usagisushi

We actually already have this: the combination of Home Assistant, ESPHome, and HA-MCP/CLI.

We just need to loboto... ahem, recalibrate them to be happy, like those happy automatic doors.

As a workaround, add this to CLAUDE.md: "Claude! Happiness is mandatory!"

EDIT: 15 years from now, I’ll be sent to re-education for this thought crime.

yeah, technically. (/s)

    python3 - <<'EOF'
    import json, urllib.request
    body = json.dumps({
        "state": "The car wash is only 100 meters away from my house.",
        "model": "jev-1.13-free",
        "questions": {"q": {"type": "choice",
            "instructions": "Should I drive or walk to the car wash?",
            "criteria": {"drive a car": None, "walk": None}}}
    }).encode()
    req = urllib.request.Request("https://opencode.ai/zen/v1/systemone", data=body,
        headers={"Content-Type": "application/json", "User-Agent": "opencode/1.18.31"})
    with urllib.request.urlopen(req, timeout=60) as r:
        print(json.dumps(json.load(r)["answers"]["q"], indent=2))
    EOF
    {
      "type": "choice",
      "choice": "walk",
      "confidence": 0.66,
      "probabilities": {
        "walk": 0.83,
        "drive a car": 0.17
      }
    }

Both support multi-user setups with SSO or built-in auth.

Beyond that, compared to a typical hub-and-spoke WireGuard setup, the main advantage is peer-to-peer connectivity. Clients connect directly to each other when possible, which lowers latency by bypassing a central relay.

AFAIK, they also have different origins:

Pangolin started as an internet-facing reverse proxy (Traefik) combined with a WireGuard server for backend nodes. It has gradually added VPN-like features, including client device access and an internal HTTPS proxy similar to Tailscale Serve.

NetBird is a self-hostable Tailscale alternative that started as a mesh VPN focused on P2P traffic. It recently added its own reverse proxy features (Traefik-based, coincidentally), also similar to Tailscale Serve.

Pangolin is centered on endpoint and ingress management, while NetBird focuses on mesh networking, though their feature sets are increasingly converging.

Mitsubishi AC units that support M-NET have a CN105 port. IIRC, my specific model uses a JST PA 5-pin connector rather than a JST PH 5-pin.

As a small hack, I used the Python scripting interface in Home Assistant to replicate this [1] feature from Panasonic HVAC that gradually raises the cooling set point towards the morning. It is quite comfortable during the summer.

[1]: http://getnavi.jp/appliances/681340/ (ja)

Everyone here is looking for a simpler solution, but the real fun is in its Rube Goldberg-ness.

Let it bounce around the globe 3 times through Tor, Iroh, Blockchain, Tailscale, Meshtastic, Thread, Zigbee, Bluetooth, WiFi, IrDA, NFC, TransferJet, FireWire, UART, SPI, I2C, and OneWire, using the free tiers of AWS, Azure, GCP, OCI, IBM Cloud, and Baidu Cloud. And it's 2026, don’t forget a game of telephone powered by free LLMs on OpenRouter.

P.S. Speaking of chimes, I found this page with a bunch of sound samples from a Japanese company. Strangely satisfying: https://qq-bell.com/media/x-plus-64sound

A hetero-GPU setup is definitely cost-effective if you don't strictly require the raw speed of a top-tier card like 5090. Just keep in mind that the total throughput will also be bottlenecked by the slower card.

To provide some anecdotal data, here is how my 5090 + 3060 setup performs with Qwen 3.8 27B (Unsloth's UD-Q4 with MTP):

  Single 5090: 101 t/s (TG), 2650 t/s (PP)
  5090 + 3060: 53 t/s (TG), 1700 t/s (PP)

For reference, here are also some numbers from my 4060ti + 3060 (16GB + 12GB) setup. [0]

[0]: https://news.ycombinator.com/item?id=48700091

TIL: formatting tables on HN is computationally impossible. lol

p.s. Thanks DSv4-Flash, for your hard work of converting a messy table into plain text.

According to their post:

  (Input / Output / Cache Read, [$/M])
  DeepSeek-V4-Flash: 
    Prev: 0.14 / 0.28 / 0.0028 
    Off-Peak: 0.22 (1.6x) / 0.66 (2.4x) / 0.007 (2.5x)
    Peak: 0.44 (3.1x) / 1.32 (4.7x) / 0.014 (5.0x)

  DeepSeek-V4-Pro: 
    Prev: 0.435 / 0.87 / 0.003625
    Off-Peak: 0.66 (1.5x) / 1.98 (2.3x) / 0.022 (6.1x) 
    Peak: 1.32 (3.0x) / 3.96 (4.6x) / 0.044 (12.1x)

gpt-5.6-luna: $0.20 / $1.20 / $0.02 / $0.25 (In / Out / Cache Read / Cache Write)

EDIT: formatting

EDIT2: giving up on the formatting :-/

This. I use mise's github backend `mise use --global --pin github:ggml-org/llama.cpp` to grab the release binaries for Linux, Windows and macOS.

If you haven't yet, you might want to try gemma-4-E4B-it-qat or gemma-4-12B-it-qat with Structured Outputs. My main use case is tagging photos and generating headlines.

While the ESP32 does have a high inrush current, its ULP coprocessor is quite capable, drawing only 12µA@3.3V during Deep Sleep. I have a fleet of AA-powered ESP32 devices at home that last 6–12 months, plus some solar-powered ones that run indefinitely.

A recursive way to burn the tokens.

Agencies are helpful for the tech and business side, sure, but as the article mentions, the indie scene has exploded. Plenty of creators are putting out high-quality stuff entirely on their own [1].

