Practical insights and guides on PHP/Symfony, React/Next.js, legacy modernization, and software architecture, drawn from 18+ years of building web applications for European businesses.
Next.js Middleware runs before every request, and the Edge Runtime it runs in rejects most of what your team knows about Node.js. This post covers the production patterns that hold up: WebCrypto JWT verification, matcher configuration, redirect loops, and the debugging workflow when console.log goes silent.
Read moreTechnical due diligence can claw back valuation or kill a deal outright. Here are the five areas reviewers consistently flag, how to score your codebase before a buyer does, and how to present known issues without spooking anyone.
Read moreMonorepos promise shared code and faster builds, then deliver dependency conflicts and CI runs that rebuild everything on every commit. Here is a Turborepo workspace setup for a 3-10 engineer team that keeps builds correct, caches working, and CI under two minutes.
Read moreTutorial Dockerfiles work locally and fail in production in predictable ways: cache permissions, stale opcache responses, exhausted FPM pools, and health checks that lie. Here is a Symfony 7 container setup that survives real traffic.
Read moreA tech stack consultant evaluates your technology choices with code-level evidence and documented trade-offs. Here is what a real engagement looks like, what it costs in the EU market, and the questions that separate independent advisors from resellers.
Read moreEnterprise security questionnaires range from 20 to 500 questions, but the same 50 controls decide the outcome. Here is what buyers really check, what evidence they accept, and how to build a reusable evidence pack.
Read moreThe React data fetching landscape has split into two camps: Server Components and React Query. Most teams pick one and force everything through it. The better answer is a decision framework built on the four data patterns every B2B SaaS actually has.
Read morePlatform engineering is not just for companies with a dedicated platform team and a Kubernetes cluster. Here is how a 5-20 engineer SaaS team builds an internal developer platform from conventions, pipeline templates, and Kamal instead.
Read moreGraphQL gets adopted for the wrong reasons and abandoned for the right ones. Here are the four scenarios where it genuinely beats REST in production, the four where it backfires, and how to defend against the N+1 problem before it takes down your API.
Read moreMost teams enable PHPStan at level 0 and never move it, because jumping to level 8 surfaces thousands of errors at once. Here is the incremental migration strategy that gets real codebases there while feature work keeps shipping.
Read more