Threshold secret sharing for data (or metadata)

[Intensity]

Divide the data into n chunks, any k of which are able to recover the information (but with fewer than k shares, no information on the original data can be discerned). An example is Shamir's secret sharing.

This could be combined with other processors. The data payload could be split between a few cloud providers. By metadata, I mean for example an index or an ephemeral symmetric key. But the more common case could be to apply threshold secret sharing to the stream.

[Roman2K]

Hi @Intensity. Sorry for the late reply, my notifications were disabled 🙈.

Indeed, that processing you suggest would take the form of a two new procs, one for splitting into parts (like parity), one for joining them (like uparity).

I don't need such a proc, so I must admit I don't have the motivation to invest time in it. Would you?