Written by Saffron
Having read Ryg's elaborate and thoughtful article "Good Night, Scene" in Hugi #16, I felt compelled to write down some of my own thoughts on this matter. Although it's a rather useless thing to mention, this happens to be the first diskmag article I've ever written, so there...
To start with a positive counterpoint, I don't agree with Ryg's view that the scene would have stopped developing some years ago. Regardless of the absolute quality of present day demo coding, there's more to demos than code, and IMO this is where important progress has been taking place. Originally, code was what mattered about demos and the rest was just an afterthought, but since the first pastel-colored Amiga design demos the state of things has been changing, and this development has picked up pace in the past years.
Because of this, it seems to me that there's actually significantly more variety in demos today than three years ago. Maybe the most visible example of this is the Danish style that Purple and Blasphemy introduced last year (building on a solid foundation laid by Sonik's Amiga prods, I'd say). Effects such as solid vector 2d and wireframe/flat 3d that might be unimpressive by themselves are combined with subtle graphics and other effects, creating a coherent visual style that mirrors some movements in contemporary graphic design and could be described as minimalistic except that it still provides plenty of eyecandy. The same de-emphasization of code has affected many other demos and groups as well - Pulse's 73 Ms for example retains the traditional effect-oriented demo styling, yet it's really the graphics that are the centerpiece in this great demo, and the code serves them rather than the other way around.
These are not the only examples of code moving into the background. There now is enough CPU and graphics hardware power to do layering in demos with blending modes and real alpha channels. This kind of code is hardly glamorous, but it opens a whole new world for demo design, one much closer to the "real world" of video effect and compositing work.
Thanks to this, demos are gaining a more complex look than ever - I'd even say that there's a revolution of demo visuals going on that's the biggest one in demoscene history so far. However, looking at the bigger picture, things start looking much grimmer. The visual revolution seems to be touching only the top groups for the most part, while the rest linger in a sea of mediocrity that can appear truly depressing at times.
Here, I'll have to agree with Ryg: on average, the quality of coding has decreased steeply over the past years. Boring and ugly 3d scene flythroughs have become a notorious tradition of today's demo releases and more and more CPU power is being used for rendering them, yet the quality (of both the modelling and rendering) has pretty much remained on the same low level as some three years ago. This is especially ridiculous in the current age of 3d hardware acceleration, as the rendering and geometry power available is probably quadrupling every year, and the (game) coders who are making use of this technology are reaching quality levels that start to approach the output of packages like 3D Studio.
A paradoxical situation for sure. Demo coders were very late to the 3d hardware rendering game, and consequently the 3d quality gap between games and demos became huge compared to what it used to be. This has led to the perception that a different standard applies to games and demos, and so the situation is feeding itself, as the standard that new coders try to attain is set by the quality of the average demo rather than the state of the entire industry.
Some people even think that this is how it should be; they believe that no demo coder can match the quality of today's games created by large professional teams, and therefore the demoscene should stray away from all competition with the game industry, becoming its own isolated pocket of the computing world. This kind of thinking is the complete opposite of what the scene used to be about: showing what can be done on a computer when the restraints of commercialism are removed.
By this definition, the PC demoscene must keep up with the current standard of PC hardware in order to provide the masses with demos that use their computers to the fullest. This also places the scene in direct competition with the game industry, which has always been a significant force in pushing the computer technology envelope, but is now even more so in this age of extremely technology-oriented gaming (does anyone think Unreal's sales numbers had anything to do with its gameplay qualities?).
This is however just the theory and provides no solution to the dilemma of how the scene is actually supposed to compete. I think the answer lies in the definition I gave above: the non-commercial nature of the scene. The fact that demos don't have to deal with interactivity seems to provide the scene some essential breathing room. Game developers are stuck to anticipating and dealing with the player's impulses and all worst-case scenarios of what he might do, whereas a demo has a fixed environment in comparison.
This should empower demo makers to do things that simply might not be practical or even possible in a game environment. For example, I'm fairly positive that with modern 3d hardware there are plenty of things that would be possible but are simply of no use for a game. Seems like this kind of "loopholes" could be compared in a sense to what demo coders were doing on the earliest demo platforms, especially the C64: finding ways of doing things that should be "impossible" (or in other words, something that no one has had a reason to do before).
The difference is that a 9-pixel scroller that runs in an outer screen border could only impress people who were very familiar with the details of the machine, whereas the things that a 3d accelerator might be wrought into doing could be very appealing to anyone who's seen a few modern games and is under the impression that that's all the visuals a computer can do. This would potentially create an entirely new demo audience and in turn would expand and refresh the demoscene with new members, ones that actually contribute and not just hang around.
Ok, I'm not a coder, so you're free to bash this graphician's little democoding utopia. And certainly I don't think all this would automatically happen - on the contrary, it looks improbable with the way things are going, despite the good results that some top groups are reaching. The combination of low standards and a self-satisfied atmosphere that Ryg was so pissed with really appears to be the major obstacle here. (To be fair, this is not just a coders' problem - too many graphicians are nowadays doing those notoriously mediocre pics of women's heads and don't show any signs of even wanting to improve... Content was never any strong point of scene graphicians, but in the pixelling times we at least had superiority of technique and style; now even that seems to be disappearing as so many people are satisfied with blurry, oversaturated Photoshop redraws that have absolutely nothing personal about them.)
Ryg provided a solution with his nine points of advice, and I'll just have to wholeheartedly agree with them as there's really nothing I can add. Read them again, and then take a good look at what you've done and what you've tried to do. I'm certainly not happy with what I see: there's too much sloppy work I've released, too many things I've failed to really think about, too many times I've taken the easy way instead of trying to be better than last time. Keep Ryg's advice in mind, and maybe that scene renaissance will happen after all.
More discussion on these issues would be welcome. So think, write to Hugi, and save the world...
Saffron / TBL