HNHacker News
TopNewBestAskShowJobs

chaosprint

1,681 karma · joined January 2, 2022

a computer music worker developing Glicol (graph-oriented live coding language): https://glicol.org
submissionscomments
chaosprint··on Recreating Daft Punk's Something About Us
Very interesting article and process.

> "A huge part of the joy came from working in Ableton Live 12, which now feels like an extension of how I think" I feel the same way, but for Reaper

Seeing the author mentions that decomposition of the snare is the hardest, and that’s what I was trying to solve in https://github.com/chaosprint/RaveForce (just an idea).

Multimodality AI is much more powerful now. I wonder how helpful it would be for music and art education if AI could help us deconstruct some songs.

When I was teaching kids with Glicol, I often used KraftWerk’s Das Model as an example:

https://glicol.org/demo#themodel

This rough midi version is very different from the original, but the kids had a lot of fun messing around with it.

chaosprint··on Study finds solo music listening boosts social well-being
is there any good app to know local music event?
chaosprint··on A 6-Hour Time-Stretched Version of Brian Eno's Music for Airports
sounds like paulstretch is heavily used. you can get similar results when applying this to almost any sont.
chaosprint··on The demoscene as a UNESCO heritage in Sweden
Really interesting to read about this! That's wonderful validation for a vital digital culture and its heritage.

As the creator of Glicol (https://glicol.org/), based in Oslo and working in the digital arts space, I'm always fascinated by how different countries foster creative technology. Sweden's approach in recognizing the demoscene this way is particularly encouraging.

It makes me reflect on the pathways to support here in Norway. While academic environments can be very supportive (as my previous supervisors have been), navigating the broader public arts funding structures for newer digital art forms sometimes feels more challenging, especially perhaps for those working outside of long-established networks.

Seeing Sweden's success in formally recognizing this kind of digital heritage is genuinely inspiring and offers food for thought.

chaosprint··on Koto Programming Language
You are absolutely right. Thanks for the correction and info!

I gave Lua a shot, but getting the toolchain set up in wasm was a hassle:

https://bytedream.github.io/litbwraw/introduction.html

So at least I can say Rhai has some advantages on syntax and compatibility with wasm

chaosprint··on Koto Programming Language
Should compare with Rhai (https://rhai.rs/)

I found rhai's syntax very straightforward, and I could almost accomplish my needs just by looking at some basic examples.

I use Rhai in wasm, and it can handle real-time audio blocks, which is really impressive:

https://glicol.org/tour#meta2

chaosprint··on OpenAI adds MCP support to Agents SDK
First, I don't want to pay Claude Code at all...

Second, isn't that doable with API calling :)

chaosprint··on OpenAI adds MCP support to Agents SDK
claude needed these those tools in 2024, so having the community contribute for free was actually a smart move.

service providers get more traffic, so they’re into it. makes sense.

claude 3.5 was great at the time, especially for stuff like web dev. but now deepseek v3 (0324) is way better value. gemini's my default for multimodal. openai still feels smartest overall. i’ve got qwq running locally. for deep research, free grok 3 and perplexity work fine. funny enough, claude 3.7 being down these two days didn’t affect me at all.

i checked mcp since i contribute to open source, but decided to wait. few reasons:

- setup’s kind of a mess. it’s like running a local python or node bridge, forwarding stuff via sse or stdio. feels more like bridging than protocol innovation

- I think eventually we need all the app to be somehow built-in AI-first protocol. I think only Apple (maybe Google) have that kind of influence. Think about the lightening vs usb-c.

- performance might be a bottleneck later, especially for multimodal.

- same logic as no.2 but the question is that do you really want every app to be AI-first?

main issue for me: tools that really improve productivity are rare. a lot of mcp use cases sound cool, but i'll never give full github access to a black box. same for other stuff. so yeah—interesting idea, but hard ceiling.

chaosprint··on Qwen2.5-VL-32B: Smarter and Lighter
no. it's nothing to do with deepseek. it's openrouter and providers there
chaosprint··on Qwen2.5-VL-32B: Smarter and Lighter
it seems that this free version "may use your prompts and completions to train new models"

https://openrouter.ai/deepseek/deepseek-chat-v3-0324:free

do you think this needs attention?

chaosprint··on Show HN: I built a website for sharing drum patterns
there is also a cli:

https://github.com/glicol/glicol-cli

chaosprint··on Show HN: I built website for sharing Drum Patterns
there is a login
chaosprint··on Show HN: I built a website for sharing drum patterns
I would be happy to port some patterns to Glicol (https://glicol.org/)

only suggestion: support https...

chaosprint··on Show HN: I made a Sonic Pi MCP to create music with prompts
I am also working on MCP for Glicol (https://glicol.org/). But I think Gemini's multimodality may be more suitable than Claude. Have you tried to connect to models other than Claude?
chaosprint··on MCP – Flash in the Pan or Future Standard?
yeah you can simply replace mcp with langchain and have a new article
chaosprint··on Show HN: Codemcp – Claude Code for Claude Pro subscribers – ditch API bills
I was wondering if I could use Claude Desktop to control VS Code directly (perhaps a plugin in it) so that I only need to open two windows.
chaosprint··on Hacking Your Own AI Coding Assistant with Claude Pro and MCP
If you use bolt.new or v0's free tier to make a design and perfect it in cursor, the speed will be much faster than roo, and there will probably be a lot of fast credits left. cost $20
chaosprint··on Hacking Your Own AI Coding Assistant with Claude Pro and MCP
I have cancelled the cursor subscription, but it is more for functional reasons.

