One byproduct of the fitful filter fight (or so I'm calling this debate about relaying transactions people don't like) is that the peer-to-peer part of the Bitcoin network is getting a workout.
Libre Relay
In January, 2024, @petertodd released a fork of Bitcoin Core called Libre Relay. Libre Relay had these additional features:
- Removed op-return limits
- Connected to an additional four Libre Relay nodes (via a service bit)
- Implemented "pure replace by fee"
So, here we have a node that intentionally maintains connections to nodes that are like itself (along some metric, doesn't really matter) in order to increase the propagation of certain transactions.
Garbageman
In early May of 2025, Chris Guida (@cguida6 on X, I couldn't find his username on SN) released a fork of Bitcoin Knots called Garbageman
- A knots node that pretends to be a Libre Relay node
- Has much tighter relay policy and does not relay transactions libre relay nodes want to relay
This is a node that looks for some identifying attribute found in a set of nodes and intentionally connects with these nodes in order to decrease the propagation of certain transactions.
We're under attack!
In late May, Todd started a thread on the Bitcoin Mailing List stating
Recently proponents of transaction "filtering" have started sybil attacking Libre Relay nodes by running nodes with their "garbageman" fork
He proceeded to discuss several ways nodes could become more sybil resistant:
- Most fees metric
one straight-forward approach would be to simply keep track of a decaying average of new fees/sec each peer had advertised to you prior to you advertising the transaction to them. Periodically, you could drop the peer with the lowest new fees/sec ranking, and then connect to a new peer.
However, Todd notes that this might not be the easiest thing to implement for a variety fo reasons.
- Most-fees next (one or two) blocks mempool
Cluster mempool is a functionality that will allow nodes to store transactions in clusters ordered by fees. With cluster mempool
Nodes can cheaply compute the total fees in the top one or two blocks worth of transactions they currently have in their mempool, and advertise this fact to their peers.
Periodically you would drop the peer advertising the lowest double-block of fees, and then connect to a new peer to see if they're better.
Even a small difference in transaction relay policy will show up as missing transactions. This difference will translate into the sybil attacking node(s) getting dropped, and honest nodes with policy compatible with yours eventually being found.
Guida replied to these suggestions in a more philosophical vein, but concluded
I’m sure there are strategies for getting LR nodes to detect GM nodes and banning them. And I’m equally sure that, if implemented:
- Very few people will run them. Only LR nodes are likely to run the garbage-maximizing strategies Peter outlined above. I don’t know of any noderunners in their right minds who would run them.
- The pro-spam-filtration noderunner community will work around these detection methods any way we can, and we will never give up.
After this, the conversation turned toward filtering and not filtering and that's not what I'm interested in here. What is interesting is that the participants in the discussion dropped lots of suggestions for attacking/defending against each other (I only included a few examples).
Ban everybody!
Then on June 22, 2025, @aeonBTC (finger's crossed this is their SN handle) released a "Bitcoin Knots Banlist" with instructions for how to implement the ban list on a Bitcoin Core node.
This is a node banning connections to other nodes based on the version or type of software a specific node is running.
The results in the social sphere were predictable. On the surface, these are horrible ideas! Bitcoiners shouldn't be thinking about how to wreck the network!
Whatever can be done by Bitcoiners to attack each other can be done by the state with much greater intensity.
gmaxwell made an interesting note in the mailing list thread:
Not building the censorship infrastructure (even though you intend it for 'good' purposes) and instead building anti-censorship infrastructure leaves us all with a better world.
I used to be pretty aligned with this mode of thinking. But these recent attempts by the various camps of the filter fight are making me think perhaps there is some good here.
Let's take off the kid gloves
Bitcoin is an open, permissionless network. The value of the coins comes directly from the network's ability to resist the state (censorship resistance). If we expect to actually deliver on this promise, Bitcoin needs to relentlessly pursue battle-testing. If it can be done, a motivated attacker will try it.
While I'm sure all Bitcoiners are highly motivated to probe for weaknesses in the Bitcoin p2p network, nothing motivates people like righteous anger. Get after it filterers and core-ers.1
Footnotes
-
Is there a better name for the side of the op-return debate that is pro relaxing the op-return limit? ↩