Actually it would be fine! The creation of timestamp proofs is both very cheap, and scalable. The hashing of the data is done client side, so all that the servers have to do is build a merkle tree of all requests. That part scales well as we have aggregator software that can build different parts of the merkle tree on different servers.
At the moment demand is low enough that I only need a single aggregator per calendar. But if load increases we'll just spin up more.
Peter Todd
npub1ej…3ndrm
2025-04-05 08:43:45
in reply to nevent1q…v2yr
Author Public Key
npub1ej493cmun8y9h3082spg5uvt63jgtewneve526g7e2urca2afrxqm3ndrmPublished at
2025-04-05 08:43:45Event JSON
{
"id": "40d1b5ee6280888017d491eb13cada4e47854a8a8e96e368cbf670394f8a575a",
"pubkey": "ccaa58e37c99c85bc5e754028a718bd46485e5d3cb3345691ecab83c755d48cc",
"created_at": 1743842625,
"kind": 1,
"tags": [
[
"e",
"cb89297616dd7e79db5b9319dcd375916820516e5c351fe469c546538c52918e",
"",
"root"
],
[
"p",
"ccaa58e37c99c85bc5e754028a718bd46485e5d3cb3345691ecab83c755d48cc"
],
[
"p",
"9e0d3a2c5bdf7161b6f248e45af480a3d5ad923984d70ca42179b525e27caab8"
]
],
"content": "Actually it would be fine! The creation of timestamp proofs is both very cheap, and scalable. The hashing of the data is done client side, so all that the servers have to do is build a merkle tree of all requests. That part scales well as we have aggregator software that can build different parts of the merkle tree on different servers.\n\nAt the moment demand is low enough that I only need a single aggregator per calendar. But if load increases we'll just spin up more.",
"sig": "ce504d0749d5d30e8ddcd5b468f6ebefe39faf79d1ee562b12be4a0cfea055735bd8fffbe1009a257c302edc533d433580224fd2271001af40604296be47685b"
}