3D acceleration kills the demoscene
category: offtopic [glöplog]
I love you too, Rasmus! ;-)
are you rasmus?
Direct3d accelerated photoshop killed the OpenGL Realtime Raytracing Amiga chiptune demoscene (which was of course more imaginative and stylish than the AGA doskpop ravedemo scene running under DemOS with Gravis Soundblaster Ultrasound).
Furthermore, happy newbies get more and more harsch critics by hillbilly f**kheads. And Kozmik's ass is sweet. Or something. Or something else. Or something completely different.
And now pick up random words from this post and start a new trolling thread.
Furthermore, happy newbies get more and more harsch critics by hillbilly f**kheads. And Kozmik's ass is sweet. Or something. Or something else. Or something completely different.
And now pick up random words from this post and start a new trolling thread.
a good coder knows you can achieve the same kind of thing now on geforce class cards in hardware, or in software. but the implementation and final result will be a bit different.
you can do per pixel effects in hardware, it just takes a lot of effort and tricks with texture combiners, render to texture, etc. you have to work your way around the problem using a method that is fastest to your system, even if it isnt the most obvious way to you. that takes a lot of skill.
a bad coder thinks you can only do some things in hardware and some things in software, and never try to work out a clever way around to exploit hardware's power to do the same things.
you can do per pixel effects in hardware, it just takes a lot of effort and tricks with texture combiners, render to texture, etc. you have to work your way around the problem using a method that is fastest to your system, even if it isnt the most obvious way to you. that takes a lot of skill.
a bad coder thinks you can only do some things in hardware and some things in software, and never try to work out a clever way around to exploit hardware's power to do the same things.
hmmm, time 4 my opinion :-) years earlier i coded a bit on the c64 doing things like fld's, rasterbars and all the fancy oldschool stuff (the 80's oh yeah:) my opinion: THIZ SIMPLY RULEZ! because code is also an art, i would never prefer hardware accelleration before a god software routine (even if it looks badder)... i think hardware-only demos are for the
"normal-user" who wants to see such effects, but they are nothing for an "experienced" demowatcher. but, thats the industrie... faster... more ram... i can remember the times
where coders were lucky about doing one or two sprites more... but today it is just lame coding... and at least: sorry 4 bad, bitchin n-glish :-) peace
"normal-user" who wants to see such effects, but they are nothing for an "experienced" demowatcher. but, thats the industrie... faster... more ram... i can remember the times
where coders were lucky about doing one or two sprites more... but today it is just lame coding... and at least: sorry 4 bad, bitchin n-glish :-) peace
zaphod: as an "experienced" scener ( I began around 1989 on atari, and counted the ticks taken by every common GFA basic instruction to get the very best of this language ( ie. 800dots in overscan without any line of assembly ^__^ ) ) I enjoy both stuffs
yes, code can be artistic, and I dare to think my twisted (coder) brain can appreciate the inner art lied within a higly optimized sprite routine, but I reserve the same respect to less brain blasting stuffs.
I consider the demoscene releases as a new form of Art.
SW, HW, oric, 256b, 20mb .... are just exercice of style and deserve the same consideration.
yes, code can be artistic, and I dare to think my twisted (coder) brain can appreciate the inner art lied within a higly optimized sprite routine, but I reserve the same respect to less brain blasting stuffs.
I consider the demoscene releases as a new form of Art.
SW, HW, oric, 256b, 20mb .... are just exercice of style and deserve the same consideration.
If your imagination (or ability) is limited to a few rasterizers and a couple of image processing tricks, then software rendering rules, and hardware is for lamers who cannot hide their incompetence. But this is what ignorants tend to believe nowdays...
If your imagination (or ability) is limited to a few rasterizers and a couple of image processing tricks, then software rendering rules, and hardware is for lamers who cannot hide their incompetence. But this is what ignorants tend to believe nowdays...
shut the fuck up with all your attitude and shit. inactivity is what kills the scene. SO GET YOUR FAT HANDS OFF YOUR DICK AND START RELEASING (DEMOS).
I have been following the scene since the late 80s, first on C64, then on Amiga, and finally PC.
