Bitcoin developer Leonidas unveiled DOG Mode, a client that skips the miner vote a soft fork normally requires. Foundry, the largest mining pool, has stayed publicly silent on BIP-110, and si
Bitcoin’s fight over data-heavy transactions took an unexpected turn on July 17 when Ordinals developer Leonidas published the specification for DOG Mode, a node client built to make the entire BIP-110 voting battle irrelevant. BIP-110 needed a supermajority of miners to activate a soft fork capping transaction outputs at 34 bytes and OP_RETURN data at 83 bytes, and that vote has stalled: signaling sits at roughly 1%, nowhere near the 55% threshold required before an early-August activation deadline. DOG Mode doesn’t ask a single miner to vote on anything. It rewrites the local relay rules that decide which transactions a node forwards, and if even one major mining operation runs it, that miner can pull oversized, high-fee transactions straight out of the mempool no matter what the rest of the network agreed to reject.
Foundry USA, which controls roughly a third of Bitcoin’s network hashrate, has issued no public statementon BIP-110 and has not signaled support. AntPool has stayed silent as well, while F2Pool has openly opposed the proposal. That near-total absence of large-pool backing leaves the ~1% signaling that does exist carried almost entirely by Ocean, the pool co-founded by Luke Dashjr, along with a handful of small independent operators. Michael Saylor has been blunt about whyhe thinks the proposal deserves to fail, arguing that turning a dispute over spam into a change to consensus rules risks invalidating transactions that are otherwise perfectly valid.
The distinction that makes this fight different from a normal Bitcoin governance dispute is the line between consensus rules and relay policy. Consensus rules define what the network as a whole will accept as a valid block. Relay policy is just local software configuration, the settings an individual node uses to decide what it bothers passing along to its peers before a block ever gets mined. Bitcoin Core’s 400,000 weight unit cap on a single transaction was never a protocol requirement in that sense; it was a default. The chain itself can hold blocks up to 4,000,000 weight units. DOG Mode simply removes the self-imposed filter, letting a transaction of up to 3,900,000 weight units through, and slashes the dust limit from the usual 294 to 546 satoshis down to exactly 1. Leonidas framed the move as overdue, saying Bitcoin Core and Bitcoin Knots have spent years enforcing restrictions the protocol itself never actually prohibited.
DimensionBIP-110DOG ModeLayer targetedConsensus rulesRelay policyPath to activation55% miner signaling, UASF flag dayOne miner running the clientMax transaction sizeRestricts output/data size400,000 WU raised to 3,900,000 WUDust limitEnforced, 294-546 satsCut to 1 satoshiCurrent status (July 17)~1% signalingSpec published, seeking miner adoptionA block that follows DOG Mode’s relay settings still has to satisfy Bitcoin’s actual consensus rules, so the rest of the network has no valid reason to reject it. That’s the entire point. Miners chase fee revenue first and ideology second, and a project willing to pay tens of thousands of dollars to push one enormous transaction into a block gives whichever miner accepts it a payday nobody else on the network can veto.
Adam Back has framed this asthe honest version of how Bitcoin was always supposed to work, telling BIP-110’s backers that if they want a restriction imposed on the network, their only legitimate option is to organize a fork rather than lean on a soft-fork vote to settle what counts as acceptable use.
Dropping the dust limit to 1 satoshi does something concrete beyond the ideological argument: it unlocks roughly $25 million in bitcoin that’s currently sitting idle as mandatory padding on Ordinals and Runes transactions. That capital becomes usable the moment a miner running DOG Mode starts accepting transactions built around the new minimum. The tradeoff shows up on the other side of the ledger. Jameson Lopp has warned thatonce a network accepts that any actor can push through changes to what counts as valid activity, the argument for the next restriction gets easier to make, not harder, and Bitcoin’s image as permissionless money erodes either way. Blocks that regularly fill to the 3,900,000 weight unit range will also grow the chain’s total size at a faster clip, and if that growth outpaces what a consumer-grade machine can comfortably store and verify, running a home node stops being realistic for a meaningful share of today’s operators.
The open question isn’t whether DOG Mode is technically valid, since nothing about it breaks a single consensus rule. It’s whether any miner beyond the initial backers actually switches over before the BIP-110 flag day arrives in early August, and whether exchanges and wallet providers treat transactions built under DOG Mode’s relaxed limits as standard or flag them as non-conforming until the rest of the ecosystem catches up. That compatibility gap, separate from anything miners decide, will likely determine how fast the unlocked liquidity actually moves.
The post DOG Mode Bitcoin Doesn’t Wait for a Miner to Say Yes appeared first on ETHNews.
