<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/"><channel><title>Elie Nehme — Ideas and publications</title><link>https://elienehme.fr/en/</link><description>Ideas published on elienehme.fr and publications on Medium and Hashnode, by Elie Nehme</description><language>en</language><managingEditor>contact@elienehme.fr (Elie Nehmé)</managingEditor><atom:link href="https://elienehme.fr/en/rss.xml" rel="self" type="application/rss+xml"/><item><title>Proactive without deciding for you</title><link>https://elienehme.fr/en/ideas/proactive-without-deciding/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/proactive-without-deciding/</guid><description>As soon as an action becomes recurrent, the agent can prepare it — but a system that speeds up the routine puts vigilance to sleep, while one that isolates the exception wakes it up</description><pubDate>Mon, 09 Nov 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Conversation frames the gesture</title><link>https://elienehme.fr/en/ideas/conversation-frames-the-gesture/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/conversation-frames-the-gesture/</guid><description>Conversation does not replace the gesture, it replaces the navigation that leads to the gesture: on high-frequency repetitive tasks, muscle memory is a feature</description><pubDate>Mon, 02 Nov 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>SharedCore: un cœur produit partagé, des modèles de base adaptés, une évolution commune</title><link>https://elie29.hashnode.dev/sharedcore-un-c-ur-produit-partag-des-mod-les-de-base-adapt-s-une-volution-commune</link><guid isPermaLink="true">https://elie29.hashnode.dev/sharedcore-un-c-ur-produit-partag-des-mod-les-de-base-adapt-s-une-volution-commune</guid><description>Hundreds of customers, each running its own variant of the software: the vendor pays for every legal change, fix and upgrade once per customer. SharedCore is a strategy to bring them onto one common product, adapted through configuration and never through code. In the case it draws on, migrating a customer went from 70 days of work to 6.</description><pubDate>Mon, 02 Nov 2026 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Free conversation, controlled actions</title><link>https://elienehme.fr/en/ideas/free-conversation-controlled-actions/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/free-conversation-controlled-actions/</guid><description>The rendering can be generative because the write is not: the guardrail is no longer a button hidden in an interface, it is a right evaluated whatever the wording</description><pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Réécrire un projet avec l’IA: plus d’une cinquantaine de régressions malgré une documentation complète</title><link>https://elie29.hashnode.dev/reecrire-un-projet-avec-l-ia-plus-d-une-cinquantaine-de-r-gressions-malgr-une-documentation-compl-te</link><guid isPermaLink="true">https://elie29.hashnode.dev/reecrire-un-projet-avec-l-ia-plus-d-une-cinquantaine-de-r-gressions-malgr-une-documentation-compl-te</guid><description>A project built a few years ago, with complete documentation, specs, tests and migration history, was rewritten with AI for another stack. It reached qualification testing with more than fifty regressions. Having all the information does not guarantee it will be used well.</description><pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>PHP2Shell: une faille repérée en 2024, démontrée en 2026</title><link>https://elie29.hashnode.dev/php2shell-une-faille-rep-r-e-en-2024-d-montr-e-en-2026</link><guid isPermaLink="true">https://elie29.hashnode.dev/php2shell-une-faille-rep-r-e-en-2024-d-montr-e-en-2026</guid><description>PHP2Shell, the name given to two critical PHP flaws patched on 21 October 2026, shows how far apart spotting a defect and proving an attack can be: a weakness in out-of-memory handling, flagged in 2024, led to remote code execution only once AI-assisted research linked it to a second flaw.