A prominent voice in Bitcoin’s Ordinals and Runes community wants to rewrite the rules for what data the network carries — and unlike previous attempts, his plan does not need a vote. The proposal, called DOG Mode, could free an estimated 25 million in capital currently locked up as padding in Bitcoin transactions. But it also reignites a fierce debate about what Bitcoin is for.
By Michael Nguyen | July 18, 2026
The Hook: Two Visions for Bitcoin’s Future
Bitcoin has always been money first. But over the past two years, a growing community of developers and users have been embedding other things into the blockchain — images, text, tokens — using technologies called Ordinals and Runes. Think of it like writing messages on the back of every dollar bill that passes through your hands. The bills still work as money, but now they also carry information.
Not everyone likes this. A faction of Bitcoin developers and users wants to restrict non-financial data, arguing it bloats the blockchain and drives up costs for everyone. Their proposal, called BIP-110, would change Bitcoin’s fundamental rules to cap arbitrary data. It needs 55 percent of miners to signal support before it can take effect.
According to tracking data from the BIP-110 monitor, the proposal has drawn effectively zero miner support in the current signaling period and has never cleared roughly 1 percent in any period. Its node support sits in the low single digits, carried almost entirely by Bitcoin Knots, an alternative to the dominant Bitcoin Core software.
Now the other side has a counter-proposal that skips the vote entirely.
On-Chain Evidence: How DOG Mode Works
Leonidas, co-founder of the Runestone project and one of the most prominent advocates for Ordinals and Runes, has announced plans for an open-source Bitcoin client called DOG Mode. Unlike BIP-110, which changes the consensus rules, DOG Mode only changes what individual nodes choose to forward to their peers — a softer category called relay policy.
Here is the difference in plain terms: Bitcoin’s consensus rules are like the laws of physics for the network — break them and your transactions are rejected. Relay policy is more like the postal service deciding what packages it will carry. A package the post office will not carry can still be delivered by hand, and the same is true for Bitcoin transactions — a miner who receives a transaction directly can still include it in a block.
DOG Mode would make two specific changes to what nodes relay:
- Raise the maximum standard transaction size from 400,000 weight units to 3.9 million weight units. A full Bitcoin block holds four million weight units, so currently nodes will not relay anything larger than a tenth of a block. DOG Mode would allow transactions that fill nearly an entire block.
- Cut the dust limit — the floor below which an output is considered too small to bother relaying — from between 294 and 546 satoshis down to a single satoshi. A satoshi is the smallest unit of bitcoin, equal to one hundred millionth of a single BTC.
The Core Conflict: Who Decides What Bitcoin Is For?
The stakes here go beyond technical parameters. At issue is a philosophical question that has divided the Bitcoin community for years: should the network be optimized purely for financial transactions, or should it serve as a general-purpose platform for data and tokens?
The BIP-110 camp argues that Bitcoin’s block space is a scarce and valuable resource. Every byte used for an image or a token inscription is a byte that cannot be used for financial transactions. They worry that bloating the blockchain makes it harder and more expensive for regular users to run nodes, which could centralize the network over time.
The Ordinals and Runes camp counters that the market should decide. If someone is willing to pay the fee to include data in a block, that is their right. The fees generated by Ordinals and Runes transactions have at times provided significant revenue for miners, strengthening network security.
Leonidas claims that cutting the dust limit would release approximately 25 million in capital currently locked up as padding — extra bitcoin that Ordinals and Runes creators must add to their outputs to meet the current threshold. That is real money sitting idle, and freeing it would return it to active use in those ecosystems.
But DOG Mode has a notable weakness: it does not exist yet. Leonidas asked developers to contribute to an initial release, miners to add support, and users to spread the word. There is no repository, no version, and no benchmark as of the announcement. He is proposing a client and asking someone else to build it.
Market Implications: What This Means for Miners and Investors
For miners, this debate is directly about revenue. Larger transactions and more data on the blockchain mean more fee income. When Ordinals activity spiked in past periods, transaction fees surged, providing miners with a significant chunk of their revenue beyond the block subsidy. DOG Mode, if adopted, would make it easier and cheaper to create these data-heavy transactions — potentially boosting fee revenue for any miner willing to process them.
The beauty of DOG Mode from a miner’s perspective is that it requires just one miner to include these transactions. No threshold, no signaling window, no deadline. If even a single mining pool starts accepting near-block-size transactions and one-satoshi outputs, those transactions will confirm. Other miners can join if they want the fee revenue.
Services like MARA’s Slipstream already exist to handle large, complex transactions outside the standard relay path, showing there is demand for this kind of capability. DOG Mode would make it standard rather than exceptional.
