Feed aggregator

Article URL: https://atomichabits-calendar.com

Comments URL: https://news.ycombinator.com/item?id=47057956

Points: 1

# Comments: 0

Article URL: https://isonomiaquarterly.com/wp-content/uploads/2026/02/iq-4.1-spring-2026-poenicke-rudeness-and-liberty.pdf

Comments URL: https://news.ycombinator.com/item?id=47057943

Points: 1

# Comments: 0

I built a pipeline for querying fleet telemetry (AVs, robots, vehicles) in natural language. Load Parquet or CSV into cuDF, then ask things like "Which vehicles exceeded 120 km/h in region X?" and get back IDs and metrics. Tech: cuDF for GPU ingest/analytics, NVIDIA NIM on GKE for LLM inference, format-aware model selection (GGUF local, TensorRT prod). Three notebooks: data ingest with pandas vs cuDF vs cudf.pandas benchmarks, local Gemma 2 inference, and full NIM deployment. Runs on Colab with a T4 for notebooks 1 and 2; notebook 3 uses GCP and NIM on GKE.

Comments URL: https://news.ycombinator.com/item?id=47057926

Points: 1

# Comments: 0

Article URL: https://dpaint.app

Comments URL: https://news.ycombinator.com/item?id=47057919

Points: 1

# Comments: 1

I built MeterLogs because most energy tracking solutions require smart devices, installers, or expensive hardware.

MeterLogs takes a pragmatic approach: you enter your monthly meter readings manually, and it generates simple yearly dashboards with grouped energy types (electricity, water, gas, etc.).

It’s designed for homeowners who just want a clear overview without IoT setups or subscriptions to hardware ecosystems.

I’d love feedback, especially on usability and pricing.

Comments URL: https://news.ycombinator.com/item?id=47057911

Points: 1

# Comments: 0

Article URL: https://jellyocean.com/

Comments URL: https://news.ycombinator.com/item?id=47057909

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=eIoohUmYpGI

Comments URL: https://news.ycombinator.com/item?id=47057884

Points: 1

# Comments: 0

Article URL: https://github.com/taleshape-com/shaper

Comments URL: https://news.ycombinator.com/item?id=47057879

Points: 2

# Comments: 0

Article URL: https://nvie.com/posts/15-years-later/

Comments URL: https://news.ycombinator.com/item?id=47057829

Points: 33

# Comments: 1

Article URL: https://gist.github.com/jake-stewart/0a8ea46159a7da2c808e5be2177e1783

Comments URL: https://news.ycombinator.com/item?id=47057824

Points: 3

# Comments: 1

Article URL: https://www.premeth.com/

Comments URL: https://news.ycombinator.com/item?id=47057805

Points: 1

# Comments: 0

Article URL: https://hackaday.com/2026/02/16/a-computer-that-fits-inside-a-camera-lens/

Comments URL: https://news.ycombinator.com/item?id=47057778

Points: 1

# Comments: 0

Article URL: https://github.com/laminardb/laminardb

Comments URL: https://news.ycombinator.com/item?id=47057770

Points: 2

# Comments: 1

Article URL: https://github.com/zavg/linux-0.01

Comments URL: https://news.ycombinator.com/item?id=47057765

Points: 1

# Comments: 0

I've spent the last week trying to use Claude Opus in an IDE via a localhost proxy that wraps the Claude Code CLI.

The proxy exists because there's no way to get programmatic API access to Claude on a Pro ($20/mo) or Max ($200/mo) subscription.

This isn't an oversight. Anthropic has actively taken steps to prevent subscription-based API access. They've shut down third-party bridges, tightened terms of service, and made it clear this is corporate policy, not neglect, not a missing feature, but a deliberate business decision.

Meanwhile, OpenAI opened API access to ChatGPT Pro/Plus subscribers. The result is predictable.

A CTO friend is prototyping an electronic warfare detection system. His existing customers include a major European MNO that operates across dozens of countries. He's currently splitting ~$80/month across four AI subscriptions, constantly rationing tokens, juggling providers. He'd switch to Claude Max tomorrow if it included IDE API access. Instead, he's moving his entire prototype to Codex/OpenAI, because they made it possible.

Here's why this matters beyond one developer: The prototype becomes the demo. The demo becomes the procurement spec. The spec becomes a multi-year enterprise contract. This particular MNO signs seven-figure deals per country rollout. Once the prototype is built on OpenAI, Claude never enters the conversation. Anthropic doesn't just lose one deal, they lose an entire multi-country enterprise pipeline they'll never even know existed.

This isn't an isolated case. In my circle, senior engineers, CTOs, infrastructure people, nearly everyone would switch from Pro to Max if it meant using Claude in their IDE the way they can now use OpenAI. That's going from $20/month to $200/month per user, zero acquisition cost. Even if only 10-20% of the broader Pro base converts, the revenue math is overwhelming.

The proxy ecosystem (claude-max-api-proxy, Antigravity, and others) is direct proof of unmet demand. Developers are writing code to work around a billing boundary. That's not abuse, that's a market signal Anthropic is choosing to ignore.

Claude Opus is technically superior to Codex for most development tasks. That means nothing when developers can't access it in their workflow.

Anthropic built the best model and then made a policy decision to keep it out of the environment where long-term platform adoption is decided.

Comments URL: https://news.ycombinator.com/item?id=47057752

Points: 2

# Comments: 0

Hi HN,

I built TokenMeter after seeing how unpredictable LLM costs become as usage scales.

Most teams optimize prompts but lack real-time cost observability.

This project tracks token usage and exposes cost insights to help AI teams stay financially aware.

Would appreciate feedback on architecture and missing observability features.

Thanks.

Comments URL: https://news.ycombinator.com/item?id=47057746

Points: 1

# Comments: 1

Article URL: https://github.com/openclaw/openclaw

Comments URL: https://news.ycombinator.com/item?id=47057712

Points: 1

# Comments: 1

Article URL: https://github.com/DatanoiseTV/tinyice

Comments URL: https://news.ycombinator.com/item?id=47057707

Points: 3

# Comments: 1

Article URL: https://twitter.com/ID_AA_Carmack/status/2023805426345689198

Comments URL: https://news.ycombinator.com/item?id=47057705

Points: 2

# Comments: 0

I built a 4-tier self-healing runtime for AI agents running 24/7.

When the gateway goes down, it escalates: watchdog → HTTP health check → Claude Code "doctor" (reads logs, diagnoses, autonomously fixes) → Discord alert.

The interesting part: Claude Code is both the patient and the doctor. Open source, bash-only, macOS/Linux.

Comments URL: https://news.ycombinator.com/item?id=47057694

Points: 1

# Comments: 0