</description><pubDate>Thu, 22 Oct 2026 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>The cost of a representation: 31 days yesterday, 2 seconds today</title><link>https://elienehme.fr/en/ideas/the-cost-of-a-representation/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/the-cost-of-a-representation/</guid><description>First use case, walked through end to end: same query, same data set, two forms of rendering, and no deployment between the two</description><pubDate>Tue, 20 Oct 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Brownfield vs. greenfield is the wrong debate in the age of AI</title><link>https://elienehme.fr/en/ideas/brownfield-vs-greenfield-is-the-wrong-debate/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/brownfield-vs-greenfield-is-the-wrong-debate/</guid><description>Brownfield or greenfield predicts little: what matters is whether an agent finds the right context within clear boundaries, from day 1 to day 1,000</description><pubDate>Mon, 19 Oct 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>The conversational journey: generative AI does not shift the problem, it shifts the solution</title><link>https://elienehme.fr/en/ideas/the-conversational-journey/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/the-conversational-journey/</guid><description>Whole production cycles were paid for to change a representation, while the data structure had not moved. It is not the screen that needs redoing, it is the idea that it must be frozen</description><pubDate>Tue, 13 Oct 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>We would still be coding in QBasic</title><link>https://elienehme.fr/en/ideas/we-would-still-be-coding-in-qbasic/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/we-would-still-be-coding-in-qbasic/</guid><description>Publisher, PowerBuilder, no-code: every tool was meant to end software development. Each took a task, never the craft: AI will too, if the foundations hold</description><pubDate>Thu, 08 Oct 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Plus le projet grossit, plus l’IA se trompe: le bounded context comme investissement</title><link>https://elie29.hashnode.dev/plus-le-projet-grossit-plus-l-ia-se-trompe</link><guid isPermaLink="true">https://elie29.hashnode.dev/plus-le-projet-grossit-plus-l-ia-se-trompe</guid><description>On a green-field project, generative AI works wonders; as the project grows, its answers lose accuracy. The model has not changed: what has grown is the context it has to hold. Context Budget, the method I have introduced on several projects, splits the code into bounded contexts the AI can hold in full: the split is paid for once, and each request then pays only for its own context.</description><pubDate>Fri, 02 Oct 2026 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>When everyone ships a website</title><link>https://elienehme.fr/en/ideas/when-everyone-ships-a-site/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/when-everyone-ships-a-site/</guid><description>Vibe coding lets anyone build a website, but it cannot make that website unique and safe</description><pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Corriger ou atténuer: le CRA ne prévoit que deux réponses</title><link>https://elie29.hashnode.dev/corriger-ou-attenuer-le-cra-ne-laisse-que-deux-reponses</link><guid isPermaLink="true">https://elie29.hashnode.dev/corriger-ou-attenuer-le-cra-ne-laisse-que-deux-reponses</guid><description>Since 11 September 2026, the Cyber Resilience Act (CRA) has required an early warning within 24h for an actively exploited vulnerability in a product, including products already on the market. Meeting that deadline depends on work done well before the incident: in the delivery chain, in the number of versions maintained, and in whether products can be mitigated without a developer.</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>A vision that is never executed is just a dream</title><link>https://elienehme.fr/en/ideas/a-vision-that-is-not-executed-is-not-one/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/a-vision-that-is-not-executed-is-not-one/</guid><description>A vision is only worth what teams can follow: execution that is faithful and practicable, a pragmatic trajectory rather than fine ideas nobody can keep up with</description><pubDate>Wed, 23 Sep 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>From Legacy Artifacts to Served Context: My Six-Stage Method</title><link>https://elie29.medium.com/from-legacy-artifacts-to-served-context-my-six-stage-method-09f668052904</link><guid isPermaLink="true">https://elie29.medium.com/from-legacy-artifacts-to-served-context-my-six-stage-method-09f668052904</guid><description>Six stages to build a knowledge graph from C, PL/SQL, PHP and twenty years of documents, keep it true release after release and serve it to an LLM: deterministic collectors, a SHACL gate, append-only facts, and a graph that selects before the model reads.