed to battling with the Javascript syntax checker and hurting my head trying to think of linear text as a way to represent non-linear logic.

When I’m programming in visually, all mistakes I run into are mistakes I have made myself. Logical mistakes, oversights, etc. As opposed to the constant struggle with logistics that is coding in script for me. Universe allows me to get my work done.

And more than that. The way in which a visual programming language lays out the logic, inspires me too be more creative. Looking at some dead text with odd characters is meaningless to me. But seeing the living graph light up while my programming is running gives me ideas.

Bientôt l’été has been entirely programmed visually. Until today.

I decided to have a closer look at the strange hiccups I was hearing in the wonderful music Walter Hus has composed for the game. And sure enough, my Universe graphs were the culprit. In Unity’s profiling tools I noticed regular spikes that were identified as something mysterious called Garbage Collection. When I turn off the graphs, the spiking stops. The average framerate of the game is fine, but the spikes in performance mess up the music.

When I transcribe the visual logic to scripts, the problem does not occur. So the fault is not with my programming. Once again, my creativity is being mangled by engineering flaws. I love computers, I adore what we can make them do. But they’re so sensitive, so weak. The slightest thing that happens out of the ordinary, throws them off. You have to be so terribly careful with them. Careful in a way that only obsessive engineers can be. How am I supposed to make art with this thing?

I have been working on the animation of the waves today. With not much success.

I had outlined the basic animation with primitive shapes with the idea to replace these by more recognizable forms later. That outline felt very good. The timing was right, the motion was right. It really felt like waves flowing over a beach.

But trying to also make this look like waves is proving more difficult than expected.

I like the cleverness of the abstract version, though. I find it amusing. I don’t really want to replace it by something that looks realistic. Realistic things are so easy to ignore. They’re hardly worth the effort. But what I have now can only be loved by its creator.

I’ll try again tomorrow.
I should probably look at the sea a bit more.

Being fond of the interactive medium seems to imply that I suck at directing. For me this medium is about exploration and curiosity. And when I work with other people, I’m curious how they will interpret their role. In a way, I’m happy to accept whatever they do, especially if it’s different from what I expected.

This makes me a lousy director. But it’s exactly how I approach the medium: I want to create circumstances without any expectations of how players will act in them. I am curious how they will respond. I want to see how people respond, to see if there are things in common in our responses, and to be surprised.

What was experienced as lack of control by my colleagues web designers back in the day, I always found liberating. I liked that users could change the size of the browser window to change the composition of the page. I actually actively designed for this in some cases.

The surprises that can happen during interaction are what interests me most about this medium. I don’t want to direct somebody’s experience. I want them to have their own experience. Even if it is different than mine, or what I envisioned as creator, it’s vastly preferable for me if the experience was personal. I think this is related to the importance of agency.

But this openness makes it very difficult for me to tell the people I collaborate with what to do. I’d rather adapt my design to their contribution than force them to do what I expected. So I’m very happy that composer Walter Hus took on the role of directing Christophe Poulain when recording the voice of the male character in Bientôt l’été. I loved everything the actor said, but Walter remained critical and made him repeat things over and over until it sounded right. Of course this was another way for me to let go. He just happened to have the same taste as I did. Not sure what I would have done if he didn’t.

I’m rather proud of the very simple system I have created for simulating the time of day on the beach in Bientôt l’été.

There’s a half dome that represents the sky. But it’s very small. There’s a special camera inside that rotates around like the main camera but remains in the middle otherwise. This camera is simply rendered first and then the rest of the game world is rendered, in real size, on top of that.

This sky dome has a texture with gradient colors: night to dawn to day to evening to night horizontally, and sky to ground vertically.

The UV mapping is straight on. To change the time of day in the sky, I simply move the texture horizontally. I chose to do this because I noticed that the sky never changes color uniformly. There is a distinct difference in color between the part of the sky where the sun is and the opposite part. This is especially noticeable during and after sunset and sunrise.

At the same rate of moving the texture on the sky dome, I change the ambient light color (which represents the color of shadows in a 3D engine) and the color of the fog to match the color of the bottom of the texture (ground). And I change the color of the light and of the bright disc that represents the sun to the top color of the texture (sky). The eight source colors for this are sampled from the texture on the sky dome directly on startup and then blended according to time of day.

A great source of inspiration, by the way, was this wonderful description of light effects by Richard Yot.

In keeping with the sparse aesthetic style that I chose, the sky has no other features than these gradient tones. There’s a sun and a moon, but no clouds. Given the choice of colors, it’s not entirely clear whether we are seeing the sky, or just a very uniform cloud cover.

Next to the motion of the texture that corresponds to the time of day, there is also a movement of a transparency map, in the perpendicular direction, that corresponds to the position of the character on the beach. When he or she walks far enough, the sky opens up to reveal a sight of the planets and the stars that surround the space station. But that is stuff for another post!

Today we have recorded the female voice for the multiplayer part of Bientôt l’été. In this part of the game, you sit down with another player at a café table. The objects you collected on the beach can be put on the table and the phrases you collected can be spoken. Each player puts or says something in turn. Like a game of chess.

Over 200 phrases were selected from multiple novels by Marguerite Duras. Some are very poetic, but many are plain. For the tone we follow Duras’ own style of reading from her work, as she did in many of her films: very dry and neutral. This fits perfectly with her writing style. On the surface the words may seem banal, and sometimes indifferent or cruel, but that only heightens the sense of extreme passion that the characters are going through. At least that’s the effect this has on me.

There’s two characters in the game, a man and a woman. Both have the same phrases at their disposal. For the female voice, we were as fortunate as to find an even bigger Duras fan that I in Mademoiselle Fabienne Mésenge, a multi-talented actress/director herself.

She clearly enjoyed performing Duras’ words, surrounded by the organ pipes in the studio of Bientôt l’été composer Walter Hus in Brussels. The recording went much more smoothly than anticipated. But the second session, after un verre de vin was clearly the better one!

It was very moving to me to hear the text. I can’t wait to record the male voice and put it all in the game and then play!