I wrote a very basic overview of watermarking, fingerprinting, timing analysis and supernodes for Bitcoin Magazine's last print issue, which is pretty much an unsolicited advertisement for why I think we need a second mempool (and also mixnets, but thats a longer story). Since no one cares about stuff like this on Twitter anyway, I'll explain here.

Bitcoin has a privacy issue on baselayer. I know this. You know this. Everybody knows this. The problem is that there's a lot of stuff we can't do to solve this issue without completely fucking up how Bitcoin works, like, say, anonymous amounts. But there is some stuff we *can* do to increase privacy on the Bitcoin baselayer. One of those things is incorporating a second mempool to integrate Dandelion++, the routing protocol used in Monero. Hear me out.

One of the ways blockchain surveillance firms identify who what transactions belong to on the Bitcoin blockchain is by operating so-called supernodes. A supernode sets up as many connections to other nodes as it can, and by doing so can establish where a transaction was first seen in the peer-to-peer network, ergo ascribe whom a transaction belongs to.

Here's where Dandelion++ comes in. Instead of propagating transactions to *all* connected peers, Dandelion++ propagates transactions like, well, a Dandelion.

In Dandelion++ propagation, Bitcoin nodes send transactions to *one* peer, instead of to all of them. This peer sends it to another peer, they send it to another peer, and so on and so forth. This is called the "stem phase".

When we've established enough plausible deniability, Dandelion++ reaches the "fluff phase". At this point, a node that did not *create* the transaction, but is simply relaying it, propagates it to all nodes in the network it is connected to, including supernodes, and the next node does the same, and so on and so forth – business as usual.

Incorporating Dandelion++ (or any other anonymizing propagation protocol, like Dandelion, Dandelion Lite, or Clover) would arguably seriously fuck up the blockchain surveillance stick as we are taking away the most obvious attack vector for blockchain surveillance firms. It's also not a trivial task, see ajtowns' overview of stempools (and no one wants to maintain another mempool on bitcoin, if we're honest). But it's a really interesting proposal to think about to increase privacy on Bitcoin that, yes, would be a lot of work to implement and maintain, but also does not get talked about enough imo for everyone yapping about Bitcoin baselayer privacy.