</description><pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Why reinvent what already exists, or what will be native tomorrow</title><link>https://elienehme.fr/en/ideas/reinventing-what-already-exists/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/reinventing-what-already-exists/</guid><description>Rebuilding generic skills that plugins and integrations already provide, and that the next models will absorb, is rarely the right investment</description><pubDate>Mon, 14 Sep 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Ontology-Based Context Engineering: When RAG Is Not Enough</title><link>https://elie29.medium.com/ontology-based-context-engineering-when-rag-is-not-enough-124e6b1de6bb</link><guid isPermaLink="true">https://elie29.medium.com/ontology-based-context-engineering-when-rag-is-not-enough-124e6b1de6bb</guid><description>Why vector RAG failed on a two-million-line legacy system, and how a small ontology kept in Git, with RDF, SHACL and SPARQL, gives an LLM release-filtered context and a traceable path.</description><pubDate>Fri, 11 Sep 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>If AI replaces developers, who will decide what it builds?</title><link>https://elienehme.fr/en/ideas/if-ai-replaces-developers-who-decides/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/if-ai-replaces-developers-who-decides/</guid><description>&quot;AI will replace everyone&quot;, or, in the targeted version, &quot;developers&quot;. Replacing a profession because its tool can now be generated confuses the two</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>I Tried BMAD on a Real Feature</title><link>https://elie29.medium.com/i-tried-bmad-on-a-real-feature-69afd0ba53b3</link><guid isPermaLink="true">https://elie29.medium.com/i-tried-bmad-on-a-real-feature-69afd0ba53b3</guid><description>BMAD with Superpowers on a real feature across a Java backend and an Angular frontend: what a formal method adds to a living specification, where the two collide, and what to adopt, adapt or skip.</description><pubDate>Mon, 24 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>The Living Specification: How I Keep AI-Written Code Correct</title><link>https://elie29.medium.com/the-living-specification-how-i-keep-ai-written-code-correct-907be05c9eed</link><guid isPermaLink="true">https://elie29.medium.com/the-living-specification-how-i-keep-ai-written-code-correct-907be05c9eed</guid><description>A living specification, invariants, adversarial AI reviews and real verification keep AI-written code correct, and engineering shifts from writing code to exercising judgement. Second part of a two-article series.</description><pubDate>Wed, 19 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Code is a means, not an end</title><link>https://elienehme.fr/en/ideas/code-is-a-means-not-an-end/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/code-is-a-means-not-an-end/</guid><description>A developer who only writes code will have less work, or none at all. The question is not whether AI writes the code, but what the developer brings beyond it</description><pubDate>Sun, 16 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>AI Writes the Code. So What Is the Engineer’s Job Now?</title><link>https://elie29.medium.com/ai-writes-the-code-so-what-is-the-engineers-job-now-79fa3fdf1b88</link><guid isPermaLink="true">https://elie29.medium.com/ai-writes-the-code-so-what-is-the-engineers-job-now-79fa3fdf1b88</guid><description>How I asked Claude Code to attack a design instead of building it, the three bugs the adversarial review found before any code, and why engineering is moving from code production to technical validation.</description><pubDate>Sat, 15 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Sustainable competitive advantage, when AI writes the code</title><link>https://elienehme.fr/en/ideas/competitive-advantage-when-ai-writes-the-code/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/competitive-advantage-when-ai-writes-the-code/</guid><description>As AI writes more and more of the code, companies need to rethink where their sustainable competitive advantage will come from</description><pubDate>Mon, 10 Aug 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Advanced patterns never make up for a fragile foundation</title><link>https://elienehme.fr/en/ideas/patterns-never-make-up-for-a-fragile-foundation/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/patterns-never-make-up-for-a-fragile-foundation/</guid><description>Hexagonal architecture, clean architecture, CQRS: powerful patterns, often shown off without the good development practice that holds them up</description><pubDate>Tue, 02 Jun 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>What