Randomness in Demos - a reaction
Optimus/Dirty Minds^Nasty Bugs
While enjoying another night of HUGi reading, I stumbled upon a debate on randomness in demos. I think there were more discussions on this matter somewhere, but I really can't remember whether it was in pouet or in a diskmag. Since many thoughts are always surrounding my mind while reading articles, I decided to formulate these ones in a notepad before it's too late. Usually, the article idea that motivates me to plan for some writing, fades away if I don't write down exactly what I was feeling at the time I got the passion about it. That must be the reason of many subjects I once wanted to write about but don't know why would I like now. I decided to start writing and even finish this article after midnight, which is very unusual for me, 'cause I am usually not willing to write anything, even if I had the whole day in front of me. Fortunatelly, I persuaded myself that this would be fast, easy and small, which made me open the notepad exactly after finishing reading the article that motivated me.
I would first like to say that I have witnessed exactly the same situation stated in the previous article, the one of a nonscener wondering about realtime nature and nonrandomness of demos. It was a question from my big brother: "What's the meaning of realtime algorithms in demos, if it always looks the same?" He was also wondering why these demo engines are non-interactive. My smaller brother btw (out of subject again, but just for the record), dislikes that most of the demos have no concept but are just a bunch of random parts/effects show without a meaning. Coming back, it seems that us sceners haven't even wondered about this contradiction of our hobby. It is something common for us. But I have an answer to the main question. Why do demos have to look exactly the same at each run?
Personally, it would be scary to see demos transforming into anything else than what I have got used to watching, because each one of them is a unique piece of art, similar to a painting in an art gallery. It is designed to be viewed as the original author wants to at every run. And that's how each one of them remains in my mind. Uniqueness is the most important element of a demo, the one that makes it the way it is.
Concering interactivity, I have found out that demowatching is the most relaxing thing I can do with one single click in my computer. I only have to sit down then, and stare at the screen while listening to the music, but not controlling anything. (Even games are tiring sometimes ;P) Interactivity in demos is encouraged, only if optional. Usually allowed, in order to impress people by proving to them that the demo engines are truly realtime. Examples are in Nature Suxx, where by pressing C you can move around the raytraced world, Oneder by Oxyron on the C64 with the very impressive and controllable 3d dots and raycaster engine parts, or many CPC onescreeners which let you play with the parameters, change the sine curves, colors, etc. This complemental way of interactivity is great and I encourage such works, but without spoiling the unique, one and only flow of the original demo of course.
As for randomness. It's interesting. I have used easy randomness in my quickbasic demos (different sine trajectories of blobs for example), but it applies in my case, since they are not real demos rather than effect shows. There exists no design in them, that could be spoiled by randomness. The only demo of mine that looks exactly the same, is my CPC demo "A Step Beyond", but wouldn't be possible to make it more random even if I wanted to, not only because of the way I started writing the code, but also 'cause I am not aware of a good and easy way to have different random numbers after each run, through Z80 assembly.
I believe though that randomness is not a big threat to the uniqueness of a demo. At least I think so, depending on how the coder uses it. And that's the most interesting thing about it! Surely it's not when applied in the easy way, for thousands of particles in a fountain, etc., since it will still look the same as a whole. Medium work approaches are very interesting algorithmically, especially that swarm of planes in Paper (Oh.. do you also remember the swarm of spiders in Kasparov?), still not destroying the whole feeling. (Btw, some of the times I tried to run Paper, I did not get the swarm but only one plane. And also, when this is happening, a specific cool part with bump mapped plasma and letters doesn't appear either! Does this have to do with the randomness of this intro? Not showing the good scenes??? Strange..) I also have to say that the third approach of different camera paths sounds very cool! It will not destroy the whole feeling, if the different paths meet each other at specific points and the demo ends up in the same way. Some parts being unique, others varying. I believe that using randomness while succeeding to keep the uniqueness of a demo will be very surprising and clever. It's an interesting direction to move on.
We've forgot another very interesting and unique piece of randomness, Rand by Fuzzion. I have never seen anything like that before. It consists of a bunch of cool software effects and filters, blended all together at different combinations and colors, fast changing and synced with the music and the lyrics which remain the same. But as a whole it's always the same at the end. Randomness never destroyed the demo feeling. It's obvious, since there are a lot of effects changing (so at the end, you won't remember which was first and which came after) but some of the important elements remain the same (music, syncing, lyrics) engaging randmoness with uniqueness perfectly. I think this is the best example of demo randomness, and it looked pretty impressive into my eyes when I discovered its individuality from usual demos. I didn't understand the worth of this one at the beginning, but it came out later after rewatching it for many times. Oh, and it's even a 64kb intro! I want it as a screensaver!!! :)
It's true that randomness and interactivity go far beyond common demomaking, being even more interesting and harder technically. It's also a hard fact that the demoscene seems to be a limiting community in conclusion. Perhaps that's because we don't like that our demos will look like games or technical/commercial demos in the future. We love the fact that they will be always something unique that real life can't give us. This happens with me too, who prefers the features of today's demos, mainly because I've got used to and grown up with them! But surprisingly, cleverly used randomness in some demos and optional interactivity have never spoiled the demo feeling I got throughout these years. In fact, examples of both branches, like Rand or Nature Suxx, have made the demos much more unique in my eyes!
Correct me if I am wrong. But it's very interesting and I would like to hear more opinions on the subject.