The Outbox That Keeps Your Events Honest
Save the order, publish the event, crash in between - the dual-write problem, and how Convey's outbox and inbox turn at-least-once messaging into something you can build a business on.
Archive
Every article
Newest first, grouped by year. Search or filter by topic.
- 388 articles
All
.NET
AI Engineering
Architecture
Azure & DevOps
Data Warehousing
Engineering Practice
Microservices
Web APIs
Cards
List
Vulnerability Metrics Are an Operations Discipline
Vulnerability density, severity-tiered remediation clocks and hotspot review coverage - running the findings backlog as an operational queue with owners and deadlines, not an annual scare.
DNS Zone: One Module, Many Optional Shapes
The public DNS template exposes every record type - A, AAAA, CAA, CNAME, MX, PTR, SRV, TXT - as an optional object list on a single module, and its twelve-kilobyte README quietly becomes the reference manual. When one module with optional shapes beats fifteen micro-modules, and when it stops being true. Part 13 of the Terraform Module Catalog.
The Files That Keep Changing
Code churn, defect density and hotspot scoring - reading a repository's git history to find the files most likely to break next, and the thresholds I use to decide when movement has become rework.
Commands Over gRPC, No Proto Files
FastEndpoints.Messaging.Remote executes the same ICommand on another machine over gRPC, serialized with MessagePack instead of protobuf contracts - plus an event hub mode that turns a server into a lightweight broker. Reading it explains both the magic and the trade. Part 12 of FastEndpoints in Depth.
Coverage Is Not Testing
Line, branch and condition coverage each assert less than you think - mutation testing is the honesty check that reveals whether your suite would actually notice a bug, and the thresholds worth enforcing are conventions, not laws.
What FeedR Gets Right, and What Production Would Add
The retrospective: after ten parts inside DevMentors' real-time feeds sample, the honest scorecard - five decisions worth stealing for real systems, the sharp edges catalogued in one place, and the concrete backlog that separates this teaching codebase from a production deployment. Part 11 closes the FeedR deep dive.
Coupling, Cohesion and the Main Sequence
Measuring the lines between the boxes - afferent and efferent coupling, the instability ratio, Robert C. Martin's main sequence, and the cohesion metric that catches a class doing two jobs, plus where every one of them needs a human to overrule it.
Commands and Events on the Wire
Convey's broker-CQRS bridge is seventy lines of code that make the same handler serve local dispatch and RabbitMQ delivery — one line per subscription, correlation context riding along for free.
Front Door: Edge TLS, WAF Links and Two Instructive Defects
The classic Front Door template does global entry properly - custom domains behind WAF policies, Key Vault TLS, HTTP-to-HTTPS redirect rules - and then teaches by counterexample: a hardcoded ARM resource ID where a lookup map should be, and a missing environment key silently swallowed by a forgiving default. Part 12 of the Terraform Module Catalog.
The Maintainability Index and the Debt Ratio
The composite score that hides more than it shows and the percentage that translates code quality into money - where the maintainability index formula came from, why its logarithms flatten real problems, and how the technical debt ratio earns its place as a trend.
An Event Bus and a Command Bus, No Extra Package Required
FastEndpoints ships in-process messaging that covers most MediatR use cases - events with three wait modes, commands with middleware pipelines, generic and streaming handlers. The dispatch code makes some sharp lifetime choices worth knowing. Part 11 of FastEndpoints in Depth.
Cyclomatic Counts Paths, Cognitive Counts Pain
Two complexity metrics that look interchangeable and answer different questions - McCabe's 1976 path count is a testing instrument, SonarSource's cognitive score is a readability instrument, and only one of them earns a place in the merge gate.
Six Services on a Laptop
FeedR ships three ways to run its six services - pm2, Project Tye, and a deliberately split pair of docker compose files - plus per-service Dockerfiles and an environment-keyed configuration layer that lets the same build run on localhost and inside a compose network. Part 10 of the FeedR deep dive.
When a Measure Becomes a Target
Dashboards go green while codebases rot - the opener of a fourteen-part series on measuring code quality and software delivery without staging the theatre that ruins both.
Retries, Dead Letters, and the Plugin Chain
What Convey's RabbitMQ subscriber actually does when your handler throws - fixed-interval retries, an exception-to-message escape hatch, a dlx- safety net, and one plugin that can quietly defeat all of it.
Application Gateway: When Every Input Is a List of Objects
Inside the module with the biggest surface in the catalog: listener, pool, settings and probe objects cross-referenced by name, Key Vault TLS through a user-assigned identity, a private-by-default frontend - and a deliberately invalid placeholder IP that turns an unfilled template into a loud failure. Part 11 of the Terraform Module Catalog.
What a Scorecard Cannot See
The honest retrospective on two weighted competency frameworks - the blank templates at their centre, the qualities no weight can hold, the situations where a weighted model is the wrong tool, and what I would build differently.
Teaching NSwag What Your Endpoints Mean
FastEndpoints' Swagger support is a thousand-line translation layer that knows which DTO properties are route params, which FluentValidation rules are schema constraints, and which endpoint versions belong in which document. Part 10 of FastEndpoints in Depth.
Weighted Planning and the 90-Day Loop
How the framework becomes a form and the form becomes a plan - per-sub-discipline ratings with evidence, a 500-point ceiling, a gap-times-weight priority formula and a 90-day goal loop, walked through with a worked example.
The Supporting Disciplines
The four disciplines that operate across my AI competency stack rather than inside it - agent architecture, evaluation, governance and domain translation - and why 37 of the framework's 100 points live outside the core.
What Two Hundred Copies Teach You
The honest retrospective that closes the series - how template scaffolding metastasizes mistakes, why version pinning without governance curates a museum, what secrets in locals really cost, and the upgrade list I would run on this estate today. Part 12 of the Terraform on Azure series.
Testing a Message You Can't Await
How do you assert that an endpoint published an event when publishing is a side effect with no return channel? FeedR's end-to-end test answers with WebApplicationFactory, a real Redis round-trip, and a TaskCompletionSource - plus the timeout and determinism traps the sample leaves open. Part 9 of the FeedR deep dive.
Why Specification Weighs 22 and Prompt Craft 8
The four core disciplines of my AI engineering framework and the argument behind their weights - why the prompt sits at the bottom of the stack at 8 points, why the specification sits at the apex at 22, and the sentence in my own document that its own table contradicts.
Running the Warehouse — the V7 Runbook
A pile of subsystems isn't an operable system until you can run it. KC Star's docs distil into a runbook - daily health checks, weekly index hygiene, monthly reindex and security review, PostgreSQL tuning, backups, and one consistency check that catches a whole class of corruption.