Here we publish anonymized experience reports from incident response, data hygiene, and Git-native compliance — precisely the topics in which small mistakes have a big impact.

First article

When the backup is missing and a script deletes: A missing backup and an unquoted deduplication script delete ~6,400 unique files. How we recovered ~98% from a Nextcloud mirror, external backup, and Git origin.

Read the experience report →

Articles

Date Article Preview
2026-08-25 When the backup is missing and a script deletes — Data loss caused by a faulty deduplication script and how ~98% were recovered via Nextcloud mirror, backup, and Git origin. Preview: Resilience through Git
2026-09-20 One receipt, one hash, one authoritative location — How Git-native compliance prevents redundancy and errors in networked systems: original docs exist once, other locations hold only structured references (path + SHA-256 + object ID). Preview: One hash, one source
2026-10-03 From concept to prototype — How a conversation sharpened an architecture: sphere check as a resolving call-back, reproducible compliance artifacts in Git. Preview: From concept to prototype
2026-10-05 Same session, different device — Session handover between browsers: an agent session lives server-side, a derived link (plus QR code) suffices for the handover — “one instance, many clients”. Preview: Same session, different device
2026-10-11 From scanned receipt to digital self-generated receipt — How a scanned hospitality receipt becomes audit-proof via OCR, document type recognition, and a signing process — and why a separate sidecar is no longer needed. Preview: From receipt to self-generated receipt
2026-10-13 Presentation is not a detail — Theming, languages (i18n/.resx), and the framework question: why presentation is a state, languages are decided per request — and why the official documentation beats any forum pattern. Preview: Presentation is not a detail
2026-10-25 The false finding that wasn't — PHP2Shell: how a patch verification, four log layers and two wrong search assumptions produced a false finding — and which counter-checks exposed it. Preview: The false finding that wasn't

Further articles to follow.