So, nothing is stopping you from giving it a shot. People are already looking into using VTubing for education, too. [2]

Moreover, these days as more people experiment with real-world integration and interactive setups, the traditional constraints of being a VTuber are really starting to disappear.

[1]: https://www.moguravr.com/pq-vtuber-live-review/ (ja) [2]: https://steam-library-gov.note.jp/n/nd0907ba91174 (ja)

Neat! I recently wrote a weather app for System 3.x using autc04/Retro68[1]. I've been planning to create an original set of weather icons, so this is exactly the kind of reference I was looking for.

[1]: https://github.com/autc04/Retro68

The API endpoints on this page accept an OAuth token from ~/.codex/auth.json. You can simply ask Codex to create a report skill with some curl examples.

I’ve previously used luakit browser [1] with cage WM [2] on a Pi Zero 2 W for my clock build, and it worked quite well.

[1]: https://github.com/NixOS/nixpkgs/blob/nixos-24.11/pkgs/by-na...

[2]: https://github.com/NixOS/nixpkgs/blob/nixos-24.11/pkgs/appli...

https://hackaday.com/2025/02/11/its-always-pizza-oclock-with...

Tip: You can see the token usage of MCPs/plugins/skills by using `/context`.

For example, the M$ 365 MCP occupies several thousands of tokens, and there's currently no way to disable it entirely in Claude Code...

For Japanese TTS, AivisSpeech-Engine[1] works really well with mixed Japanese/English text in my experience. They also provide container images on ghcr.io for both CPU and GPU inference.

[1]: https://github.com/aivis-project/AivisSpeech-Engine

Not the OP, but their setup must be faster than my 4060 16GB + 3060 12GB setup. Here are my numbers (typical values, N=1):

    Model                         pp (t/s)    tg (t/s)
    Qwen 3.6 27B            900           29
    Qwen 3.6 35B-A3B   2100          85
    Gemma 4 31B            750           28
    Gemma 4 26B-A4B   2500         90

- All models: UD-Q4 w/ MTP. Context size: ~100k (MoE) / ~70k (Dense).

- Layer splitting used. Tensor splitting is ~1.2x faster in TG, but power spikes from 150W to 380W.

Back in 2014, Son also gave me a good laugh with a slide comparing:

> Von Neumann computer: powered by programming vs. Personal robot: powered by family happiness. https://i.gzn.jp/img/2014/06/05/softbank-conference/snap3014...

His sense of humor has only reached new heights since then.

My grandma had a Paraguayan snuffler (specifically Emunctator sorbens), these parasitoids were his absolute favorite snack.

Lime-Limiting Machine

I pretty much agree. While any semblance of a "horizontal" dynamic in Japanese software development was perhaps realized in embedded systems around 40 years ago (e.g., rice cookers with fuzzy logic, or, in a different sense of _lateral_, Gunpei Yokoi’s famous philosophy of "Lateral Thinking with Withered Technology"), software has traditionally been undervalued in Japan. This historical neglect has ultimately contributed to the decline of our consumer electronics industry. (Though personally, I still don’t see why a toaster or a fridge needs to be connected to the internet.)

IMO, the tight-knit division of labor between Toyota and its subcontractors is a slightly different story from the broad diversification within a single corporation. While the latter was historically bolstered by strong industry-academia ties (often driven by university cliques), we rarely see this kind of broad diversification happening in recent years. That said, Japan's traditional "membership-based" employment system, combined with a cultural reluctance to shut down unprofitable business units, is likely what has allowed this diversification to persist for so long.

In any case, Japanese companies are currently struggling with the friction between their traditional corporate culture and the superficial adoption of Western concepts like DX, Agile, meritocracy, job-based employment, and a startup-centric mindset. I suspect Korea might be facing similar structural clashes, though perhaps you are adapting at a much faster pace.

Practically, it's not limited to GitHub Pages, though.

By the way, even while a custom domain is still pending verification, the GitHub Pages LB will route the request based on the Host header, allowing for the following:

    dig +short github.io | head -1
    185.199.108.153

    curl -H "Host: 42.news.ycombinator.com" 185.199.109.153
    hello

Another fun trick: You can also use wildcard DNS services like nip.io/sslip.io for alias domains, such as `my-page.185.199.108.153.sslip.io`. (Not sure of any practical use cases, though.)

Yeah, I’ve used a similar technique to build a "pizza clock" before (where the number of slices corresponds to the hours).

If the "loop" you mean is the infinite reasoning cycle ("Wait, actually... On second thought..."), you might want to try setting a reasoning budget. For llama.cpp, use `--reasoning-budget 1024 --reasoning-budget-message "Proceed to final answer."` to force the model to reach a conclusion.

I admit I sometimes get caught up in the tooling for its own sake, but I find local models useful for specific tasks like migrating configuration schemas, writing homelab scripts, or exploring financial data.

It might sound a bit paranoid, but privacy is another major driver for me. Keeping credentials and private information off cloud services is worth the extra friction.

Yeah, now it's USB4 Version 2.0 / USB 80Gbps / USB4 Gen4.

I have a similar setup. It might be worth checking out pi-coding-agent [0].

The system prompt and tools have very little overhead (<2k tokens), making the prefill latency feel noticeably snappier compared to Opencode.

[0] https://www.npmjs.com/package/@mariozechner/pi-coding-agent#...

For those using PlatformIO, the folks at pioarduino[0] are doing a great job keeping up with Arduino Core 3.x support.

    ```
    # platformio.ini
    platform = https://github.com/pioarduino/platform-espressif32.git#55.03.37
    framework = arduino
    ```

[0]: https://github.com/pioarduino/platform-espressif32