To Unity or Not to Unity
category: general [glöplog]
@smash : I'm still interested in the answer to this question: why real-time?
Quote:
@smash : I'm still interested in the answer to this question: why real-time?
It sounds a bit like asking a poet, why come up with rhymes, and obsess over rhythms, if you could have just written prose instead.
soundy: hmm. "real-time" as end delivery is obviously pure dickwaving. it's still really hard to make a top-end real-time render, it's a big challenge, so there's still the competition factor.
however, "real-time" as a creation medium is different entirely. there's still a pretty big divide (although its shrinking) between how you work for real-time vs non-realtime, a very different workflow. real-time is a great way to work and the demoscene has been doing it for years longer than anyone else more or less.
however, "real-time" as a creation medium is different entirely. there's still a pretty big divide (although its shrinking) between how you work for real-time vs non-realtime, a very different workflow. real-time is a great way to work and the demoscene has been doing it for years longer than anyone else more or less.
Quote:
asking a poet, why come up with rhymes, and obsess over rhythms, if you could have just written prose instead.
A poet would tell you the end result sounds more beautiful, and the person who hears the poem can enjoy the "sound" of it.
@smash : that's exactly my point. If we're talking pure art, pure self expression, then real-time is way less relevant. If you think technical challenge is still a big part of a demo, then real-time (still for the moment) makes sense, and in that case, means that a demo without technical challenge misses something.
this thread sure has a lot of words in it.
Quote:
The fact that you used speech marks suggests that you do not necessarily believe in the objective reality of "sound". Hence, the objective reality of "real-time" is about as relevant to the discussion.A poet would tell you the end result sounds more beautiful, and the person who hears the poem can enjoy the "sound" of it.
You can just say, some people appreciate it more this way. I don't think you need more of an argument.
Real-time is a vanishing concept as computers are becoming more and more powerful (still talking high-end demos). Looks like the tech challenge is not a defining element for a demo, as some demosceners don't give a fuck, and some care.
I still do think that tech achievements count a lot, and not necessarily optimizations or real-time. for example, the idea behind https://www.pouet.net/prod.php?which=57446 to animate lines is just brilliant.
I still do think that tech achievements count a lot, and not necessarily optimizations or real-time. for example, the idea behind https://www.pouet.net/prod.php?which=57446 to animate lines is just brilliant.
Quote:
Real-time is a vanishing concept as computers are becoming more and more powerful
By "vanishing" you mean "not vanishing at all", since more and more things are realtime?
Quote:
I still do think that tech achievements count a lot, and not necessarily optimizations or real-time. for example, the idea behind https://www.pouet.net/prod.php?which=57446 to animate lines is just brilliant.
Can we just use the above as an example to how the scene has this really weirdly warped sense of what a "tech achievement" is?
Also: an example of language that is incomprehensible
the real-time singularity, where real-time is so fast, it's not real-time anymore
I love it when someone random on the internet gets to define the absolute truth of what constitutes "a tech challenge" and by inference who is following their gospel or not. Good 'ol internet; still got it.
Yes, we are all random people, and each of us is trying to discuss and give his opinion. Did you spot anyone claiming he had the truth?
I still don't care what tool or engine is used, if I looks and sounds good where is the problem? I think from a artist perspective, if you use a homebrew scene tool or something off the shelf makes no difference.
Still, Smash said some interesting things, so quoting some of them:
The golden age of GL_ONE/GL_ONE blending. I think Sunflower won, end of story :-D
Yeah, but this is not a bad thing.
Wise words.
The first rule about 4K coding is you don't talk about shadertoy! 4K looks impressive these days, but on the other side, I wish there were some hard spiky triangles. But that's just my taste. There are very creative people releasing things. Good "alone" coders can shine here ... and I think shader showdowns are a result of this.
And the rest of what Smash said :-D
But as a boring side note, what keeps me away from creating anything that somehow looks a bit like a 64K is the overload of new technology. Instead of finishing anything there is always a new toy to play with. It started with the moment shaders were invented. And then every week you had a new toy to play with. Cool that works, next toy please. I think many coders are lost in technology, instead of doing something creative, they play with the new tech they have. And then there is the API surface, you spend more time on the framework than on the real renderer.
There are so many papers on JCGT I have to cast in code - not that I will use the result ever. And the current trend of opening up anything distracts me even more ... look at this Graphics Studies Compilation and don't tell me you wouldn't try implementing some of the tricks they came up with. Then there comes Nanite, I bet we (coders) have an inefficient compute shader at hand to test it by ourself, do we? If not, take a leftover from a 10 year old test you made in rendering an occlusion buffer in compute :-)
Hats off to everyone releasing stuff, I really enjoy and love it. Thank you. The streams from the parties are always streamed here. And yes, even Spectrum scene drama if available - I'm always impressed what you pull off that rotting hardware. Yepp, your capacitators :-(
P.S. I. This is an objection from my little son about Excess: Can you stop making "Vulkan Hello World" and start releasing demos again? He loves Excess with a passion, you should see his YouTube playlist. But as new toys comes in, there is just a Vulkan proof-of-concept thing ... we demand Function Over Fame II and Hurtquake II.
