🏠 Developed on: GitLab · 🪞 Mirror: GitHub · 🌐 Website: pantheon-souverainete.eu

🏛️ The Pantheon of Sovereignty

A 100% sovereign, redundant and secure cloud infrastructure, designed to break free entirely from the GAFAM. Version 2.1.0-alpha — Full structural overhaul. Starting from scratch.

🙏 A word about the silence (2 months)

I owe you an apology. The project has not progressed publicly over the past 2 months.

The truth: it was not an abandonment. I worked under the radar on a complete structural overhaul. V1.x had shaky foundations, so I made the decision to start over from a blank page rather than patch things up.

Concrete proof that the project is alive: the showcase website is online (pantheon-souverainete.eu).

Lesson learned: I no longer promise specific dates. The guides will arrive "when they're ready".

🧠 Philosophy & Methodology

Versioning & History

Important note: The earlier versions (1.0.0, 1.1.0) were alphas before the suffix was formally adopted. No stable version has ever been released — everything is in alpha.

Version	Status	Description
v1.x	❌ Outdated	Initial architecture, guides published as they went
v2.1.0-alpha	🚧 In progress	Full structural overhaul. Starting from scratch.

🤔 Philosophy on Forks

Recurring question: "Why do you use [X] rather than its fork [Y], which is more popular in the self-hosted community?"

My answer: I deliberately limit the use of forks. A fork creates a structural lag in updates relative to the upstream project, which can introduce security vulnerabilities or incompatibilities.

I only favour a fork if it brings a SIGNIFICANT gain in:

✅ Security (audited code, vulnerabilities fixed faster)
✅ Privacy (telemetry removed, privacy-first features)

In all other cases: I stick with the upstream project.

Example 1: Debian > Ubuntu

Criterion	Debian	Ubuntu
Origin	Source	Fork of Debian
Stability	✅ Rock-solid (rigorous testing)	⚠️ Sometimes regressions
Security patches	✅ Direct (Debian Security team)	⚠️ Delay (Canonical has to merge)
Philosophy	Community-driven, sober	Commercial, "user-friendly"

Why Debian?

"We can't afford to wait for a patch on the system itself. It's a matter of security and stability."

Example 2: AmneziaWG > WireGuard (kernel)

Critère	WireGuard	AmneziaWG
Origin	Source protocol	WireGuard kernel fork
Obfuscation	❌ No	✅ Native (anti-DPI)
Confidentiality	Excellent	✅ Even better
Collaboration	—	✅ Works with the WireGuard team

Why AmneziaWG?

"A deliberate fork because of the significant gain in confidentiality (anti-DPI obfuscation). Moreover, the Amnezia team works closely with WireGuard — it's not a 'wild' fork."

Example 3: Bitwarden > Vaultwarden

Critère	Bitwarden (official)	Vaultwarden (fork)
Origin	Source project	Rust fork (lightweight)
Clients	✅ Official	⚠️ Uses Bitwarden clients
Compatibility	✅ Guaranteed	⚠️ Incompatibility windows
Synchronized updates	✅ Yes	❌ No (possible lag)

Why Bitwarden and not Vaultwarden?

"Vaultwarden itself states that there may be incompatibility windows where the server is inaccessible to official Bitwarden clients. Beyond security, there is a real compatibility issue. I prefer the official version."

In summary:

✅ Debian (source) > Ubuntu (fork) — System stability & security
✅ AmneziaWG (fork) > WireGuard — Confidentiality gain (obfuscation)
✅ Bitwarden (source) > Vaultwarden (fork) — Compatibility & security

Hosting & Sovereignty

Primary: Infomaniak (Switzerland) — Data in Switzerland, signed DPA, genuine environmental commitment
Backups: Hetzner (Germany & Finland) — Cold storage replicated across 2 DCs
Golden rule: No data rests on a single site.

Can I use V1 while waiting for V2.1?

Technically: Yes, the HS-02 (Firewall) and HS-03 (Gold Image) guides are technically valid.

In practice: ❌ No, it is not recommended.

Why?

Only 2 of the 3 Special Editions were published in V1 (HS-01 missing)
Security has been entirely rethought for V2.1
New security tools will be integrated (e.g., Lynis for hardening audits)
The V1 guides are therefore outdated in practice

Recommendation: Wait for the publication of the V2.1 guides for an optimal experience.

Author Profile & Transparency

I am not a certified IT professional. This project is the result of a passion for system administration, built through self-training over ~5 years.

Risk of error: I rely on official documentation, but I am not immune to mistakes.
AI assistance: The documentation and architecture benefit from AI assistance, but every command is human-verified.
Ethical stance: Immediate abandonment of any tool whose privacy policy is questionable.

📒 This project is a shared technical logbook.

This is my personal infrastructure, built in my free time and for my own needs. I document it and share it because I hope it can help others in their journey toward digital sovereignty.

The guides progress at my own pace — when I have time available. If the content helps you or inspires you, I am sincerely glad. If the publication pace or certain technical choices don't suit you, I completely understand: everyone has their own constraints, priorities and vision.

The goal remains simple: to pass on what I learn, in my own way, without pretension.

Limitation of Liability

Educational and personal project. The guides are provided "as is", without express or implied warranty.

No Guarantee of Accuracy: The author does not guarantee the absence of errors or future incompatibilities.
Security & Vulnerabilities: No architecture is "vulnerability-free." The guides reflect the best of my knowledge as of their writing date.
Deployment Responsibility: It is imperative to test any procedure in an isolated environment before production. The administrator assumes full responsibility for the security of their data.

🗺️ Architecture: The 12 Deities

⚠️ Important V2.1: All deities are undergoing revision/reintegration into the new architecture.

Le Panthéon de la Souveraineté — Updated!