Is a Name Discoverable?

When Wellzai first consumed the RSM kernel as an outside application, through its public protocol and nothing else, one small thing could not be done properly: showing the name of a participant it had discovered.

Discovery returned a subject's canonical identity but no label, because person and organization names are not discovery data. Whether they should be is, in the words of the record kept from that adoption test, a privacy policy question RSM has not yet governed. Wellzai labelled its own demonstration identities from its development setup and otherwise fell back to a short identifier. The gap was deliberately not closed in the runtime. Closing it properly needs a governed decision about which profile fields are discoverable.

The surrounding design suggests what such a decision could look like. The kernel already separates broad, discoverable geography from protected precise location: a person can be discoverable within a broad area while their exact location stays undisclosed until disclosure policy permits. Discovery returns discovery-safe facts; the details are the subject's own records, released only through an authorized disclosure. The federation architecture's broader rule is minimum necessary information. A discovery index may need to know that a participant has a capability without holding anything more about them.

The implication is that a name is not neutral metadata. In a self-sovereign network, being findable by name is itself a disclosure, and it should be the participant's to make.

The open question is the default. A network where no one can be found by name is safe and strange; one where everyone can is convenient and exposing. Which profile fields should a self-sovereign network expose by default, and should a person's name follow the same rule as an organization's?