DOG Mode Challenges Bitcoin Core's Relay Policy Governance
Leonidas announced DOG Mode on July 17: an alternative Bitcoin client that lifts relay transaction size to 3.9 million weight units and drops the dust limit to 1 satoshi, without touching consensus rules.
An alternative Bitcoin client announced July 17 would raise the relay size ceiling to nearly a full block and cut the dust limit to 1 satoshi, forcing a long-sidestepped governance question into the open.
Key takeaways
- Leonidas announced DOG Mode on July 17: an alternative Bitcoin client that raises the maximum relayable transaction size from 400,000 to 3.9 million weight units and reduces the dust limit from roughly 294-546 satoshis to 1 satoshi, without altering consensus rules.
- DOG Mode requires no supermajority miner signaling and no BIP approval. It needs only willing node operators and a single cooperative miner, making it the tactical mirror image of BIP-110, which has stalled near zero miner support.
- The real stakes extend beyond Ordinals: if relay policy can be changed by a small group of Core maintainers without a vote, it is de facto governance, and DOG Mode is the first serious counter-client designed to prove that point.
Leonidas, co-founder of the Runestone project, announced DOG Mode on July 17, 2026, describing an open-source alternative Bitcoin client that targets relay policy rather than consensus rules. The announcement landed with no published code repository and a call for developers and miners to build and signal support.
"Bitcoin Core and Bitcoin Knots have spent years enforcing rules that Bitcoin itself does not have. The $DOG Army is done asking for permission. It is time to remove even more of these frivolous restrictions." , Leonidas (@LeonidasNFT), X, July 17, 2026
What DOG Mode Actually Changes
Bitcoin's consensus rules define what is valid. Relay policy defines what nodes forward to miners before a block is confirmed. Those are two separate layers, and DOG Mode only touches the second.
The two specific changes proposed: raise the maximum standard transaction size from 400,000 weight units to 3.9 million weight units (a full Bitcoin block holds 4,000,000 WU, so current Core defaults cap relay at roughly one-tenth of a block), and reduce the dust limit from the current 294-546 satoshis (depending on output type) to 1 satoshi. Leonidas claims removing the dust limit would free approximately $25 million in satoshis currently locked as padding in Ordinals and Runes transaction structures. That figure is his own unaudited estimate.
DOG Mode is not a fork. It does not create a new chain. A node running DOG Mode accepts the same blocks as every other node. It just relays a broader set of unconfirmed transactions.
The Governance Question DOG Mode Forces Open
Bitcoin's consensus rules change rarely, publicly, and with extraordinary friction. Relay policy changes quietly, at the client level, without BIPs or community votes, by the handful of contributors who dominate Bitcoin Core. That asymmetry has gone largely unexamined because nobody built the counter-client.
DOG Mode forces the question into the open. If relay policy is just neutral operator discretion with no network effect, DOG Mode is a minor curiosity. But if a single alternative client plus one willing miner can confirm previously "blocked" transactions at scale, then Core maintainers hold governance power they have never formally been granted. DOG Mode is designed to test exactly that.
The comparison to BIP-110 sharpens the point. BIP-110 proposed tightening relay and consensus rules to restrict non-financial data inscriptions. It requires 55% miner signaling to activate and has never exceeded roughly 1% in any signaling period, standing at zero in the current one, per the BIP-110 signaling monitor.
DOG Mode needs neither a supermajority nor a BIP. One node, one miner. The asymmetry in activation requirements reflects exactly the governance asymmetry DOG Mode is exposing.
The DOG community has already demonstrated this mechanism works in practice. Large transactions that Core refused to relay were routed directly to miners through private channels, bypassing default relay entirely. Services like MARA's Slipstream exist precisely because non-standard transactions need a private lane to reach miners. DOG Mode, if it gains node adoption, would put that lane back in the hands of ordinary node operators rather than institutional transaction brokers.
If mempool fragmentation widens as different node populations relay different transaction sets, fee estimation becomes less reliable and confirmability of certain transaction types grows unpredictable. That affects everyone, not just Ordinals users.
What Needs to Happen Next
The falsifiable trigger here is Core's response. If DOG Mode attracts meaningful node and miner adoption and Core contributors respond with a counter-BIP, a policy adjustment, or even a public statement, it confirms that relay policy is governance and that the counter-client successfully applied pressure. If Core stays silent while DOG Mode grows, it proves the point a different way: the defaults are not as sticky as the maintainers assumed.
As of the July 17 announcement, no code repository had been published. Whether a repo has since materialized and whether any miners have signaled interest are the two near-term data points to watch.
Sources
- Leonidas (@LeonidasNFT), X, July 17, 2026
- First reported by CoinDesk
Frequently Asked Questions
Consensus rules determine what blocks and transactions are valid across the entire network. Every node enforces them.
Relay policy determines what unconfirmed transactions a node forwards to its peers before a miner includes them in a block. Relay policy is set at the client level and varies by software. DOG Mode changes only the relay layer, so nodes running it accept the same blocks as Bitcoin Core nodes and cannot cause a chain split.
No. DOG Mode changes what a node forwards across the peer-to-peer network, not what blocks it accepts. Consensus rules are untouched.
A DOG Mode node and a Bitcoin Core node will always agree on what constitutes a valid block. The difference is only in which unconfirmed transactions each node will propagate.
Miners can include any valid transaction regardless of relay defaults, but they can only include what they know about. Transactions that nodes refuse to relay typically reach miners only through private channels like dedicated relay services. That gives institutional players with direct miner relationships a consistent advantage over ordinary users whose transactions rely on p2p propagation. Broader relay defaults reduce that information asymmetry.
