pouët.net

Dogma demomaking

category: general [glöplog]
Ok Krill, if you think you depend on a PC to do on a PC what you think is cool on a C64, then choose level 1. Or don't. :-)
added on the 2024-05-31 16:09:04 by bifat bifat
I don't hugely see why making the logistics difficult again is going to make much of a difference. What really changed between then and now was a gradual progression towards unified aesthetics across disciplines and coherence in the finished work.

If you are going for some sort of Dogma 95 for demos, as wysiwtf pointed out, surely it'd be a return to the '86-'90 "pre-design/extracted graphics/effects over visual clarity" styling and attitude you'd be looking for? Taking that as a focal point without the usual nostalgia angle would be a different take certainly.
added on the 2024-05-31 16:09:34 by 4mat 4mat
as far as i'm concerned if you posses any information beyond the year 1989 in your brain you might as well not bother with any of the proposed methods since that's where the real advantage lies—experience and knowledge of a wider context—not in artificial asceticism that just inconveniences and slows you down.

just make demos in the Here and Now and instead find other non-artificial ways to challenge yourself, preferably conceptually, since that's what you likely aspire to do anyway.
added on the 2024-05-31 16:20:50 by noby noby
Quote:
Ok Krill, if you think you depend on a PC to do on a PC what you think is cool on a C64, then choose level 1. Or don't. :-)
A (big) Amiga would work as a PC substitute as well in this case, i guess. It still remains a fact that you can come up with things that are at least impractical (if not impossible) to create on the target machine, yet easy to run on it once created. =)

And i was somewhere between dogma levels 1 and 2 (i guess) until 2002 or so. Never looked back after finally fully committing to cross-devving, of course with the actual computer as the ultimate target and regular test machine. (No emulator is good enough to warrant not testing on real machine.)
added on the 2024-05-31 16:30:16 by Krill Krill
4mat: I speculate that adopting such tenets would react in some way on how things would look like. Maybe it even allows for finding new ways of expression. For me it's not just technical, and it's certainly not nostalgic. I'm not entirely happy with how demos look and sound nowadays, they often look very slick and consistent, maybe a bit too slick and consistent for my taste.
But that's just my personal background and should not influence anybody's motivation for adopting such rules.

noby:
Quote:
as far as i'm concerned if you posses any information beyond the year 1989 in your brain you might as well not bother with any of the proposed methods since that's where the real advantage lies—experience and knowledge of a wider context—not in artificial asceticism that just inconveniences and slows you down.

I see your point, but it's also a wild guess, maybe it's not so ascetic and inconvenient at all. More like selecting some unconventional tools for this time and age, which need to be mastered first, like everything else.
added on the 2024-05-31 16:39:44 by bifat bifat
Quote:
For this C-64 4K demo i packed some tables (sine, arctan and others) lossily to a bunch of just a few floats (Nth-degree polynomials, with N=3 to 5 or so), for those floats to be run through the POLYX ROM function by the C-64 at demo init-time, to recreate the curves.

Sine and arctan are bad examples, since trigonometric functions have simple power series which don't require any "fitting" at all if you want to approximate them with polynomials. You don't need a powerful PC or a C-64, you just need a pen and some paper if anything.

Which DOGMA LEVEL would that be?
added on the 2024-05-31 16:54:06 by fizzer fizzer
Quote:
Sine and arctan are bad examples, since trigonometric functions have simple power series which don't require any "fitting" at all if you want to approximate them with polynomials. You don't need a powerful PC or a C-64, you just need a pen and some paper if anything.
The point was to pack them efficiently (albeit lossily) for a 4K demo, and have integer (fixed-point) variants of them in unrolled state. To generate them wasn't the computationally-hard part, but to polyfit the curves (and tweak the parameters until the output was good enough).

Also, please read carefully before ranting, and refrain from implying that i'm an idiot. Also the source is here.
added on the 2024-05-31 17:01:36 by Krill Krill
Why not just make better demos?