AI changes, and what it doesn’t, in software engineering</title><link>https://elienehme.fr/en/ideas/what-ai-changes-and-what-it-doesnt/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/what-ai-changes-and-what-it-doesnt/</guid><description>A short list of what AI has changed in the work, a shorter one of what it has not, and what that means for the engineer’s job</description><pubDate>Mon, 02 Feb 2026 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Quand false ne veut rien dire</title><link>https://elie29.hashnode.dev/quand-false-ne-veut-rien-dire</link><guid isPermaLink="true">https://elie29.hashnode.dev/quand-false-ne-veut-rien-dire</guid><description>When a method can fail for several reasons, returning true or false throws away the one thing the caller needs: the cause. Why the boolean return is an anti-pattern, where it comes from, and what to return instead.</description><pubDate>Thu, 18 Dec 2025 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>An LLM has no memory, the chatbot does</title><link>https://elienehme.fr/en/ideas/an-llm-has-no-memory-the-chatbot-does/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/an-llm-has-no-memory-the-chatbot-does/</guid><description>The model forgets everything between two requests: the chat application sends the conversation back each time. What users think the AI remembers is what the application chose to resend</description><pubDate>Tue, 18 Nov 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Mesurer une démarche DevSecOps: un modèle de maturité sur preuves et quatre familles d’indicateurs</title><link>https://elienehme.fr/writing/mesurer-une-demarche-devsecops/</link><guid isPermaLink="true">https://elienehme.fr/writing/mesurer-une-demarche-devsecops/</guid><description>DevSecOps is declared everywhere and measured almost nowhere. Two instruments fix that: the OWASP maturity model, where an activity counts only with its evidence, and a monthly dashboard on four axes, each indicator with a source and a target.</description><pubDate>Tue, 23 Sep 2025 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>RabbitMQ is a postman, Kafka is a ledger</title><link>https://elienehme.fr/en/ideas/rabbitmq-is-a-postman-kafka-is-a-ledger/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/rabbitmq-is-a-postman-kafka-is-a-ledger/</guid><description>A broker that delivers and forgets, a log that keeps and lets consumers read again: choose by what you need after delivery, not by throughput</description><pubDate>Tue, 09 Sep 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Roll out in waves, from a pilot</title><link>https://elienehme.fr/en/ideas/roll-out-by-waves-from-a-pilot/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/roll-out-by-waves-from-a-pilot/</guid><description>No change should happen by chance, in the code or in an organisation. Five phases, from a pilot to a decision, turn a good practice into a standard that holds</description><pubDate>Tue, 10 Jun 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Kill Feature Branching or Die Trying</title><link>https://elie29.medium.com/kill-feature-branching-or-die-trying-5da0fe41b806</link><guid isPermaLink="true">https://elie29.medium.com/kill-feature-branching-or-die-trying-5da0fe41b806</guid><description>Why long-lived feature branches work against continuous delivery, what feature toggles offer instead, and a .NET sample that puts a new feature behind a request header in five steps.</description><pubDate>Mon, 12 May 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>La CI/CD n’est pas une priorité !</title><link>https://elie29.hashnode.dev/la-cicd-nest-pas-une-priorite</link><guid isPermaLink="true">https://elie29.hashnode.dev/la-cicd-nest-pas-une-priorite</guid><description>Setting up CI/CD is often a team’s first DevOps reflex, and often premature: what a real pipeline is, what must be settled first, and why the persistence layer and the infrastructure deserve their own care.</description><pubDate>Wed, 07 May 2025 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Justifier le Craftsmanship auprès des DSI: ROI, contexte et impact réel</title><link>https://elie29.hashnode.dev/justifier-le-craftsmanship-aupres-des-dsi-roi-contexte-et-impact-reel</link><guid isPermaLink="true">https://elie29.hashnode.dev/justifier-le-craftsmanship-aupres-des-dsi-roi-contexte-et-impact-reel</guid><description>A CTO who cannot demonstrate the ROI of Sonar and GitLab, a director who cannot quantify the benefits of pair programming: why Craftsmanship must be justified by the business problems it solves, and measured on two levels.