Blog
Project notes and guides to technology decisions, costs, and risks in business operations.
125 articles found
For Technical Teams
SQLite WAL Bug: The Limits of Mature Technology
A 16-year-old SQLite bug silently dropped committed data. Why boring tech stops being boring once you use it differently.
For Technical Teams
Agentic Tool Loops for Multi-Step Business Work
Typed tools, a bounded loop, per-tool guardrails: how agentic chatbots handle multi-step work single-shot RAG can't.
For Technical Teams
Integrating Access Control and Camera VMS
Getting events out of closed access-control panels and camera VMS, then mapping them into your own domain model.
For Business
Software Handover That Reduces Vendor Dependence
Obtain accounts, code rights, backups and instructions tested by a new team so operations can continue after a vendor change.
For Technical Teams
Recovering Fragmented Video When File Carving Gives Up
When signature carving returns broken files, cluster-level reassembly with global verification loops can still rebuild fragmented video.
For Business
Fixed Price vs Time and Materials for Software Projects
Fixed price, T&M, or milestone hybrid — who carries the risk, why fixed quotes run high, and red flags in both.
For Business
Room-Level Energy Control With Smart Breakers
Smart breakers combine room-level metering and switching. Check installation, reading accuracy, billing rules, and disconnection limits.
For Business
Who Owns the Code? IP Clauses Every Client Contract Needs
Agree code ownership, toolkit licenses, third-party duties, source handover, and payment-linked rights before a software project begins.
For Business
The PMS Integration Mistake That Quietly Costs Hotels Revenue
Three documented PMS integration failure modes quietly cost hotels revenue after go-live — and why the fix is what you connect first.
For Business
AGPL Disputes: Lessons for Software Buyers
ONLYOFFICE vs. Nextcloud and SFC vs. Bambu Lab: two live AGPL disputes showing what the license actually restricts.
For Technical Teams
Rule Zero of Data Recovery: Image the Drive First
Image a suitable drive before recovery attempts. Use ddrescue's mapfile, bound retries, and assess suspected physical damage with a lab.
For Business
Taking Over Legacy Systems: Stabilize Before Rebuilding
Inherited codebase, developer gone, no docs? The order that works: run it, back it up, test it, then rebuild the worst part.