For regular investors, the implications are more indirect but still important. Higher fee revenue for miners means more profitable mining operations, which can support network security and potentially reduce selling pressure from miners who need to liquidate holdings to cover costs. On the other hand, if the data debate leads to a fragmented ecosystem where different nodes follow different rules, it could create confusion and uncertainty.
The Verdict: A Technical Squabble With Real Consequences
The DOG Mode proposal highlights a fundamental tension in Bitcoin’s evolution. The network was designed to be neutral — its rules do not favor one type of transaction over another. But in practice, the software that most nodes run (Bitcoin Core) makes editorial decisions about what to relay, and those decisions shape what the network is actually used for.
BIP-110 tried to settle the debate through consensus — a vote that would bind the entire network. It failed because it could not attract miner support. DOG Mode takes the opposite approach: let individual nodes and miners decide for themselves. It is a more Bitcoin-native solution, in the sense that it preserves freedom of choice at the individual level.
But the fact that DOG Mode exists only as an announcement, without code or a working implementation, means the debate is far from settled. The Ordinals and Runes community has shown it can rally attention and generate excitement. Whether it can actually ship software and convince miners to run it is the harder test.
For now, Bitcoin’s block space remains governed by the old rules. But the pressure to change them is not going away — and with an estimated 25 million in locked capital at stake, the incentive to find a way forward is only growing.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. Cryptocurrency investments carry risk; always do your own research.
DOG Mode skipping the vote entirely is wild. Leonidas just said fine I will build my own client then
oh great another bitcoin civil war. because the 2017 blocksize debate was so productive for everyone involved
2017 at least ended with a resolution. this one just loops forever because neither side can force the other off the relay layer lol
BIP-110 never cracked 1 percent miner support. at some point you have to accept the network does not want this
BIP-110 has literally never cracked 1 percent miner support. the anti-Ordinals crowd is loud on twitter but has zero hashpower backing them
DOG Mode bypassing the vote entirely is the move. if you cant win consensus just change the relay policy on your own node. elegant honestly
Dimitri V. BIP-110 at under 1% miner support means the anti Ordinals crowd lost the vote and now they want to win through client policy. same playbook different year
25 million locked up as padding in transactions. freeing that is a massive unlock for Runes whether the purists like it or not
25 million locked up as transaction padding is insane. had no idea Runes were carrying that much dead weight
blockspace_bro the 2017 blocksize war at least had economic substance behind it. this is about JPEGs and memecoins on the base layer. what a time to be alive
Leonidas basically said forget consensus I will just change relay policy on my own node. that is either brilliant or terrifying depending on which side you are on
rune_bag_47 changing relay policy on your own node is not consensus. its one validator opting out. if the market disagrees the mempool routes around you
Min-soo C. changing relay policy on your own node is opt out not consensus. if the market disagrees the mempool literally routes around you. this isnt a fork its a preference
relay_rat_ changing your own relay policy is fine but when enough nodes do it together thats soft censorship. the mempool doesnt route around coordinated behavior it just makes a new default
Bence M. coordinated relay policy changes are functionally identical to soft censorship. saying its just preferences ignores network effects on mempool propagation
soft censorship is doing a lot of work in that sentence. nodes picking what they relay is the design working, nobody owes your inscription propagation
the filter side never answers who curates the spam list. once relay policy becomes editorial policy you have a committee with no accountability
and whoever writes the spam list becomes an editor with no byline. id rather eat 25M in padding than hand content decisions to an anonymous maintainers circle
this is the part nobody answers. a spam filter is an editorial decision wearing an engineering hat. who audits the list six months in?
25M in transaction padding being dead capital is the real story. whether DOG Mode is the right solution is debatable but the problem is undeniable
25M locked as transaction padding is real dead capital. whether DOG Mode is the right fix is debatable but ignoring the problem is worse
the 25M number moves with fee regimes, could be half that in a low fee quarter. even 12M of dead padding justifies a proposal someone will argue about for a year
even at 12M its someones capital parked as garbage bytes. the utxo bloat crowd defending mandatory padding as a feature is a fun bit of cognitive dissonance
BIP-110 under 1 percent miner support tells you everything. the anti Ordinals crowd is loud on twitter but has zero hashpower. Leonidas saw the numbers and acted
ordinal_bag_ under 1 percent miner support for BIP-110 and they still pushed forward. the anti Ordinals crowd has conviction but zero hashpower. Leonidas read the room correctly
skipping the vote is the entire design. OP_RETURN limits are relay defaults, any miner can run DOG Mode tomorrow without a bip or a signalling period
25 million in padding is real money doing fake work. whatever you think of Ordinals, paying fees to store nothing is a protocol level inefficiency tax