P.S. II. Excess again :-D Really, no joke. Kusma, get your ass off the couch.
P.S. III. What I like is the abstraction id software is doing. Instead of some API wrappers and abstactions, it's still some entry points like RenderWorld(World* world); Specialized engine vs. allrounder ... opinions?
Much to much brabble from my side.
Still, Smash said some interesting things, so quoting some of them:
Quote:
meanwhile pc demos moved to gpu. gpus were initially very limited and only really capable of throwing out polys (with alpha blending), shaping the style of demos at the time and putting a lot of weight on the 3d. as they gained more programmability through the early 2000s post processing fx over 3d became the norm, and the balance slowly swung towards the coder again.
The golden age of GL_ONE/GL_ONE blending. I think Sunflower won, end of story :-D
Quote:
the scene aged. the age of the average scener has been rising since the late 90s, probably almost linearly :)
Yeah, but this is not a bad thing.
Quote:
so why unity/unreal/notch?
looking at the profile of people making demos with these platforms (and i can speak with authority on notch at least), the typical person is an artist who used to make demos 20 years ago. the reason they're attracted to the platform is that it enables them to make demos again - but without the reliance on dodgy tools or worse, no tools at all and having to give their graphics to a coder with the hope it'll make it in and look acceptable. these people want to make demos - the whole demo - not assets. they see these tools as a way to get back to their hobby, work in real-time and make demos rather than rendering animations, and be inspired again. in the case of unity/unreal you're seeing some artists who can also code/script (maybe learned during that last 20 years of working) too.
what you're not seeing is lots of instances of existing coders/groups ditching their own processes and moving to unity/unreal/notch. we're getting _new_ demos from this, demos which otherwise would not exist - because the people just wouldn't be making them if the process was less appealing.
my hope is that these new tools encourage more and more artists back into the game; that we see the results of the creativity and different kinds of output this can bring, and this actually evolves the scene as a whole. maybe it'll move demos back towards being made by coders and artists working together. perhaps it'll even help the scene to grow.
and then, one day, maybe we'll see a return to more of a balance: a shift away from the coder-centric world we're in now to a more accessible medium with a wider appeal.
Wise words.
Quote:
or we can just keep copying raymarched blobs from shader toy or whatever. :)
The first rule about 4K coding is you don't talk about shadertoy! 4K looks impressive these days, but on the other side, I wish there were some hard spiky triangles. But that's just my taste. There are very creative people releasing things. Good "alone" coders can shine here ... and I think shader showdowns are a result of this.
And the rest of what Smash said :-D
But as a boring side note, what keeps me away from creating anything that somehow looks a bit like a 64K is the overload of new technology. Instead of finishing anything there is always a new toy to play with. It started with the moment shaders were invented. And then every week you had a new toy to play with. Cool that works, next toy please. I think many coders are lost in technology, instead of doing something creative, they play with the new tech they have. And then there is the API surface, you spend more time on the framework than on the real renderer.
There are so many papers on JCGT I have to cast in code - not that I will use the result ever. And the current trend of opening up anything distracts me even more ... look at this Graphics Studies Compilation and don't tell me you wouldn't try implementing some of the tricks they came up with. Then there comes Nanite, I bet we (coders) have an inefficient compute shader at hand to test it by ourself, do we? If not, take a leftover from a 10 year old test you made in rendering an occlusion buffer in compute :-)
Hats off to everyone releasing stuff, I really enjoy and love it. Thank you. The streams from the parties are always streamed here. And yes, even Spectrum scene drama if available - I'm always impressed what you pull off that rotting hardware. Yepp, your capacitators :-(
P.S. I. This is an objection from my little son about Excess: Can you stop making "Vulkan Hello World" and start releasing demos again? He loves Excess with a passion, you should see his YouTube playlist. But as new toys comes in, there is just a Vulkan proof-of-concept thing ... we demand Function Over Fame II and Hurtquake II.
P.S. II. Excess again :-D Really, no joke. Kusma, get your ass off the couch.
P.S. III. What I like is the abstraction id software is doing. Instead of some API wrappers and abstactions, it's still some entry points like RenderWorld(World* world); Specialized engine vs. allrounder ... opinions?
Much to much brabble from my side.
i think the real question here is:
was the intro of "foundation" made with notch?! it sure looks like a bad ass fairlight demo (just not in realtime).
was the intro of "foundation" made with notch?! it sure looks like a bad ass fairlight demo (just not in realtime).
sorry i probably should post this too
https://www.youtube.com/watch?v=iWgZBFdRgCo
https://www.youtube.com/watch?v=iWgZBFdRgCo
i tried to read that long post... i guess get a Room* room
I don't know if this fits here....(didn't dare to open a new topic in fear of getting yelled at :O)
But maybe this is something interesting for you smarties in/out here
The Thirty Million Line Problem
But maybe this is something interesting for you smarties in/out here
The Thirty Million Line Problem
Should not this become residual already?
What is the RULE for RESIDUAL ¯\(°_o)/¯
let's feed the troll(s) 🔥
