Local LLM JSON Output That Doesn't Lie
Constrained decoding makes local LLMs emit valid JSON in llama.cpp, Ollama, and vLLM. It guarantees syntax, not truth. Working curl, Pydantic, and retry code.
Tag: self-hosting
All the articles with the tag "self-hosting".
Catch Your Self-Hosted Apps Phoning Home
Self-hosted is not silent. Audit outbound calls with DNS query logs, watch live connections, block egress with Docker networks, then flip telemetry opt-outs.
Sync Game Saves Without a Cloud Account
Sync game saves between PC, Linux, and Steam Deck without a cloud account. Ludusavi backs up the saves, Syncthing moves them, and Proton paths get sorted out.
Crawl4AI: Feed Your RAG From the Web
Crawl a docs site to clean markdown with Crawl4AI, embed it with Ollama, store it in Qdrant, and query it. A working local RAG ingest pipeline, with politeness.
Migrating 10 Years of Bookmarks
Merge Chrome, Firefox, Pocket, Raindrop and Pinboard exports into one deduplicated file with tags, triage dead links, and import it all into linkding.
Radicale vs Baikal: CalDAV That Syncs
Radicale or Baikal for a shared family calendar? Compose files, a rights file that shares one calendar, Caddy redirects, and the DAVx5 and iOS sync gotchas.
Temporal vs Cron: Durable Home Lab Jobs
Temporal vs cron plus bash for multi-step home lab jobs: what durable workflows buy you, the setup cost, and when a systemd timer is the smarter choice.
journald Structured Logging, No ELK
journald already stores structured key/value log fields. Learn to write custom fields, query them, and ship logs to one box before reaching for Loki or ELK.
AppFlowy vs Notion: Self-Host Your Notes
AppFlowy vs Notion in 2026: the self-host server is now proprietary and the free tier is one browser seat. Licensing, Docker Compose setup, and real prices.
LinuxGSM: Boring Game Servers Done Right
LinuxGSM manages 140 game servers from one script. Compare bare-metal tmux plus cron against the official Docker image, using a Valheim server as the example.
SurrealDB vs Postgres: Real or Toy?
SurrealDB promises documents, graphs, search and realtime in one binary. I tested it against Postgres to see where that pitch holds up for self-hosters in 2026.
Batch Inference: One Model, 100 Users
Loading model weights, not math, is your GPU bottleneck. Continuous batching with llama-server serves many concurrent requests for close to the cost of one.