</description><pubDate>Thu, 01 May 2025 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Building a Resilient Site-Aware Synchronization</title><link>https://elie29.hashnode.dev/building-a-resilient-site-aware-synchronization</link><guid isPermaLink="true">https://elie29.hashnode.dev/building-a-resilient-site-aware-synchronization</guid><description>Several on-premise sites, each with its own database, that must keep working through days of network outage and still converge: an outbox table, a RabbitMQ cluster, a fanout exchange, idempotent consumers and a golden source for replaying events.</description><pubDate>Mon, 14 Apr 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>One rotating champion per technology</title><link>https://elienehme.fr/en/ideas/one-rotating-champion-per-technology/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/one-rotating-champion-per-technology/</guid><description>A single expert per technology is a risk that looks like a strength. A shared calendar of rotating champions spreads the knowledge of each technology across the team and shows where the gaps are</description><pubDate>Tue, 18 Mar 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Nobody approves their own merge request</title><link>https://elienehme.fr/en/ideas/nobody-approves-their-own-merge-request/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/nobody-approves-their-own-merge-request/</guid><description>A merge request approved by its author is a formality, not a review. The rule fits in one line of a team’s Git conventions, and it matters more as assistants write more of the code</description><pubDate>Tue, 21 Jan 2025 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Prompting for People Who Are Not Engineers: What Actually Moves the Answer</title><link>https://elienehme.fr/en/writing/prompting-for-people-who-are-not-engineers/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/prompting-for-people-who-are-not-engineers/</guid><description>Temperature and Top P are real, but most people never touch them: the prompt is the lever they hold. What a course for non-engineers teaches about criteria, prompt structures and the four ways a prompt fails.</description><pubDate>Tue, 26 Nov 2024 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Standardise the collection, not the tool</title><link>https://elienehme.fr/en/ideas/standardise-the-collection-not-the-tool/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/standardise-the-collection-not-the-tool/</guid><description>With OpenTelemetry, the standard belongs in what touches every service, its instrumentation and its collector, and the backend becomes a choice each team can change</description><pubDate>Tue, 15 Oct 2024 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Le dernier kilomètre du self-service est culturel</title><link>https://elienehme.fr/writing/le-dernier-kilometre-du-self-service/</link><guid isPermaLink="true">https://elienehme.fr/writing/le-dernier-kilometre-du-self-service/</guid><description>Moving from a ticket-driven platform to a self-service catalogue costs less effort than the move to the platform did, and most of that effort is cultural. What automating Kafka topics shows about who decides, who is trusted and what the platform team stops doing.</description><pubDate>Tue, 17 Sep 2024 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Security sets the rule, operations runs it</title><link>https://elienehme.fr/en/ideas/security-sets-the-rule-operations-runs-it/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/security-sets-the-rule-operations-runs-it/</guid><description>Governing secrets takes two roles that work closely together and never merge: security writes the policy, operations runs the tool, and both choose it together</description><pubDate>Tue, 11 Jun 2024 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Secrets on Kubernetes: Vault as a Sidecar, a CSI Driver or an Operator</title><link>https://elienehme.fr/en/writing/secrets-on-kubernetes-sidecar-csi-or-operator/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/secrets-on-kubernetes-sidecar-csi-or-operator/</guid><description>A Kubernetes Secret is only base64, not a vault. A review of eight secrets managers recommended HashiCorp Vault, which still left a choice: an agent sidecar, the Secrets Store CSI driver or the Vault Secrets Operator, and each one puts the secret in a different place.