What does it mean for it to be a 64k demo. That all the code required to produce the video fits in 64k? Can they call external libraries? How do they talk to graphics hardware?
They need to produce an .exe below 64kb in size that can run on clean versions of Windows and display the demo. That 64kb includes everything (textures, models, music, code etc). I think most modern demos use DirectX.
The executable file may not exceed 65536 bytes. No external data files are allowed. Any and all external libraries available on the compo machine may be used, including OpenGL/DirectX. The full specs of the compo machine, including software configuration are announced beforehand. In general, you can pretty much expect a stock win7 machine with modern hardware.
In 64k competition, data such as textures, models, samples/synth, etc. are either precomputed or generated on the fly. Most groups have their own tools or use tools available in the community.
I am always a bit disappointed though that such creativity is always restricted to demo comps while we are stuck with relatively mundane, static visuals in games.
The creativity is in no way restricted to the demoscene - just more conspicuous. Creating games involves a different set of constraints, and there are plenty of examples of ingenuity & creativity to be found there.
In fact, many game programmers are also involved in the demoscene; For instance, the guys who created Max Payne were a part of the infamous Future Crew.
Well, it is harder in a game to indicate to the player when they should press the Rhinoceros button - sure, running man to WWI fighter plane is fairly intuitive, and the squiggly-line-to-polygon-to-point-cloud powerup basically speaks for itself, but finding some fairly obvious method of indicating this is the point in the game story/world where the rhino is called for can be more of a challenge.
Nor did you get to drive it in the demo - it's the scripted composition that makes it interesting. It works brilliantly as a demo, but I suspect the novelty of driving the morphing squigglemobile in the demoverse would wear off pretty fast.
And once you do start incorporating other elements like AI, level geometry, physics, etc to make things interesting, you'll no longer have the luxury of allocating all your computational resources for graphical eyecandy.
Having seen the demo scene from its inception in the late eighties I find it fascinating that after all these years we are close to saturating our analog senses with this digitally created content. 32bit pixels saturated color perception, compressed music maxed out hearing and we are now pretty close with form and structure.
There are still nuances that we can keep chasing but looking at the big picture we are pretty close to generating a comprehensive audio visual experience.
I love the demoscene, I wish it were a much bigger phenomenon than it is (especially outside of Northern Europe).
That being said, one of the constraints on really getting why a demo is impressive is being able to appreciate what's going on. It's like understanding wines or beers or fine cheeses. You need to have a large base knowledge going in to get why the winning demo is unbelievably amazing.
Most people, even very technically or artistically inclined people don't get it. They think it's an interesting art project or something. But really understanding why it's impressive compared to say http://www.youtube.com/watch?v=9oXRP5OM6tA requires a pretty deep understanding of the capabilities of modern computing.
Nonsense. I'm neither technically nor artistically top of the bill, I have no good clue about the "capabilities of modern computing", but still I've been an active demoscener (producing and consuming) for years. To say I don't understand why these things are impressive is, well, a bit weird.
For me, it's simply (and only) freeing yourself of the limits of <your favourite animation tool> and replacing that by the limits of your programming skill and your computer's raw power.
This differs per person, but in the winning Assembly demo shown at the OP, for instance, some people may find the sea-particle-system impressive for its technological prowess; i'm more impressed by how code, not keyframes, was used to choose where which wave flows.
This also implies that I can find the same awe in watching Processing sketches, even though those hardly ever push the hardware limits.
I've always thought of 'Codename Chinadoll' as an unheralded beauty.
It seemed unusual in 1999 and still stands out stylistically today.
It featured some really nice 2-d elements, nods to print media.
Its poor placement in competition, difficulty running on a wide variety of hardware, and general lack of direction was mostly due to the sad circumstance of its main coder dying before it was completed.
33 comments
[ 3.7 ms ] story [ 64.1 ms ] threadEdit: Second place is a 64k demo: http://www.youtube.com/watch?v=69Xjc7eklxE
http://www.assembly.org/summer11/compos/realtime/desktop-pla...
In 64k competition, data such as textures, models, samples/synth, etc. are either precomputed or generated on the fly. Most groups have their own tools or use tools available in the community.
I am always a bit disappointed though that such creativity is always restricted to demo comps while we are stuck with relatively mundane, static visuals in games.
In fact, many game programmers are also involved in the demoscene; For instance, the guys who created Max Payne were a part of the infamous Future Crew.
And once you do start incorporating other elements like AI, level geometry, physics, etc to make things interesting, you'll no longer have the luxury of allocating all your computational resources for graphical eyecandy.
Smash of Fairlight who wrote the 2nd placed demo is in the games industry (at Sony Europe IIRC): http://directtovideo.wordpress.com/about/
(And one can take shortcuts in games that cannot take in real simulations.)
http://www.youtube.com/watch?v=LIIBRr31DIU
Listen to those hard drives!
Netscape Navigator didn't come out until over a year later (December 1994) and IE 1.0 another year after that. Wow.
Does anyone know if there is a current good site that reviews/indexes good new mod/scene releases besides index sites like pouet.net or scene.org?
There are still nuances that we can keep chasing but looking at the big picture we are pretty close to generating a comprehensive audio visual experience.
Actually, most demos are written in C these days, I believe.
That being said, one of the constraints on really getting why a demo is impressive is being able to appreciate what's going on. It's like understanding wines or beers or fine cheeses. You need to have a large base knowledge going in to get why the winning demo is unbelievably amazing.
Most people, even very technically or artistically inclined people don't get it. They think it's an interesting art project or something. But really understanding why it's impressive compared to say http://www.youtube.com/watch?v=9oXRP5OM6tA requires a pretty deep understanding of the capabilities of modern computing.
For me, it's simply (and only) freeing yourself of the limits of <your favourite animation tool> and replacing that by the limits of your programming skill and your computer's raw power.
This differs per person, but in the winning Assembly demo shown at the OP, for instance, some people may find the sea-particle-system impressive for its technological prowess; i'm more impressed by how code, not keyframes, was used to choose where which wave flows.
This also implies that I can find the same awe in watching Processing sketches, even though those hardly ever push the hardware limits.
It seemed unusual in 1999 and still stands out stylistically today.
It featured some really nice 2-d elements, nods to print media.
Its poor placement in competition, difficulty running on a wide variety of hardware, and general lack of direction was mostly due to the sad circumstance of its main coder dying before it was completed.
http://www.youtube.com/watch?v=4q0xv2123Ic