pouët.net

BPL information 5 glöps

• general:
• level: user
• personal:
• first name: B.
• last name: P.

• 64k JavaScript Demise of the Scene by Anksai Collective [web]
• Motivational piggie, because this looks like a newcomer, and honest, constructive feedback is how people get better. Everyone starts somewhere: 30 years ago that meant copying existing asm and spending afternoon after afternoon working out what every line did.
  
  But then there's the statement: "a conceptual excavation of what technological mediation has hollowed out... its dissonance is not a flaw but the argument". That's a lot to put on this. If the idea is to shake up the scene, it can't be done with a hello world, and that's what this is, plain and simple. It proves nothing, demonstrates nothing and doesn't push anything.
  
  Looking at what's actually in there:
  Visuals: the scene rendered twice (once mirrored for the water reflection) into an MSAA target, ray-marched SDF boxes, the rest boxes, quads and a heightfield valley, a fullscreen sky gradient, additive particles, a down/up bloom chain, text rasterized to a texture, and a final pass with glitch, chromatic aberration, flashes and fades. No shadows, no AO, no GI, no fog or volumetrics, no real materials, and not a single effect that hasn't been done a bazillion times before. Every one of those is a tutorial you can find in an afternoon; there's nothing in there you haven't seen in hundreds of prods since the early 2000s.
  
  Camera and editing: straight lines and orbits with easing, scenes joined by cuts and fades.
  
  Music: rendered up front from sines, noise and a couple of filter types, then a textbook bus: ping-pong delay, Freeverb-style reverb, EQ and a limiter.
  
  All of it is standard, done many times before, and none of it is used in a way that makes it yours. Dissonance doesn't turn defaults into an argument. The mood and the sound have something, but that's it.
  
  Honestly, take a look at the most viewed 64ks: fr-08: .the .product, heaven seven, candytron, poemtoahorse, Project Genesis, Chaos Theory, Panic Room, Dead Ringer, Planet Potion... most of them made between 2000 and 2007, on a fraction of today's hardware and with tools people had to write themselves. Then ask yourself, honestly: can this be compared to any of them? Twenty years later, with everything in our favour, it should at least be in the same conversation.
  
  A 64k that really pushes the limits sits on decades of accumulated work: engines, tools, synths, techniques refined over years, plus months of direction, design, sync and pacing for each prod. It doesn't come out of two afternoons and a couple of prompts, with AI or without it.
  
  If you just want to do quick releases for fun, that's completely legit. But a message this big needs weight behind it.
  
  None of this is meant to put you down. It's constructive criticism, like revival's, and I'd read the rest of the thread the same way: as a reason to try harder, not to quit. Come back with a new version or a new 64k that really impresses. Let's raise the bar and push the limits in what is, imho, the hardest category in the scene.
• isokadded on the 2026-10-29 11:24:07
• demo JavaScript LOL FUCK AI by xplsv [web]
• @mrdoob Hi Trace,
  
  Not judging whether this prod is good or bad.
  
  When I asked in the fix-me thread how to make the scene healthier, you answered constructively and talked about how hard it is to keep a positive culture in the three.js communities you moderate. So, honest question: does this prod help with that, or does it just add fuel to the fire?
  
  Arguing hard has always been part of the scene. Insulting people directly is something else, and yes, some people crossed that line with you and other sceners these past months. But someone with your reach and history sets the tone whether you like it or not, and answering in kind just makes it everyone's fight. Sometimes being the adult in the room means not throwing the next punch, even when you'd be entitled to. It feels worse than ever lately. After months of reading these threads, I understand better why someone who gave so much and loved the scene like Gargaj burned out and left.
  
  On the "proving a point" part: making a half-baked prod in 3 hours is easy now. Hard problems still aren't. Look at iq's Elevated-in-the-browser thread. It's been open for weeks with lots of skilled people in it, and a 1:1 port still sits around 4.6-4.9KB, not 4096. The browser sandbox is brutally strict for sizecoding, so you have to get really creative. Another example: why not try a real 64KB browser port of Tokyo, that wonderful demo of yours? IMHO that would be an order of magnitude harder, and no agent or a few hours of vibecoding would get you there, because the problem itself is inherently complex. Every asset has to become procedural code, the synth has to be rebuilt, the browser gives you no Crinkler-style packer so you fight the compression and bootstrap tricks yourself, and every byte you save in one place changes the trade-offs everywhere else. It's thousands of small decisions, measuring, failing and trying again, guided by taste and experience. That's pushing limits like in the old days. (For the record, and as an experiment, porting this prod fully to native without electron whatsoever took me about an hour... But, what's the value in this? That's not pushing limits at all.)
  
  Tools change, the challenge doesn't. Transpiling and decompiling are basically solved problems now, so the merit is in making something new. If you port a 64k, keep it at 64KB on the target platform. PC, Amiga, browser, PICO-8, it doesn't matter: rules are rules. (I'm biased: 64ks are my favourite by far, and I've been chasing one I'm happy with for decades...) Plenty of bloated demos in pouet that could fit in 64KB. But... Is anyone using all this new tech to prove that?
  