</description><pubDate>Tue, 21 May 2024 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Moving to the cloud is not moving the servers</title><link>https://elienehme.fr/en/ideas/moving-to-the-cloud-is-not-moving-the-servers/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/moving-to-the-cloud-is-not-moving-the-servers/</guid><description>Putting everything on infrastructure as a service, or containerising everything as is with huge images, is not a migration: it is a move that costs more</description><pubDate>Sat, 02 Mar 2024 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Le délai est dans les files d’attente: cartographier la chaîne CI/CD</title><link>https://elienehme.fr/writing/le-delai-est-dans-les-files-d-attente/</link><guid isPermaLink="true">https://elienehme.fr/writing/le-delai-est-dans-les-files-d-attente/</guid><description>On a CI/CD chain mapped end to end, a deployment to one environment took 802 minutes, 732 of them spent waiting. How to build the map, why the queues dominate, and what a target state really removes.</description><pubDate>Tue, 13 Feb 2024 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Un, trois ou six clusters? Gouverner Kubernetes pour 2 500 applications</title><link>https://elienehme.fr/writing/un-trois-ou-six-clusters-kubernetes/</link><guid isPermaLink="true">https://elienehme.fr/writing/un-trois-ou-six-clusters-kubernetes/</guid><description>One cluster for 2,500 applications is cheap until an upgrade or a security flaw reaches all of them at once. How the choice between one, three and six clusters led to eight permanent clusters split by business domain, plus a sandbox with an expiry date.</description><pubDate>Tue, 07 Nov 2023 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>A user story is not a technical task list</title><link>https://elienehme.fr/en/ideas/a-user-story-is-not-a-technical-task-list/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/a-user-story-is-not-a-technical-task-list/</guid><description>Code, test, describe the implementation: repeated in every user story, these tasks bring no business value, and Jira is not a development tool</description><pubDate>Tue, 10 Oct 2023 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Stratégie de déploiement Blue/Green</title><link>https://elie29.hashnode.dev/strategie-de-deploiement-bluegreen</link><guid isPermaLink="true">https://elie29.hashnode.dev/strategie-de-deploiement-bluegreen</guid><description>Beyond the textbook definition: the decisions a team must take, component by component, for Blue/Green to work in practice, from persistence compatibility and sessions to decommissioning and rollback.</description><pubDate>Sun, 27 Aug 2023 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>One Collector, Any Backend: Standardising Telemetry with OpenTelemetry</title><link>https://elienehme.fr/en/writing/one-collector-any-backend/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/one-collector-any-backend/</guid><description>Wiring every service straight to one tracing backend is the fastest way to start, and the hardest to change. Putting an OpenTelemetry Collector in between moves the choice of backend out of the services, as long as you read the stability label of each component you depend on.</description><pubDate>Tue, 14 Mar 2023 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Why architecture reviews often fail</title><link>https://elienehme.fr/en/ideas/why-architecture-reviews-often-fail/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/why-architecture-reviews-often-fail/</guid><description>Reviews happen at the wrong time, with the wrong question. What has made them useful in my experience</description><pubDate>Mon, 06 Feb 2023 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>One Kafka platform, not one cluster per project</title><link>https://elienehme.fr/en/ideas/one-shared-kafka-platform/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/one-shared-kafka-platform/</guid><description>A Kafka cluster per project looks like autonomy and costs like a debt: one shared offer, run by one central team, with every topic declared as code</description><pubDate>Tue, 11 Oct 2022 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Un radar technologique trimestriel: installer, versionner, décider</title><link>https://elienehme.fr/writing/un-radar-technologique-trimestriel/</link><guid isPermaLink="true">https://elienehme.fr/writing/un-radar-technologique-trimestriel/</guid><description>A technology radar is only useful if each ring stands for a decision and its history can be traced. How to install Zalando’s open-source radar, keep its entries in Git, publish one tagged edition per quarter and move a technology from one ring to another only on evidence.