Closed thobson88 closed 3 days ago
First attempt:
Suppose (as in the Rebasing section above) two independent root transactions have been created by two separate user communities, each with its own network of trust relationships represented by chains of downstream DIDs.
At some later date there may be a mutual agreement to fuse the two networks into a single one, but (unlike the case above) suppose that neither of the root entities is prepared to be subordinated by the other.
In this situation the two networks can be integrated by publishing a new root DID, whose set of public keys is the union of the public keys from each of the original roots. The two original root DIDs are then updated to become downstream DIDs of the new root.
Users on both networks would need to reconfigure the root event time on their Trustchain nodes to match the publication time of the new integrated root DID but, once this is done, all credentials that were previously issued will remain valid and will now be verifiable by participants on both networks.
We already have two technical notes about ways to integrate different trust networks:
The main limitations of these approaches are:
A third possible approach is to create a new root that is upstream of both of the original roots and contains the same keys. This is a more blunt method because users on both networks have to update their root event time. But it avoids the limitations mentioned above and is therefore probably worth writing down.