  And I'm not saying this from a pedestal. When the browser ports thread started I thought it was cool idea and didn't think much about the implications. At one point I made a Windows port of Starstruck, because I love that demo forever so I wanted to learn its tricks, and wanted to prove that a hard challenge like that was doable with a perfect 1:1 port. Then I realised the original authors might not be OK with it, so I took it down and deleted the link. I hadn't considered that, and honestly it was a bit selfish, just chasing the satisfaction of learning and solving hard problems for my own sake. I still have lots of ports I've made for learning, but I keep them to myself, and if I ever release any, it will be with the original authors' endorsement.
  
  On a personal note: I've known you and the Spanish scene since way back, and you were a big part of it 🙂 I'd love to see those retired guys releasing again. Maybe new tools can help, if they bring people together instead of splitting us further. And personally I'd love to see new tech used for real challenges that help society.
  
  So I'd genuinely like to hear it from you: what was the intention behind this prod? A mirror to make people reflect on how they've been behaving? A joke that got out of hand? Just a way of saying "I don't care what you think"? Or maybe a way to let out frustration, after starting things like the browser ports thread in good faith and getting so much hate back for it? If it's that last one, I'd honestly understand it, and I think a lot of people here would too. I'm asking honestly, because I'd rather hear it from you than guess like half the thread is doing. And the question that matters more to me: does this make the scene more open and bring people together, or does it just do what so much of the world is doing right now, dividing people and feeding more hate?
  
  Not lecturing, we're all grown ups. Just a bit of reflection on ego and on the impact of what we make on others. Engineering should be about solving real problems, not making society worse.
  
  Have a nice day.
  
  PS: English isn't my first language, so an LLM helped tidy this up. Of course, the ideas are mine.
• isokadded on the 2026-10-17 15:08:47
• demo JavaScript SAVONIAN TWIST by Gray Marchers [web]
• These guys really rock hard! Really cool vibes throughout, definitely some Demonscene vibes in there. And yeah, if they could keep this level of quality while squeezing it into 64 KB, they'd be in a whole different league. But even as it is, this stuff is awesome!
• isokadded on the 2026-09-07 08:45:47
• demo JavaScript DEMONSCENE by Gray Marchers [web]
• Quote:

  As for code, I see no reason why not. The demo still packs a lot of dev time tools and unsed assets not needed in the final demo. A more ambitious size coder could make this less than 64k for sure.
  
  As for audio - I think it would need to be redone completely with some other technique to fit 64k.

  
  
  Yeah, I've got a lot of demos on my stack that I'd like to downsize to 64 KB, just to prove it's possible. So when I saw this one, I also wondered if it'd be doable considering the sdf modelling. Squeezing the song down seems quite tricky, though...
  
  Anyway, love the demo! I've watched it a few times now, and both the visuals and the song just give off really good vibes. If this had been 64 KB, it'd have been a solid thumbs up. Anyway, awesome stuff, guys! 👍
• isokadded on the 2026-08-31 14:18:06
• demo JavaScript DEMONSCENE by Gray Marchers [web]
• Wow 😮 .... Congrats guys! this one made my day. Great vibes from
  start to finish. :)
  
  One honest question, and maybe a challenge: do you think something like
  this could fit in 64k?
  
  rulez!
• isokadded on the 2026-08-23 11:30:17
• wild Animation/Video Suboceanic Reimagined by Tomkh
• Interesting to read all the reactions here — feels like this one hit a nerve.
  
  I think there’s fair criticism: the pacing, camera work, and transitions could definitely be stronger, and it does feel more like an experiment than a fully crafted prod. Given the described process — essentially feeding selected frames from the original into off-the-shelf models and assembling the results — it’s also understandable why many people perceive it as very low effort. As a standalone wild entry, it didn’t really feel that compelling to me either — beyond the initial curiosity. That said, if something like this level of visual output could be brought into a 64k intro context with proper polishing, that would actually be quite impressive.
  
  What Blueberry said earlier — “I hope this wave doesn't split the scene in half. There should be room enough for all of us.” — really resonates with me. I’ve been around the demoscene for ~30 years, mostly as a watcher, reading Pouet regularly, enjoying a lot of what people created over the years. But I’ve rarely posted, partly because the tone here can sometimes get... a bit intense :)
  
  It would be a shame if that kind of atmosphere discouraged new people from sharing experiments like this — even imperfect ones — especially in a context where they’re clearly allowed.
  
  Also, stepping back a bit: even in 64k intros, it doesn’t feel like we’ve fully exhausted what’s possible yet. There’s been a lot of amazing work over the years, but still plenty of unexplored space. To me at least, pushing new techniques into that constraint — whatever tools are used — feels like a natural continuation of that spirit of exploring ideas within tight limits and seeing where they lead.
  
  So yeah, I’d be curious to see this pushed further, maybe even towards a proper 64k intro with more direction and polish.
  
  In any case, whether people liked it or not, it definitely got people thinking — and that counts for something.
• isokadded on the 2026-04-24 12:16:33