It's not like demos nowadays are uniformly extremely badass, so it seems to me like most sceners need all the (technical) help they can get.
added on the 2024-05-31 17:04:12 by sagacity sagacity
Fast iteration enables content that would not otherwise be feasible. It's not just a matter of how long it takes; being able to see the result of edits within a fraction of a second enables a different kind of creative thought process.

My favorite way of making demos is to create a custom tool or programming language along with a simulator of the demo effect (running on a PC) in order to allow such fast iteration. While you could theoretically create such tools for the real hardware, the CPU would typically be way too slow to enable the desired iteration times.

Examples of Amiga productions made this way are Ikadalawampu, My Lucky Number, and the productions using Rose.
added on the 2024-05-31 17:14:03 by Blueberry Blueberry
Quote:
Also, please read carefully before ranting, and refrain from implying that i'm an idiot.

That was unnecessary...
added on the 2024-05-31 17:15:18 by fizzer fizzer
Quote:
That was unnecessary...
You see, it wasn't just sine or arctan.

And the general computational asymmetry point applies to anything codec (which includes compression), with the "dec" part being a lot cheaper. And other things besides codec.

Also what Blueberry said.
added on the 2024-05-31 17:29:16 by Krill Krill
Quote:
Why not just make better demos?


'cos then he'd need to switch to PC ;)

Joke (but no joke) aside, I don't mind this whole idea, I think it's entertaining and the unenforceable nature of it will probably ensure it will have little to no effect on the long run; ultimately the biggest factor in making better prods anyway is what you know, and you can't really limit the use of knowledge.
added on the 2024-05-31 17:31:08 by Gargaj Gargaj
Quote:
You see, it wasn't just sine or arctan.

No-one said that it was. I pointed to those specifically for a good reason, but that's no reason to feel like someone is calling you an idiot.
added on the 2024-05-31 17:42:32 by fizzer fizzer
Gargaj, you can talk to me in second person if you wish. So let's discuss PCs and why making demos for PC would be an improvement over making demos for Amiga, for me. Please explain.
added on the 2024-05-31 18:12:00 by bifat bifat
I only parsed through the thread superficially but the thought occured to me: how would you apply the same concepts to say pc 64k/4k intros? What would be the levels?
added on the 2024-05-31 18:53:17 by BoyC BoyC
BoyC: First, make the whole demo on Amiga 500. Then port it to PC.

That would be Dogma Level 3.5 at least.

Just kidding, I guess. :]
added on the 2024-05-31 18:57:27 by ham ham
While sometimes trying to work on the target machine is fun, and for some platforms it's very doable, let's not kid ourselves that this is the "true" or "traditional" way. It's the "sad penniless teengager who can't affort a more powerful development machine" way.

Since the dawn of time, people used more powerful computers for development. Zork was written on big UNIX computers. Lucasfilm games in the 80s had a development setup with big UNIX machines, and custom interface hardware to cross-compile and immediately run on commodore64. Burger becky used a super-upgraded Apple2gs to write for commodore, again with custom interface hardware. John Carmack wrote doom on NeXT workstations.
Nobody who had the option stuck with underpowered personal computers for development.

But we can do it Dogme95 way. Let's get together, decide on the rules, and all but one of us will disregard them immediately :)
added on the 2024-05-31 20:14:13 by Nuclear Nuclear
Quote:
Gargaj, you can talk to me in second person if you wish. So let's discuss PCs and why making demos for PC would be an improvement over making demos for Amiga, for me. Please explain.


Why would it? I thought the whole point of making demos was to show off the versatility of the LIMITED platforms of the time, especially in the 1980s with the Amiga, ST/E and 8-bits and early 1990s with 32-bit Amiga AGA/MS-DOS as the best being just as limited, and with the frenetic competition between groups pushing for better and more innovative effects.

Ever since Windows became dominant in league with Pentium processing power, not to mention graphics cards, multi-core CPUs and now realtime ray-tracing with the GeForce RTX series, PCs are so powerful now they can achieve ANY effects you want, in ANY scope, at ANY resolution at ANY framerate. So what is there to push for? You may like this aspect, but I personally don't. I personally find these high-end PC demos very BORING.

