Ideas to improve pronto

pronto is a new galacteek protocol that uses linked data to build a distributed RDF graph using an ontological chain system.

When a new object is recorded, it is signed and chained with URIs in what's called an ontolochain. Chains are synced via SparQL over IPFS streams. Chain sync status is also stored in the history RDF store.

The most obvious issues:

• If an ontolochain record is lost, the synchronization of this chain is compromised (we only keep pointers to the previous object/block)

• Scaling. Each node pulls everything it can from other nodes. While this makes content distribution almost instantaneous, this will not scale well, as rdflib (using bsddb or other backends) is really not the fastest RDF library there is.

• Encryption. Encryption of all objects is not desirable but it will be supported. Using raw unencrypted DAGs makes things easier to study in the beginning as you can explore the raw IPLD from galacteek easily.

Will post an example of pronto's history graph so that things make more sense ..