On Fri, Feb 26, 2021 at 11:40:35AM -0700, Keagan McClelland via bitcoin-dev wrote:
> Hi all,

Hi Keagan,

> 4. Once the node's IBD is complete it would advertise this as a peer
> service, advertising its seed and threshold, so that nodes could
> deterministically deduce which of its peers had which blocks.

Although some of the details differed, I believe this general idea of
sharded block storage was previously discussed in the context of BIP159,
which warns:

    "Peers may have different prune depths (depending on the peers
    configuration, disk space, etc.) which can result in a
    fingerprinting weakness (finding the prune depth through getdata
    requests). NODE_NETWORK_LIMITED supporting peers SHOULD avoid
    leaking the prune depth and therefore not serve blocks deeper than
    the signaled NODE_NETWORK_LIMITED threshold (288 blocks)."

- BIP: https://github.com/bitcoin/bips/blob/master/bip-0159.mediawiki#counter-measures-for-peer-fingerprinting
- Discussion thread 1: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-April/014186.html
- Discussion thread 2: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-May/014314.html
- Discussion thread 2, continued: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-April/014186.html
- BIP159-related PR, review comments: https://github.com/bitcoin/bitcoin/pull/10387

> If you have thoughts on
> 
> A. The protocol design itself
> B. The barriers to put this kind of functionality into Core
> 
> I would love to hear from you,

I think it would be unlikely for any popular node software to adopt a
technique that could make specific nodes easily fingerprintable on an
ongoing basis unless it solved some other urgent problem.  Luke Dashjr's
rough data collection currently shows 5,629 archival listening nodes,[1]
which is a substantial fraction of the roughly 10,000 listening nodes
reported by Addy Yeow,[2] so I don't think we're near the point of
needing to worry about the unavailability of historic blocks.

    [1] https://luke.dashjr.org/programs/bitcoin/files/charts/services.html
    [2] https://bitnodes.io/dashboard/

However, if there's a reasonable solution to the fingerprinting problem,
I do think node developers would find that very interesting.

-Dave