And everytime there was a big 'shift', there were people doing new stuff on one side, and people saying how good the old stuff was. I wonder if in a while there will be people saying how good the old 3d accelerated demos were.
Personally, I found that all new things grow on me sooner or later. Copper tricks were nice, ultrasmooth one-frame flatshaded poly 3d was nice, ultrafast PC demos were nice, and ever since I got a GeForce2 card, I can see the beauty of hardware 3d demos aswell.
I agree, there are demos that purely play some 3d scenes with barely any effects at all.
But that's not all there is. I have found some really good demos aswell.
It's always been like that, there have been good releases and bad releases. And there's always been optimists and pessimists. Just when you think the scene is going to get boring and die, someone will do something new and refreshing. 64k intros and realtime raytracing seem to be interesting these days. And a while back, the Java scene was interesting imho.
If you think the scene sucks, I say you should try to look around more. There's so many different styles and disciplines, there's bound to be something that you like.
There always has been for me, even though they were different things at different times.
And everytime there was a big 'shift', there were people doing new stuff on one side, and people saying how good the old stuff was. I wonder if in a while there will be people saying how good the old 3d accelerated demos were.
Personally, I found that all new things grow on me sooner or later. Copper tricks were nice, ultrasmooth one-frame flatshaded poly 3d was nice, ultrafast PC demos were nice, and ever since I got a GeForce2 card, I can see the beauty of hardware 3d demos aswell.
I agree, there are demos that purely play some 3d scenes with barely any effects at all.
But that's not all there is. I have found some really good demos aswell.
It's always been like that, there have been good releases and bad releases. And there's always been optimists and pessimists. Just when you think the scene is going to get boring and die, someone will do something new and refreshing. 64k intros and realtime raytracing seem to be interesting these days. And a while back, the Java scene was interesting imho.
If you think the scene sucks, I say you should try to look around more. There's so many different styles and disciplines, there's bound to be something that you like.
There always has been for me, even though they were different things at different times.
Not exactly my line of words, macaw, but I'll drink to that.
i think we have to separate the words "dos" and "software". dos is after all a crappy os that we all abandoned years ago, however; software demos is really what gets my blood pumping again. yes; we find new "hacks" and "ways" to do software-lookalike-stuff in opengl/d3d but it just isn't the same. i love software demos (as long as they are good - a crappy demo is a crappy demo, regardless of "platform").
..and macaw is right; inactivity is what kills(killed) the scene, but luckily, that seems to be changing. thank <insert your religios highest power here> for that. :)
..and macaw is right; inactivity is what kills(killed) the scene, but luckily, that seems to be changing. thank <insert your religios highest power here> for that. :)
AHAHAHAH. trop merci.
What's the point? We've all been argumenting this for a long time, but saying demoscene should only do software rendering is stupid. Shouldn't the demoscene keep up with what technology is availible? Even the C64 and Amiga had hardware for drawing graphics, people didn't ignore them, instead they use(d) them to create wonderful demos. Personally, I still love a lot of software demos, but computers have changed. Why shouldn't we keep up with the equipment we've got?
Stop complaining, try to do something innovative instead. Use whatever you want do use. Just make sure it's good.
And what the hell is that "harware" that skarab is talking about all the time? ;)
Stop complaining, try to do something innovative instead. Use whatever you want do use. Just make sure it's good.
And what the hell is that "harware" that skarab is talking about all the time? ;)
(a bit late)
moT: I code on irix. :)
moT: I code on irix. :)
I see 3d-acceleration as opportunity. Ofcourse it is ease to make texture mapped cubes with HW, but it is allso easy to plot pixel in mode 13h. We should make some new nice effects with this extra power and not just use it to make 3d flyovers.
I want to know how are my demos, please click on my logo and watch Yeah! : I don't know what I must change in my design. (software dos demo)
The scene was always dead ;)
Well Skarab, that's the problem with DOS demos, I can't get it to work in my Windows...
Do I really want to bother to reboot into DOS? Nah, can't be arsed, I just know my PCI soundblaster won't work anyway etc, no good vesa mode etc...
