Manifesto

I would like MTTL to become a reference, and I know that is not something you can simply declare. These are the rules I hold myself to on every page. If you find one I have broken, tell me.

I tie every topic back to the rest of the company

Observability does not belong to IT. A technical decision has a cost, it weighs on people, it has organizational consequences, and it serves or fails a business line. So every article and course I write ends with a Bridges box saying concretely what the topic changes for the business, the organization, teams and business lines. When I cannot fill that box, I take it as a sign the topic is not ready, and I do not publish it. See Bridges.

I write from what I have seen

Every article starts from a problem met in production or on an engagement, never from a product sheet. When a claim rests on my own experience, I say so.

I publish the configuration I actually ran

Published configurations are complete and tested. They are versioned in a GitLab lab you can clone and run. If I ever publish a configuration I have not tested, the article says so.

I source figures or flag them as illustrative

A figure is either sourced or presented as an illustrative scenario. I would rather give an honest order of magnitude than a spectacular percentage.

I hold every vendor to the same standard

Commercial and open source solutions go through the same criteria. My conflicts of interest are disclosed at the bottom of the page.

I write in French, and I translate

French is my first writing language. Structural pages exist in English, and I translate articles as I go. A page not yet translated is shown in its original language, with a notice.

I date and revise

Every page shows its last update date. Some topics move fast, such as OpenTelemetry semantic conventions or LLM observability. I revise them and note the revision.