Відео

And why not, it's widely considered as one of the most influential and best PC demos ever created

For the newbies here, what do you mean? I don't know about this computer, but I wonder how a 6502 can do so much, is that all prerendered? For reference, I'd expect something more like the GX4000 with that power, where drawing a single column of tiles per frame for scrolling, is already really an amazing performance 😅

It's not necessarily PRE-rendered, but it's a common homebrew technique in 8-bit computing to offload intensive computation to a peripheral coprocessor, leaving the main 6502 with nothing to do but copy data from the coprocessor to the output ports. Some feel this is cheating, other consider it a valid approach; it was in fact used during the 8-bit era, with 1984's Atari 2600 version of Pitfall 2 containing a custom chip designed to feed graphics and audio data as fast as it could be read.

@@florianb3935 In this case, the Commander X16 has the VERA chipset, which with the FX extensions is capable of acting like a primitive (1990s style) GPU. Thus, a lot of things which the original Second Reality had to do with the CPU, pixel by pixel, are offloaded to hardware, and the remainder of the work is within the capacity of an 8MHz 65C02.

I think it is custom to refer to/credit the original authors as a group, not individual members, only if they personally know them or had contact with them. Maybe during the part-credits the original individuals could still be credited.

@@jeffreyhullekes6866 yeah, the individual members are credited in the original, so would've been easy to do

The VERA recently got some "FX" extensions which allow certain types of drawing and bitmap manipulation to be done faster. Without them, this demo wouldn't have been as accurate to the PC original.

@@TheBeeshSpweesh yes, but my question still stands. Is the demo drawing directly to video memory or to some off-screen buffer and then blitting it to video mem? The CX16 was known for being very slow for writes to the Video mem because there was no direct access. So how are they drawing these effects?

​​​@darksword1 there are roughly 15 demo parts and they use all different kinds of techniques to create the (visual) effect needed. Asking if the demo uses an off-screen buffer cannot be answered (easily), since all these parts work differently. But for most of them: no. The x16 can write to VRAM as fast as you can towards RAM (linearly). It was never really the lack of DMA that slowed things down, more a lack of knowing/imagining how to do it without it. The FX extension has several clever helpers inside of VERA to speed up things quite a bit, but its still behind a 8-bit VERA port. However those helpers were made in such a way that you could still benefit from them greatly despite the port-like access to the VERA chip. And they are specifically aimed to speed up what would otherwise be cpu or bandwidth intensive tasks.

X16 has no DMA or blitter at all. This is just pure unrolled loop of memory port banging to the FX helper.

@@jeffreyhullekes6866 I still don't like the cumbersome port access but the demo is impressive, it's good for the X16 to have such a demo to show off it's capabilities.

That's right, I mean it's not even technically satanic since it's not inverted.. I guess they preferred to remove it so that it was "family friendly"..?? Maybe it could be because they inspired themselves from the C64 version which also has this symbol removed, but for technical reasons

Music feels so off though

It would be impossible to accomplish on something that is slower than a 486 PC, so they had to resort with Parallax scrolling.

It was also the last part added and there was time pressure to get the demo ready for release. There are several parts we could improve (a bit) but there is a point (after a year of work) you have to wrap it up. We are happy with the result: you can clearly see what kinds of things you can do with the x16. We hope it inspires others to create demos for this (or other) retro machines.

We'll have to see. The plasma effect is heavy on the CPU but light on RAM usage (the loading pause is preloading for the subsequent effect). I'm sure it can be improved upon with more precalc, and I look forward to seeing this demo outdone by something even more spectacular!