How we validate a release

Every XEQM release is exercised on a private, multi-host testnet that mirrors mainnet consensus before a single byte reaches the live network.

01A real network, not a single node

We stand up independent hosts, each running service-node daemons, meshed on a private overlay so they form Pulse quorums and gossip exactly as mainnet nodes do — 20 operators across 40 service nodes. Governance mints and mining bootstrap the chain through the hard fork under test.

02Gate tests decide GO / NO-GO

A release is GO only when every mandatory gate test passes; any gate failure is NO-GO. The gates:

• Emission / supply audit — Total emission across the fork matches the curve to the atomic unit — no inflation.
• Fork activation — Clean transition at the HF height; reward serialization correct; no chain split.
• Pulse stall & recovery — Chain stays live via PoW/GOV_SIGNED fallback when a quorum can't form; Pulse resumes when operators return.
• In-place upgrade / DB migration — Existing nodes restart on the new binary, migrate the DB without corruption, and don't resync from zero.
• RPC compatibility — The RPC schema stays additive-only so exchanges, wallets and dashboards don't break.

03Coverage by blast radius

Tests are prioritised by what breaks if they are skipped — a chain split, an inflation bug, a bricked upgrade, or a broken integration.

04Every run is recorded

Each run emits a machine-readable result and a detailed report. This site is generated from those results — nothing here is hand-entered.