Yes I still have my Pentium with DOS 6.22 and a 1 MB GUS around... it would probably work on that... but can I be arsed to hook it up? Nope.
You might want to move to Windows, and develop stuff that people can actually watch on modern hardware.
DOS really IS dead.
Software rendering is not, productions like eg. Live Evil and My World are still quite enjoyable.
But please, Windows.
Do I really want to bother to reboot into DOS? Nah, can't be arsed, I just know my PCI soundblaster won't work anyway etc, no good vesa mode etc...
Yes I still have my Pentium with DOS 6.22 and a 1 MB GUS around... it would probably work on that... but can I be arsed to hook it up? Nope.
You might want to move to Windows, and develop stuff that people can actually watch on modern hardware.
DOS really IS dead.
Software rendering is not, productions like eg. Live Evil and My World are still quite enjoyable.
But please, Windows.
quote: "I think 3D accelerated demos are only videos. 3D acceleration kills the demoscene, sure."
You think wrong.. 3D accelerated demos are not videos, they are realtime rendered demos.
3D acceleration does not kill demoscene. There's a whole new world to explore out there.
TIME kills demoscene, not 3d acceleration...
You think wrong.. 3D accelerated demos are not videos, they are realtime rendered demos.
3D acceleration does not kill demoscene. There's a whole new world to explore out there.
TIME kills demoscene, not 3d acceleration...
eheheh
I haven't read pouet.net's forums for some weeks (months?). And what do I see just when I'm passing by? Still the same good'ol' debate about HW/SW rendering.
Quite funny indeed. Do what you want to do. Enjoy yourself. And remember that demos are made by demomakers. Mainly for other demomakers to watch it, but not only. Demomakers will be able to judge the technical part, when the common mortals will just see some nice or ugly things moving on screen. And that's it. Accept it or not.
Then, beginners have the choice on what to do, there's no "perfect way" to follow. Everyone is different, so everyone should follow his own path through life. :)
Have a nice day :p
tuo/popsy^mandarine^skytech
I haven't read pouet.net's forums for some weeks (months?). And what do I see just when I'm passing by? Still the same good'ol' debate about HW/SW rendering.
Quite funny indeed. Do what you want to do. Enjoy yourself. And remember that demos are made by demomakers. Mainly for other demomakers to watch it, but not only. Demomakers will be able to judge the technical part, when the common mortals will just see some nice or ugly things moving on screen. And that's it. Accept it or not.
Then, beginners have the choice on what to do, there's no "perfect way" to follow. Everyone is different, so everyone should follow his own path through life. :)
Have a nice day :p
tuo/popsy^mandarine^skytech
Guys, back in the days of the Amiga, how many of you were saying that a demo was not a “proper” demo if it used the blitter or the copper? That it in some way lessened the “artistic” value of the demo? Or that using the blitter or copper resulted in a “hardware” demo as opposed to a “software” demo?
It’s important to realize that there is no such thing as a “software demo” or a “hardware demo”. When you use those terms, you create an arbitrary boundary saying that everything on one side is “software” and on the other side it “hardware”. Right now, the discussion presumes that the boundary is 3d acceleration. But it might just as well be use of the blitter. You could say that if you move data around with the CPU you are making a “software” demo, but if you use the blitter, you’re making a “hardware” demo. Using the same “logic” you could say that using a packed pixel format is “hardware”, using a resolution higher than 320x200 is “hardware”, using more than 256 colors is “hardware”, using one of the fancy new machine code instructions in a PIII or Athlon is “hardware” and so on.
Creating a demo has always been about pushing existing hardware beyond what people thought was possible at the time. Saying that it is more interesting to see a demo where some artificial limit has been imposed on the programmer is like saying that it is more interesting to see how fast a speed runner can run after you tie his shoelaces together. After the athlete’s performance, you can say “yes, his performance was great given that his shoelaces were tied together”. What’s the point in that?
