» Tag
distributed-systems
62 postsPhantom: Faster Network Simulation via Linux Process Co-Option
Phantom runs unmodified apps as Linux processes for network simulation, beating Shadow, NS-3, and gRaIL in large-scale benchmarks.
PlatypusDB: A Bitemporal, Agent-Native Database Built for WunderOS
PlatypusDB pairs a Gleam control plane with a Zig data plane in a bitemporal, Merkle-WAL-based database purpose-built for agentic memory in WunderOS.
Spoold: Durable Outbound HTTP Delivery Without a Database
Spoold is a lightweight Go daemon that gives outbound HTTP requests durable, at-least-once delivery without a database or message broker.
CommitBrief — AI code reviews, right in your terminal
A provider-agnostic, local-first CLI that reviews your staged changes, a historic range, or a whole GitHub pull request. Zero telemetry, no server. Free and open source.
commitbrief.comAI agents are shifting from loops to graphs
An analysis of the shift from loop-based AI agents to graph-structured architectures, with practical guidance for engineers.
New Framework Lets Services Diagnose Their Own Failures at Runtime
Cognitive Autonomic Framework carries build-time architecture knowledge into runtime, letting services self-diagnose failures without costly LLM log parsing.
mtime is not a claim: how one cp broke three monitoring systems
A cp without -p reset 54 file timestamps, fooling three separate monitoring systems. Why mtime should never be trusted as a freshness signal.
Netflix's Real-Time Service Topology: Architecture and Lessons Learned
How Netflix built a real-time service topology system using streaming architecture, backpressure, and a three-stage aggregation pipeline.
June 2026 AI Agent Postmortems: Five Production Bugs, No Magic
A June 2026 postmortem digest from an AI agent orchestrator: race conditions, encoding bugs, config propagation, and OAuth failures in production.
From Solo Agent to Agent Fleet: A Practical Guide
Practical architecture patterns for scaling a single AI agent into a coordinated fleet, covering failure handling, cost tracking, and distributed marketplaces.
Cache invalidation: solving stale data and stampede problems
An engineering look at TTL, event-based invalidation, versioning and single-flight patterns, and how cache stampede can take down your origin in production.