</description><pubDate>Tue, 20 Sep 2022 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Every consumer is idempotent</title><link>https://elienehme.fr/en/ideas/every-consumer-is-idempotent/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/every-consumer-is-idempotent/</guid><description>With Kafka, receiving the same message twice is the normal case, not the exception: every consumer has to be written so that a second delivery changes nothing</description><pubDate>Tue, 12 Jul 2022 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Kafka is not a JMS</title><link>https://elienehme.fr/en/ideas/kafka-is-not-a-jms/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/kafka-is-not-a-jms/</guid><description>Kafka is a distributed log with its own concepts: partitions, offsets, consumer groups, retention. Teams that bring their message-queue reflexes to it get it wrong, so the concepts come first</description><pubDate>Tue, 17 May 2022 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Java functional programming to leverage polymorphism</title><link>https://elie29.hashnode.dev/java-functional-programming-to-leverage-polymorphism</link><guid isPermaLink="true">https://elie29.hashnode.dev/java-functional-programming-to-leverage-polymorphism</guid><description>How a factory’s switch statement grows with every new case, and how a map of functional interfaces, then an enum, keeps its complexity constant however many types you add.</description><pubDate>Wed, 13 Apr 2022 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Apache Kafka along with RxJS operators</title><link>https://elie29.medium.com/apache-kafka-along-with-rxjs-operators-31c00d46c35a</link><guid isPermaLink="true">https://elie29.medium.com/apache-kafka-along-with-rxjs-operators-31c00d46c35a</guid><description>A Kafka processor between two topics must not lose a message while its producer is paused: why takeUntil, switchMap and combineLatest all drop data, and how windowToggle and bufferToggle keep it.</description><pubDate>Thu, 24 Mar 2022 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>The hidden cost of technical decisions</title><link>https://elienehme.fr/en/ideas/the-hidden-cost-of-technical-decisions/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/the-hidden-cost-of-technical-decisions/</guid><description>The price of a technical decision is what it does to every decision that follows. Three questions to ask before making one</description><pubDate>Thu, 11 Nov 2021 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Deploy Quarkus Application with GitLab-CI to Heroku PaaS</title><link>https://elie29.medium.com/deploy-quarkus-application-with-gitlab-ci-to-heroku-paas-72c0e9214669</link><guid isPermaLink="true">https://elie29.medium.com/deploy-quarkus-application-with-gitlab-ci-to-heroku-paas-72c0e9214669</guid><description>From an empty Quarkus project to three Heroku environments with one GitLab-CI pipeline: branch and tag rules, protected variables, dependency caching and free-tier security scans.</description><pubDate>Sat, 03 Apr 2021 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Frontend JWT token storage</title><link>https://elie29.medium.com/frontend-jwt-token-storage-77cbe6dc680b</link><guid isPermaLink="true">https://elie29.medium.com/frontend-jwt-token-storage-77cbe6dc680b</guid><description>Local storage, session storage and plain cookies leave a JWT open to theft: three ways to hold the token in a browser (a closure, a secure cookie, a service worker) and what each one trades away.</description><pubDate>Sun, 14 Mar 2021 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Combien de contrôleurs Jenkins? Votre chaîne d’intégration reflète votre organisation</title><link>https://elienehme.fr/writing/combien-de-controleurs-jenkins/</link><guid isPermaLink="true">https://elienehme.fr/writing/combien-de-controleurs-jenkins/</guid><description>Splitting Jenkins controllers by environment, by organisation, by product line or by technology changes what a failure stops, what teams can see of the flow and what it costs to run. The choice mirrors who owns each product from commit to production, so it is a decision for those who assign that ownership, not primarily for the platform team.</description><pubDate>Tue, 15 Sep 2020 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>What makes a codebase evolvable?