The incredibly exiting thing about new technology is that it opens doors. It makes it possible to do things that you just couldn’t do before. Pushing existing hardware beyond its limits is always going to be challenging but design and novel ideas are becoming more and more important in demos and I think that is a good trend. That is, the content is becoming more important than the technology, just like the content of a movie is more important than which chemicals were used in developing the film.
Don’t tie the demo creators’ hands by telling them to limit their expression to a fit within an arbitrarily defined boundary of what they can and can not do.
Gurf
It’s important to realize that there is no such thing as a “software demo” or a “hardware demo”. When you use those terms, you create an arbitrary boundary saying that everything on one side is “software” and on the other side it “hardware”. Right now, the discussion presumes that the boundary is 3d acceleration. But it might just as well be use of the blitter. You could say that if you move data around with the CPU you are making a “software” demo, but if you use the blitter, you’re making a “hardware” demo. Using the same “logic” you could say that using a packed pixel format is “hardware”, using a resolution higher than 320x200 is “hardware”, using more than 256 colors is “hardware”, using one of the fancy new machine code instructions in a PIII or Athlon is “hardware” and so on.
Creating a demo has always been about pushing existing hardware beyond what people thought was possible at the time. Saying that it is more interesting to see a demo where some artificial limit has been imposed on the programmer is like saying that it is more interesting to see how fast a speed runner can run after you tie his shoelaces together. After the athlete’s performance, you can say “yes, his performance was great given that his shoelaces were tied together”. What’s the point in that?
The incredibly exiting thing about new technology is that it opens doors. It makes it possible to do things that you just couldn’t do before. Pushing existing hardware beyond its limits is always going to be challenging but design and novel ideas are becoming more and more important in demos and I think that is a good trend. That is, the content is becoming more important than the technology, just like the content of a movie is more important than which chemicals were used in developing the film.
Don’t tie the demo creators’ hands by telling them to limit their expression to a fit within an arbitrarily defined boundary of what they can and can not do.
Gurf
gurf,
I agree with you. However, I see a point in what Skarab is saying aswell.
Writing software effects can be very educational... How about doing 2d metaballs first, before you try 3d, for example?
And knowing the ins and outs of triangle rasterizers can help you when debugging your 3d engine, even if it's hardware-accelerated.
Writing a raytracer can be very educational aswell.
But I wouldn't go as far as to say that demos should be software-only, let alone demos for DOS.
I can appreciate a good fast software renderer though.
In fact, I'm in the process of writing a new software renderer myself (hey, not all platforms have 3d hardware, you know :).
So in short, software rendering has its uses, and can be enjoyable, but that does not mean that hardware-accelerated demos can not be enjoyable.
I enjoy both.
I agree with you. However, I see a point in what Skarab is saying aswell.
Writing software effects can be very educational... How about doing 2d metaballs first, before you try 3d, for example?
And knowing the ins and outs of triangle rasterizers can help you when debugging your 3d engine, even if it's hardware-accelerated.
Writing a raytracer can be very educational aswell.
But I wouldn't go as far as to say that demos should be software-only, let alone demos for DOS.
I can appreciate a good fast software renderer though.
In fact, I'm in the process of writing a new software renderer myself (hey, not all platforms have 3d hardware, you know :).
So in short, software rendering has its uses, and can be enjoyable, but that does not mean that hardware-accelerated demos can not be enjoyable.
I enjoy both.
scali, "2d metaballs" is the effect commonly known as "shadebobs" with a nifty palette and has nothing to do with the 3d metaballs algorithm at all. And I don't see what coding triangle rasterizers would help you with debugging a hardware 3d engine either - oh my god, it's tracing edges scanline by scanline with linear increments and calculating deltas, how incredibly hard and unique a task.
But then, I never understood what that sw rendering hype was about.
But then, I never understood what that sw rendering hype was about.
every fucking 2D effect can be immitated on the god damn fucking 3D hardware. so get your fucking head out of your fucking ass and start fucking coding instead of fucking bitching on this fucking forum.
anyone who can fucking count how many fucking times i wrote "fucking" in this fucking entry owes me a fucking beer at MS.
anyone who can fucking count how many fucking times i wrote "fucking" in this fucking entry owes me a fucking beer at MS.