A coherent description
“The fictional survey covering 2024 has 137 respondents. It was published in June 2025 and consulted in October 2026.”
A useful page distinguishes a verified fact, a proposed method and an illustration. Here is how to read the site’s content and where its reference data comes from.
| Collection | Entries | How to interpret it |
|---|---|---|
| Web directory | 148 | Resource names, official destinations, category, scope and practical uses. Inclusion does not establish a partnership or a product ranking. |
| Web glossary | 660 | Definitions and practical notes. Similar terms have separate entries when their meaning differs; a definition does not replace a specification. |
| Practical guides per language | 102 | French and English editions. Methods and proposed checks are adapted to each project; they are not certificates or automatic release approval. |
| Mainland municipalities | 34,746 | Administrative identities, postal codes and contextual data. The catalogue does not imply an agency or a customer in every municipality. |
Technical explanations use primary references where relevant: standards, official documentation and public authorities. A linked source supports the subject identified beside it; it does not validate every statement on the website. Provider pricing, availability and contract terms must be checked at the time of a decision.
A successful link check establishes that an address responded during that check. It does not guarantee future availability or product suitability. Conversely, a bot restriction, timeout or login requirement does not establish that a service has closed.
Administrative names and codes use the INSEE Official Geographical Code edition for 1 January 2026. Postal codes, commune population values and centre coordinates come from the French administrative-boundaries API. The regional and departmental population tables use INSEE’s December 2025 publications, with 2023 as the reference year. The catalogue contains 13 regions and 96 mainland departments; it does not cover every overseas territory.
The catalogue distinguishes commune identifiers from postal codes. Population values are contextual indicators; an extraction date is not a census reference year. Centre coordinates locate a municipality approximately, rather than an agency address or an entrance.
The check recorded on November 5, 2026 covers 34 746 municipalities. It compares catalogue fields with the public API and administrative identifiers with the 2026 Official Geographical Code files.
Insee commune file: population at 1 January 2023, boundaries at 1 January 2025, effective from 1 January 2026. Matched by commune code; consulted 22 October 2026. Paris, Lyon and Marseille populations are reconstructed by summing their municipal arrondissements in the Insee file. These subdivisions differ from administrative arrondissements. Insee · Municipal population 2023.
The project examples section presents 200 educational profession scenarios. These possible situations are not client references, delivered websites or measured results. Career profiles describe tasks and exercises without announcing vacancies or guaranteed salaries. Service scope is agreed for each project.
Illustrations convey a subject and may depict fictional settings. They are not evidence of a real place, person or client project. Third-party photographs retain the required credits and usage conditions. A resource logo identifies its publisher without implying endorsement.

A precise claim should be traceable without repeating all the research. For important data, retain the relevant passage, primary URL, document date, access date and person responsible for follow-up.
Record scope: territory, software version, population, unit and period. Administrative data does not establish commercial availability. A result measured on one example does not automatically describe every use.
When correcting a claim, check other occurrences, titles, metadata, structured data, translations and related links. Separate editorial review from availability checks: HTTP 200 demonstrates a response, not content accuracy.
For media, add author, source page, licence and attribution requirements. Identify AI illustrations clearly. For logos, check the current name and represented identity before reusing an old file.
Set the next review according to change risk: deadline, product change or an error report. If a source becomes unavailable, record uncertainty before replacing or removing data. Do not turn an unknown value into zero.
In this fictional example, a survey has 137 respondents. Recently consulting it changes neither the observed period nor its sample size.
The answers describe the survey period rather than the current situation.
The document makes results from an earlier period available.
This records when the document was read without updating its results.
“The fictional survey covering 2024 has 137 respondents. It was published in June 2025 and consulted in October 2026.”
“In 2026, 137 people answered the survey.” The consultation date has been mistaken for the date of the answers.
Check the population studied, scope, question, unit and method. Two values under the same heading can describe different groups. If one of these details is missing, state the limitation rather than calculating a misleading change.
Find the edition and passage supporting the claim. Compare the new information, then correct the text and affected languages when the underlying facts change. A layout change alone is not new evidence.
Provide both public addresses, the relevant passage, the source date or edition and the difference observed. Distinguish an established error from information you have been unable to confirm.
Method reference: W3C best practices recommend describing datasets and their provenance. The dates and 137 respondents in this timeline are invented and do not come from that reference. W3C · Data on the Web Best Practices · consulted on 15 October 2026.
The 19 recent guides offer methods and checks in all ten languages. The eleven glossary topics also have native reading paths with selected definitions, distinctions and examples. French and English retain some additional detail. Older translations are not described as having undergone exhaustive human review.
Send the page address, the disputed passage and a dated primary source when available. For a broken link, include the destination and observed behaviour. For a usage problem, describe the task, device and expected result without sending passwords or personal customer data.
Send a correctionMunicipal population on 1 January 2023, published on 29 December 2025 and effective from 1 January 2026. The statistical boundaries are those of 1 January 2025; this website’s commune catalogue uses COG 2026.
| Region | Insee code | Municipal population 2023 |
|---|---|---|
| Île-de-France | 11 | 12 463 067 |
| Auvergne-Rhône-Alpes | 84 | 8 205 557 |
| Nouvelle-Aquitaine | 75 | 6 150 451 |
| Occitanie | 76 | 6 124 653 |
| Hauts-de-France | 32 | 5 992 194 |
| Grand Est | 44 | 5 563 378 |
| Provence-Alpes-Côte d'Azur | 93 | 5 218 960 |
| Pays de la Loire | 52 | 3 907 156 |
| Bretagne | 53 | 3 449 370 |
| Normandie | 28 | 3 345 842 |
| Bourgogne-Franche-Comté | 27 | 2 802 670 |
| Centre-Val de Loire | 24 | 2 587 031 |
| Corse | 94 | 355 486 |
Municipal population counts people whose usual residence is within the area. It is neither a customer count nor a forecast of enquiries. Shares use official populations from the same reference year and are rounded to one decimal place.
Sources : Insee. Consulted on 22 October 2026.