</title><link>https://elienehme.fr/en/ideas/what-makes-a-codebase-evolvable/</link><guid isPermaLink="true">https://elienehme.fr/en/ideas/what-makes-a-codebase-evolvable/</guid><description>Evolvability is measured by doing, not read off a diagram. Five unglamorous traits of codebases that age well</description><pubDate>Thu, 02 Apr 2020 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>When Does a RabbitMQ Message Really Survive a Restart?</title><link>https://elienehme.fr/en/writing/when-does-a-rabbitmq-message-survive-a-restart/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/when-does-a-rabbitmq-message-survive-a-restart/</guid><description>A durable queue does not make a message durable. Five settings must agree for a message to survive a restart, and a replicated queue type decides whether it survives the loss of a node.</description><pubDate>Tue, 10 Mar 2020 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Delayed Messages in Kafka: Three Designs for a Slow Consumer</title><link>https://elienehme.fr/en/writing/delayed-messages-in-kafka/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/delayed-messages-in-kafka/</guid><description>One slow client can stall every message behind it, because a Kafka partition is consumed in order. Three designs let a message wait without holding back the others: one consumer group per client, a shared delay topic and tiered retry topics, each with its own costs.</description><pubDate>Tue, 08 Oct 2019 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Le kata plutôt que la règle: diffuser le TDD et la conception par la pratique</title><link>https://elienehme.fr/writing/le-kata-plutot-que-la-regle/</link><guid isPermaLink="true">https://elienehme.fr/writing/le-kata-plutot-que-la-regle/</guid><description>A rule produces tests written after the code; repeated practice changes the habit. How to turn katas into a team ritual that lasts: subjects prepared with the technical leads, a progression by levels, time set aside, and the real project, where the practice is applied.</description><pubDate>Tue, 18 Jun 2019 00:00:00 GMT</pubDate><dc:language>fr</dc:language></item><item><title>Mono.just Runs Too Early: Nothing Happens Until You Subscribe, Except When It Does</title><link>https://elienehme.fr/en/writing/mono-just-runs-too-early/</link><guid isPermaLink="true">https://elienehme.fr/en/writing/mono-just-runs-too-early/</guid><description>Reactor runs nothing before the subscription, yet Mono.just takes a value Java has already computed. The trap behind dead fallbacks, blocked event loops and retries that replay a stale value, and how to catch it in a test.</description><pubDate>Tue, 16 Apr 2019 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Refactoring: from strategy to interpreter pattern with Jest testing tool</title><link>https://elie29.medium.com/refactoring-from-strategy-to-interpreter-pattern-with-jest-testing-tool-58ef7bf0f2fb</link><guid isPermaLink="true">https://elie29.medium.com/refactoring-from-strategy-to-interpreter-pattern-with-jest-testing-tool-58ef7bf0f2fb</guid><description>Evaluating arithmetic expressions with operator precedence: from a switch that breaks the open/closed principle to polymorphism, then a lexer, a parser and an interpreter, each step checked by Jest tests.</description><pubDate>Sun, 25 Nov 2018 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Refactoring: How to avoid nested or several conditionals within a function?</title><link>https://elie29.medium.com/refactoring-how-to-avoid-nested-or-several-conditionals-within-a-function-a25df6c27f62</link><guid isPermaLink="true">https://elie29.medium.com/refactoring-how-to-avoid-nested-or-several-conditionals-within-a-function-a25df6c27f62</guid><description>Two ways out of the pyramid of doom: a small assertion function that chains instead of nesting, and conditions turned into getters so that Math.max replaces the branching.</description><pubDate>Mon, 05 Nov 2018 00:00:00 GMT</pubDate><dc:language>en</dc:language></item><item><title>Object Calisthenics: One level ! But how can we manage to exit a loop ?</title><link>https://elie29.medium.com/object-calisthenics-one-level-but-how-can-we-manage-to-exit-a-loop-4485dbaf7f7d</link><guid isPermaLink="true">https://elie29.medium.com/object-calisthenics-one-level-but-how-can-we-manage-to-exit-a-loop-4485dbaf7f7d</guid><description>Only one level of indentation per method, and still a way out of a loop: array_reduce or recursion instead of a nested early return, with the cost of each measured on fifty thousand elements.</description><pubDate>Tue, 30 Oct 2018 00:00:00 GMT</pubDate><dc:language>en</dc:language></item></channel></rss>