Very minor Shamir split count leakage

[grocid]

In Vault, a set of n splits of a byte is generated as X = {1, 2, 3, ..., n} and Y = {p(1), p(2), p(3), ... , p(n)}, where the x coordinate is appended to the end of each split. Observing a split 0xAB05 gives a lower bound on how many splits being generated (n ≥ 5), which might leak some information about the infrastructure in which Vault is deployed, depending on the scenario. A very simple fix is to generate a random set of x coordinates instead of sequential coordinates.

Although this is by no means a significant leakage, it can be mitigated very easily, and leaking information is never a good thing :-)

[v6]

// , Can you give an example of the kind of info. this might leak, or of one of the scenarios?