I was reading Dusk's Network Updates page at 1am, not their homepage.
One line stopped me: new Phoenix transactions are no longer accepted.
Phoenix is Dusk's shielded model — note-based, ZK, the thing that made "privacy L1" mean something at the base layer. On the Boreas hard fork it was switched off for new transactions: mainnet at the coordinated restart block 4,414,095 on 10 June 2026, testnet at block 4,000,000 on 7 August 2026.
Moonlight is what's left. Fully transparent, account-based. Public addresses, public balances.
Nodes still keep Phoenix decoding, so old blocks stay replayable. Historical privacy is intact. But no new shielded transfer is being accepted.
The intended replacement for confidential flows is Hedger on DuskEVM.
Then I checked the homepage. DuskEVM: Testnet. Hedger: Testnet. The native L1 card, marked Live, still lists confidential shielded transfers as a feature.
So there is a window. A chain marketed as confidential by default is, at the base layer, transparent by default until the EVM privacy layer ships.
Now the fair reading. This looks like consolidation, not abandonment — one canonical transaction path while the privacy layer matures. Two ZK models in parallel is real attack surface and real audit cost. Deprecate first, ship second, is defensible.
What I do not know: whether a Phoenix successor returns to the L1, or whether Hedger becomes the only confidentiality path Dusk has. I have not seen either committed to anywhere.
So, the real question.
Should a privacy L1 keep a shielded path live at the base layer while its replacement matures, even at the cost of maintaining two models? Or is one canonical model the safer engineering call, and the marketing simply has to catch up?
#dusk $DUSK @Dusk
One line stopped me: new Phoenix transactions are no longer accepted.
Phoenix is Dusk's shielded model — note-based, ZK, the thing that made "privacy L1" mean something at the base layer. On the Boreas hard fork it was switched off for new transactions: mainnet at the coordinated restart block 4,414,095 on 10 June 2026, testnet at block 4,000,000 on 7 August 2026.
Moonlight is what's left. Fully transparent, account-based. Public addresses, public balances.
Nodes still keep Phoenix decoding, so old blocks stay replayable. Historical privacy is intact. But no new shielded transfer is being accepted.
The intended replacement for confidential flows is Hedger on DuskEVM.
Then I checked the homepage. DuskEVM: Testnet. Hedger: Testnet. The native L1 card, marked Live, still lists confidential shielded transfers as a feature.
So there is a window. A chain marketed as confidential by default is, at the base layer, transparent by default until the EVM privacy layer ships.
Now the fair reading. This looks like consolidation, not abandonment — one canonical transaction path while the privacy layer matures. Two ZK models in parallel is real attack surface and real audit cost. Deprecate first, ship second, is defensible.
What I do not know: whether a Phoenix successor returns to the L1, or whether Hedger becomes the only confidentiality path Dusk has. I have not seen either committed to anywhere.
So, the real question.
Should a privacy L1 keep a shielded path live at the base layer while its replacement matures, even at the cost of maintaining two models? Or is one canonical model the safer engineering call, and the marketing simply has to catch up?
#dusk $DUSK @Dusk