SumGuy's Ramblings
The art of wasting time.
Docker, self-hosting, AI/LLM, Linux, and DevOps — explained by someone who learned the hard way. No fluff, no enterprise jargon, just practical stuff that actually works on real hardware.
Recent Posts
-
Function Calling in Local LLMs
Local LLMs can call tools, query APIs, and run code if you set them up right. Function calling on Ollama and llama.cpp explained — patterns that actually work.
12 min read -
HACS: When Custom Integrations Bite You
HACS custom integrations can brick your Home Assistant setup overnight. Here's how to use them safely, recover when they don't, and know when to skip them.
10 min read -
ntopng vs darkstat
ntopng vs darkstat: full DPI flow analysis vs a tiny always-on stats page. Pick the right tool for home-lab network visibility without losing your mind.
10 min read -
Gemma 4 vs Qwen3.6
Gemma 4 vs Qwen3.6: sizes, reasoning, coding benchmarks, and which model you should actually pull for your home lab rig.
11 min read -
AnythingLLM as Knowledge Base
AnythingLLM is the closest thing to a real private NotebookLM you can self-host. Workspaces, RAG, agents, document chat — running locally on Ollama in 20 minutes.
13 min read -
FRR vs BIRD
FRR vs BIRD: two open-source routing daemons compared for BGP, OSPF, and home-lab dynamic routing. Which one belongs in your stack?
9 min read -
Home Assistant Add-Ons vs Docker Containers
HAOS Add-Ons vs bare Docker Compose for Mosquitto, ESPHome, Frigate — when the Supervisor magic is genuinely worth it and when it just gets in your way.
10 min read -
Local Vision LLMs Worth Running in 2026
The 2026 local vision LLM landscape: LLaVA's exit, Pixtral's OCR wins, Qwen3-VL's new tricks, and why Gemma 4 is the home-lab dark horse.
15 min read -
HAProxy vs Envoy
HAProxy vs Envoy for L4/L7 load balancing — config style, performance, observability, service mesh fit, and which one belongs in your home lab.
9 min read -
Home Assistant Voice: Wyoming vs Rhasspy
Wyoming protocol vs Rhasspy for local Home Assistant voice — which offline stack actually works in 2026, hardware requirements, and when to choose each one.
11 min read -
MCP Servers: Tools for LLMs
Model Context Protocol turns your LLM into a tool-using agent — file access, APIs, your home lab. Build your first MCP server in under 50 lines of Python.
12 min read -
Caddy vs Traefik
Caddyfile zen vs Traefik labels-everywhere — picking a reverse proxy for self-hosted services in 2026. Real configs, real gotchas.
10 min read
