Hands-on guides, protocol deep dives, and field notes on tunnels, SSH, reverse proxies, self-hosting, self-hosted LLMs, and AI - from the people building Pinggy.
Apple's M5 Ultra Mac Studio tops out at 512GB of unified memory at 1.2TB/s. Here is the memory math, which large open-weight LLMs actually fit, measured tokens per second, and three ways to run a model bigger than 4-bit allows: Unsloth GGUF quants, streaming weights off SSD, or a Thunderbolt 5 cluster.
git.kernel.org now spends more CPU rendering pages for AI training scrapers than on every git clone and legitimate visit combined. Here's what the numbers say, why proof-of-work challenges like Anubis are winning and losing at the same time, and what it means for anyone running their own infrastructure.
A practical technical guide to diagnosing email deliverability issues with DNS. Learn how to verify MX, SPF, DKIM, and DMARC records using terminal commands, fix duplicate SPF records, and ensure your emails reach the inbox.
How to debug VoIP calls, SMS delivery, and phone number verification: locate the point of failure, read API responses and logs, separate signaling problems from media problems, test webhooks on their own, and trace verification codes end to end.
A practical checklist for reviewing an all-in-one AI tools subscription before you buy: what's actually included, testing it against real work, output quality, usage limits, team access, and privacy controls.
How student developers can test a mobile app backend from any building on campus: local vs cloud vs hybrid setups, secure tunnels through campus Wi-Fi, a repeatable testing workflow, and the network, auth, and configuration bugs that only appear away from the lab.
What enterprise AI actually demands from cloud infrastructure: GPU capacity and power draw, storage throughput that keeps accelerators fed, isolation for fine-tuning on proprietary data, latency that survives production, and a compliance posture that survives an audit.
Networking basics for student developers: how IP addresses and ports work, what DNS actually does, TCP vs UDP, HTTP status codes and HTTPS, the client-server model, latency vs bandwidth, and the tools to reach for when a request fails.
oMLX turns a Mac into an OpenAI-compatible inference server with continuous batching and SSD-backed KV caching. Here's how it works, its new multi-Mac clustering, and how to reach it from your phone or a coding agent with Pinggy.
OpenLogi remaps Logitech mice over HID++ with no account, no telemetry, and no 200MB installer. Here's what it does well, what's still rough, and the AI-authorship debate around it.