I am confused that the author said he has privacy concerns about cursor, but the most satisfactory thing about cursor is that it asked me whether I want privacy mode at the beginning.

On the contrary, I am very skeptical about whether these websites with free quotas, not only claude, but also grok, chatgpt, etc., will use chat data for training.

If you are really worried about this, it is not impossible to deploy deepseek r1 locally now.

chaosprint··on Show HN: Web Audio Spring-Mass Synthesis
Great demo.

I used to do some web audio and tonejs works, but later switched to rust and glicol for sound synthesis.

For example, this handwritten dattorro reverb:

https://glicol.org/demo#handmadedattorroreverb

This karplus-stress-tester may also be interesting to you.

https://jackschaedler.github.io/karplus-stress-tester/

In short, I think to study more powerful physics synthesis, you need to consider the technology stack of

- rust -> wasm - audioworklet - sharedarraybuffer

Visual can rely on wgpu. Of course, webgl is enough in this case imho.

If it is purely desktop, you can consider using the physics library in bevy.

chaosprint··on Gemma3 – The current strongest model that fits on a single GPU
How does this compare with qwq 32B?
chaosprint··on Canon EF and RF Lenses – All Autofocus Motors
Agreed. EF's short flange focal distance can also be easily connected to Sony to utilize dual ISO. Indie uses manual focus anyway.
chaosprint··on Canon EF and RF Lenses – All Autofocus Motors
Canon's lenses are great, but they're really expensive. The camara body itself has the same issue.

Sony seems to be the first choice for indie filmmakers and youtubers right now.

L-mount and M43 also seem to have great potential as Panasonic supports Phase Detection Autofocus.

edit:

There is no need to be cynical here. Market figures are cruel. I hope they are all good. Competition is the best. I should also mention the return of Nikon. The patent threshold of RAW such as RED is what we need to oppose.

chaosprint··on I've been using Claude Code for a couple of days
I'd be very interested to know of a way to make it work with AI that doesn't require debugging if you can illustrate.
chaosprint··on I've been using Claude Code for a couple of days
It seems the original poster hasn't extensively tried various AI coding assistants like Cursor or Windsurf.

Just a quick heads-up based on my recent experience with agent-based AI: while it's comfortable and efficient 90% of the time, the remaining 10% can lead to extremely painful debugging experiences.

In my view, the optimal scenarios for using LLM coding assistants are:

- Architectural discussions, effectively replacing traditional searches on Google.

- Clearly defined, small tasks within a single file.

The first scenario is highly strategic, the second is very tactical. Agents often fall awkwardly between these two extremes. Personally, I believe relying on an agent to manage multiple interconnected files is risky and counterproductive for development.

chaosprint··on Ereader Easy Swedish
I found this very useful for learning second language: https://immersivetranslate.com/

Daily news is better resource than books IMHO.

I have no vested interest here. The thing is, since it's not open source, I'm a bit worried about data privacy, especially because the plugin can see everything I browse.

chaosprint··on Launch HN: A0.dev (YC W25) – React Native App Generator
Your title is honest, and your description of the limitations is honest. It is also pragmatic to allow you to try it without registering. This is the way to success.
chaosprint··on Music Generation AI Models
I got into AI music back in 2017, kind of sparked by AlphaGo. Started by looking at machine listening stuff, like Nick Collins' work. Always been really curious about AI doing music live coding.

In 2019, I built this thing called RaveForce [github.com/chaosprint/RaveForce]. It was a fun project.

Back then, GANsynth was a big deal, looked amazing. But the sound quality… felt a bit lossy, you know? And MIDI generation, well, didn't really feel like "music generation" to me.

Now, I'm thinking about these things differently. Maybe the sound quality thing is like MP3 at first, then it becomes "good enough" – like a "retina moment" for audio? Diffusion models seem to be pushing this idea too. And MIDI, if used the right way, could be a really powerful tool.

Vocals synthesis and conversion are super cool. Feels like plugins, but next level. Really useful.

But what I really want to see is AI understanding music from the ground up. Like, a robot learning how synth parameters work. Then we can do 8bit music like the DRL breakthrough. Not just training on tons of copyrighted music, making variations, and selling it, which is very cheap.

chaosprint··on DeepSeek-R1 for Coding in Zed
I really like Zed, especially its attitude of starting from the scratch.

But if you want to be commercially successful, maybe it's worth providing some free trials. In comparison, Cursor has 50 trials.

At least it can give people more motivation to download.

Haven't looked into the detail but a privacy mode is also important for providing R1.

chaosprint··on Hedy: Textual programming made easy
Love the mission!

My project Glicol (https://glicol.org/) also has some classroom practice for sound 101

But I think what children need most is to provide very simple to complex examples for modification.

The bigger worry is that other things, like Instagram, become more appealing than learning to code.

chaosprint··on One in four 2020 Tesla Model 3 failed the Danish periodic inspection in 2024
bought my model y in Dec 2023.

The brakes are always noisy in winter now. I went for maintenance and it got better, but it got worse recently. The automatic judgment of the wipers is very inaccurate. Sometimes they start like a ghost and scare me. The reversing camera is always blurry in the snow in Norway. Other than that, there are not many complaints. I feel that the basic battery life and control are really great. It’s just that I think Tesla’s current model design seems difficult to achieve true fully automatic driving, especially in places like Norway where there is a lot of rain and snow. Because many cameras are blocked (even if you want to simulate human eyes, but the human eyes are in the car). I don’t even dare to cruise on autopilot, which can easily cause danger in rainy days.

← PreviousPage 4 of 17Next →