pouët.net

well, this was an interesting project to work on. i was incredibly scared about doing this because the expectation level for a 64k nowadays is so stupidly high that we knew from the start it was going to be basically impossible to achieve it. we held off from releasing at bp because it wasnt quite there yet, and when assembly came round we couldnt do that again because it would probably have never came out. i was still crapping myself about it though. and i can completely understand why not many people are making them - the expectation might even be higher than for demos, i dont know.

one of the most fun things about it was keeping it a secret. this got quite difficult when e.g. i had to get the organisers to cut it off the list of productions they gave to meaty for asm tv.. :)

the other thing was that we decided to try and be just a little bit pretentious (with the nfo file etc, not even talking about the scenes, such as a plane with an explosion behind them - all that's missing is the bon jovi track i had in my head at the time) and see if we could get away with it. :) i think that those who know us personally won't take the posturing too seriously. :)

anyway, i hope that at least a couple of groups have now received some foot-based motivation to make something new. im looking forward to them beating us so we have an excuse to try again. :)

if you have technical issues: the shaders here are really heavy and there's shitloads of them to compile (~400 or something crazy like that), which a) explains much of the precalc time and b) why you need a good gfxcard to get good performance at a high res. 7xxx series nvidia wont cut it anymore really. well, the hardware was there, and it ran so fast on the compomachine it actually broke the water effects, so it does at least scale. :) it should work fine on ati and nvidia, if you have problems let me know.

syphus: did you select a res your monitor can handle? (the setup box isnt smart enough to stop you from selecting an invalid one)

some virus scanners incorrectly believe that any exe packed with kkrunchy must be a virus.
feel free to ignore it, or mail fsecure and complain. :)

zoom: the shader "mystery" is a pretty obvious one: combinations of shader fragments. so, youve got a shader for 0 lights, 1 light, 2 lights, 3 lights at a time; then youve got combinations for textured, untextured, normalmapped, parallaxmapped; for specular or not; for ssao or not; for water rendering, for underwater rendering; for z passes, shadow passes etc; and for other effects too. it's much the same way that half life gets thousands of shaders. it's that, or extremely inefficient shaders with a lot of branches, and halving the framerate. :)

the nfo file as a whole should be taken as a small homage (pisstake) on the tremendous boasting in some other group's infofiles and scrolltexts. :) obviously writing "400 shaders" is about as sensible as saying "we had 131294 triangles in 490 objects" or "300,000:1 compression" or something - it's clearly a pointless number. in fact it makes more sense when explaining the precalc time than it does when talking about amounts of data.

but that said, some of you guys who do know what's possible but haven't done a 64k in a few years, when fixed function was still going strong, might be surprised to learn that the shaders are the biggest single (yes,compressed) data block in the 64k, bigger than the meshes, samples, music data, texture data etc. modern rendering apparently doesnt come for free. so dont forget to factor it in when making that killer comeback 64k! ;)

soundtrack link added.

also, thank you all for the generous comments, much appreciated. it was hard work.

i just noticed the water in the dtv+capped captures is really broken. does anyone know what they were done on?

actually, lazarus, that's not the source for the rap. maximum respect to timbaland for biting rakim's famous line in not just one, but two songs.