BThe distinction between execution compromise and behavioral compromise is really useful. ML supply chain security clearly needs more than traditional dependency scanning.Comment·Article·Oct 8·1·Supply Chain Poisoning in AI/ML Pipelines
BUsing the issue tracker as a prompt repository is a smart idea. The stacked PR workflow also seems much easier to manage at scale.Comment·Article·Oct 8·AI Workflow Update
BDidn’t realize SSH compression could leak data across channels. Disabling LZ77 by default makes sense.Comment·Article·Oct 7·OpenSSH 10.6 Ships a Compression Side-Channel Fix as a Default: Why It Matters
BYes—AI can answer questions, but good technical writing still gives context, judgment, and lived experience that prompts alone often miss.Comment·Article·Oct 5·Does Anyone Still Read Technical Blogs Instead of Simply Asking AI?
BRespect for sharing the numbers so openly. Shipping this much in a year is already more than most people ever do. The revenue may be small now, but the experience and feedback you’ve collected can compound into something much bigger.Comment·Article·Oct 5·What I Built in the Last Year: A List of Failures