Eco-design declaration
Measured on 17 August 2026
This page states what this site does to limit its environmental footprint, and what it does not. It is published because a conformity score means nothing on its own: the reference framework requires it to come with a declaration that can be checked. Every figure below is computed from files kept in the source repository, not typed into this page.
Conformity
75 %
46 of 61 applicable criteria met, 17 not applicable to this service.
Non-applicable criteria are excluded from the denominator, as the method requires. A criterion awaiting an organisational decision counts as not met: until the decision is taken, it is not held.
The reference framework
We use the RGESN, the French general reference framework for the eco-design of digital services, published by MiNumEco with Arcep and Arcom. Its 2024 version sets out 78 criteria in nine themes. Applying it is mandatory for central government and for local authorities above 50,000 inhabitants. TAMIS falls outside that scope: we do this because it belongs in a European project about data, not because we are required to.
Scope
The public site only. The back-office is deliberately excluded: it is built on a third-party administration framework whose CSS and JavaScript we do not control, and it is used by a small number of people from the TAMIS consortium to publish content. Including it would make a public score depend on choices that are not ours, for a service visitors never see.
What the pages weigh
Measured on the templates that make up the site, not on a sample of pages: weight depends on the template and its media.
| Template | Weight | Requests | DOM nodes | EcoIndex |
|---|---|---|---|---|
| Home | 317 KiB | 14 | 588 | C (69.9) |
| News | 272.3 KiB | 14 | 257 | A (83.6) |
| Events | 217.1 KiB | 11 | 188 | A (86.7) |
| Resources | 218.5 KiB | 11 | 252 | A (84.4) |
| Search | 218.4 KiB | 11 | 231 | A (85.2) |
| About | 278.9 KiB | 14 | 331 | A (80.6) |
| Use cases hub | 221.3 KiB | 11 | 413 | B (77.8) |
| Use case detail | 219.9 KiB | 11 | 306 | A (82.3) |
| Glossary | 223.6 KiB | 11 | 357 | A (80.2) |
| How to join | 220.6 KiB | 11 | 300 | A (82.6) |
| What it costs | 218.2 KiB | 11 | 227 | A (85.4) |
Transferred weight, compression included. Images are counted as a browser would choose them at a reference window of 1280 pixels and a pixel ratio of 1; a high-density screen would take the next variant up. Without a stated convention, an image weight means nothing.
Each page has a ceiling for weight, requests and DOM nodes. The ceilings are checked against the recorded measurement by the test suite, so a page that grows past them fails the build rather than quietly making this page wrong. 400 KiB · 30 requests · 800 nodes
The measurement is a command in the repository, so anyone with the code can repeat it. php artisan eco:mesurer
What the score does not see
The EcoIndex rests on three inputs: DOM nodes, requests and transferred weight. It ignores the processor time spent on your device. An animation running in a loop costs no bytes and loses no points, yet it keeps a device working for as long as the tab is open. Our home page animates sixty elements. They now stop as soon as they leave the screen or the tab goes to the background, which no score will ever credit us for. A site can be rated A and still warm a phone.
The part that matters most
Across the life cycle of digital technology, manufacturing devices weighs far more than the electricity used to run them. So the first thing a site can do is not to push anyone into replacing their phone or laptop. That is why we care about a page working on an old device more than about shaving off a few kilobytes, and why we have not measured that yet: we say so below rather than claim it.
What is in place
- No third-party service on the public site: no external analytics, no embedded video player, no remote fonts. Nothing is loaded from anyone else.
- Audience measurement without cookies, on our own servers, with an irreversible visitor fingerprint rebuilt every day.
- Videos are links, never embedded players: nothing is loaded from a third party until you choose to follow the link.
- Cover images are served at the size they are displayed, in three widths.
- Partner logos are served in a modern format when it is genuinely lighter, with the original kept as a fallback.
- Fonts are limited to the character subsets and weights actually used, and served from our own servers.
- Looping animations stop when they leave the screen or the tab goes to the background.
- Personal data is deleted automatically after 13 or 24 months depending on what it is, which also keeps stored volume down.
- Pages are rendered on the server: your browser receives HTML rather than rebuilding the interface.
- Hosting in France on infrastructure whose energy and water indicators are published, on electricity backed by renewable guarantees of origin.
Where we fall short
Taken from the same grid as the score above, with the reason in each case. Some of these are decisions we have taken and stand by; others are simply not done yet.
Strategy
- 1.1 We have never formally assessed whether this service is worth its environmental cost. The scope was argued on other grounds.
- 1.3 No one is formally accountable for eco-design on this project. That is an organisational decision, and it has not been taken.
- 1.8 The code of this site is not published as open source. The wider TAMIS project does publish open-source connectors (opens in a new tab) Connector The software component a participant runs, or has run for it, to exchange data inside the data space. It enforces identity, contracts and policies. , which is not the same thing.
Specifications
- 2.2 We still have not run the site on the older phones and laptops we say we target. We do not own them, and our tooling cannot slow a processor down in a repeatable way. What we could measure instead is published above: weight, requests and DOM nodes per template, the three things an old device actually pays for. Extrapolating a load time from a development machine would give a presentable figure and a false one.
- 2.3 The site does not work offline. It stays usable on a slow connection because it is light, but we have not measured that either.
- 2.4 The oldest browsers we support are now written down and applied when the site is built, and the pages stay readable with no JavaScript at all. What the criterion asks for is running the site on those versions, which we cannot do here. The gap is narrower than it was; it is not closed.
- 2.7 Deployment, backups and data retention are documented. What happens when this service is eventually shut down is not.
- 2.8 We ask nothing of our suppliers about their own environmental footprint. That is an organisational decision, and it has not been taken.
Architecture
- 3.2 Server capacity is fixed rather than following actual load. Defensible at this traffic, but it is not what the criterion asks.
UX and UI
- 4.1 The animation on the home page starts on its own. We keep it: it is the identity of the page. It now stops as soon as it leaves the screen or the tab goes to the background, and it never starts at all if your system asks for reduced motion.
- 4.8 Three typefaces is a lot for one site, and we keep them as a design decision. They are cut down to the characters and weights actually used. Going further, to only the characters present today, would break the day an editor types a name with a letter we had removed.
- 4.11 The back-office does not tell editors the expected file size and format before they upload. Outside the public scope, but the criterion covers the whole service.
- 4.15 We give you no way to see what your own use of this site costs. We are not sure such a feature would help rather than decorate, and we have not built it.
Hosting
- 8.7 This site runs in availability zone fr-par-1, which we read from the machine itself. What we could not find is a published mapping from that zone to a physical building, so we cannot tell you how it is cooled. We have asked. Until there is an answer we can point at, we would rather say we do not know than pick the flattering figure.
- 8.8 We do not separate frequently read data from archives. The volume does not call for it yet, but the criterion is not met.
Telling us we are wrong
If a figure on this page looks wrong, or if something here does not match what the site actually does, please tell us through the contact form. A declaration that cannot be contested is not worth publishing.