A whitepaper should be honest about the distance between design and reality. This section consolidates where XChain is today and where it is going. Ask of any claim: is this here now, in progress, or the vision?

✅ Here now — running, and verifiable by anyone:

The Testnet MVC is live on real infrastructure: 4 nodes (xchain-001 Singapore hosts services + a validator; xchain-002 Helsinki is a second validator + chain-data backup; plus xchain-003 / xchain-004). Two active validators (≈70.4% / 29.6%).
6 of the 7 primitives are real on-chain modules (UIR, UAC, XMP, DA, VLC, SC). MVC-1..4 pass end-to-end on real nodes + a real relayer (devnet/mvc-all.sh runs green).
The official explorer www.xscan.org reads chain-id and denom at runtime — real height, validators, transactions.
The codebase is 100% English, guarded by three acceptance gates, with green CI.

🟡 In progress / a stand-in — stated plainly:

PAI's prover is MOCK. The aggregation is real; the proof is a mock (ProofSystem.MOCK, blocked off-testnet by a fail-closed chain-id guard). A real Groth16/BN254 verify is wired in behind a stand-in circuit MiMC(prev,tx,new) (~164-byte proof) — but that is not yet the real SP1 verifying key (the true SP1 wrap circuit is ~15.97M constraints, OOM locally). Swapping in the real SP1 key is a named, future milestone.
EVM on A1 has a fuller proof-of-concept (cosmos/evm, coin-type 60, 17 passing e2e assertions) on a feature branch, not merged; going live on A1 needs a re-genesis. EVM release is not done.
The public mirror repo is kept private, opened only on explicit instruction.

🔭 The vision — future tense, not here yet. The path scales the hard walls one at a time:

Diagram

Figure 5 — The phased roadmap. "Aggregate toward 1B TPS" is aggregate nominal throughput (the Execution Layer ~100M + the Data Layer ~900M), not a billion composable transactions over one global state.

Today's testnet is deliberately the minimum that proves the mechanism, not the scale: 3 Execution-Layer chains + 9 Data-Layer namespaces on one DA node.

Node	Role	Primitives proven
A0 — Root/Settlement (T0)	UIR + Asset Registry + Settlement + receives aggregated proofs	#1, #2, #4, #7
A1 — Value (T1)	Assets, value transfer, ZK/validity	#2 (security-class), XMP WRITE
A2 — Interactive (T2)	Applications, anchoring Data-Layer commitments	#6 (VLC anchor), XMP READ (ICQ)
9 namespaces (the Data Layer)	Social / IoT / Chat × 3 on one DA	fee ~0 + inclusion-proof

Beyond the technical phases, the roadmap includes progressive decentralization: widening governance from the protocol (x/gov, already live) to the project's direction, strategy and this whitepaper — so the community, not a founding team, sets the course (§12).

Takeaway: mixing "here now," "in progress," and "the vision" is the fastest way to lose trust — so We keep each in its own tense. If you can verify it on www.xscan.org, it is here now; the rest is the road We are building, in the open.