My favourite demos are the ones done at fixed-limited resolutions, fixed limits in colour palettes, fixed hardware like scrolling and sprites and special colour modes like HAM or the ST Shifter, and just that extra talented OOMPH to push the limited hardware of the time to perform interesting effects that they weren't designed for. THAT'S what I love.

But that's just my opinion, and I'm stating it here and now. Maybe you should do the same? :)

Here's a final thought, and I've long thought this: We are stating our opinions, agreed? And I've never figured why someone show say something and HAVE to state that it's their opinion in their own post! It should go without saying, right? "Every person's personal opinion is in their own posted post."
added on the 2024-05-31 20:17:41 by Foebane72 Foebane72
Yeah, let's all do it like in 1991. ;)
added on the 2024-05-31 20:20:07 by Blueberry Blueberry
Quote:
PCs are so powerful now they can achieve ANY effects you want, in ANY scope, at ANY resolution at ANY framerate

that's so clueless, i want that on a tshirt!
Quote:
Quote:
PCs are so powerful now they can achieve ANY effects you want, in ANY scope, at ANY resolution at ANY framerate

that's so clueless, i want that on a tshirt!


OK, if I'm wrong, correct me :)
added on the 2024-05-31 20:30:17 by Foebane72 Foebane72
Quote:
OK, if I'm wrong, correct me :)


I'm thinking in RELATIVE terms, mind ;)
added on the 2024-05-31 20:32:46 by Foebane72 Foebane72
What Nuclear said. I was actually thinking of making the exact same point.

I'm reminded of the way the CGI in Babylon 5 was famously created using Lightwave. That's a great way to sell Lightwave for people wanting to make 3D on their (one) Amiga, but it may not have been obvious at first back then that the CGI was rendered by a network of Amigas which the typical home user could absolutely not afford. So were the people making the visual effects cheating? What they were doing was totally contemporary with early 90s demoscene on Amiga. Is there a difference between using just the right computing power equivalent to 10 Amigas to generate lookup tables and doing it with 10 actual Amigas?

On that note, where does the boundary lie for hardware modifications? Is a GoTek out of the question?
added on the 2024-05-31 20:34:57 by fizzer fizzer
Also yeah I want that t-shirt too.
added on the 2024-05-31 20:35:33 by fizzer fizzer
Fizzer: Gotek and trackmos are an interesting topic in themselves. I have thrown the Goteks out of the window after I have learned that coding trackloaded things against a Gotek is a terrible idea. You will inevitably produce a trackloader that works nicely - on a Gotek - but is a delicate bitch everywhere else. :-)
For making trackmos on what I call level 2 you'd have to walk a few extra miles, but not terribly much. You just need a testing harness in which you can run your trackmo parts reliably over network cable. Later you just bundle the parts together on disk, and then come the extra things that overstep the formalisms of the framework, like loading and decrunching something TWO parts ahead of a part (as an example)... for Hologon I wrote out the disk not more than about 100 times, which isn't that terrible given the size of the project. I was in the test harness at least 95% of the time, and I used a system for fast-forwarding the music to the respective position in the demo. This required a lot of discipline for bookkeeping, but not much more than that. It was all written down in config files. Thanks to Krill for the fast-forwarding and a few other suggestions!
So using Gotek: Poor you! Do it, you might learn a thing or two along the way. :-)
Regarding rules and enforcement: You bend them and interpret them as you see fit, and then you have something to talk about if somebody asks. That's the idea.
I wish I have had the Dogma idea a few years earlier, because then I could have perhaps made a few demos on higher levels already. For the extra fun! The difference in effort was very small sometimes.
And then I learned that for example Moods Plateau do their demos on real machines, and I can tell you I have a hell of respect for that. They are on a higher level, so to speak.
added on the 2024-05-31 20:57:52 by bifat bifat

login