[Planning] Delta-DM V2 Planning & Discovery

[jcace]

Today, Delta-DM is modelled around the concept of datasets that are imported once and then don't change very much. It's good for static, archival use case, but is not very flexible when it comes to continuously growing datasets and dynamic data sources/participants.

Conceptually, the next iteration of Delta DM will continuous data streams / inputs, and a feedback loop to manage the dealmaking for that content.

Simply put, In DDM V2 we will model things around INPUTS and OUTPUTS

Both Inputs and Outputs will be modular and allow different components to be attached as necessary for the client's use case / strategy.

The core functionality of DDM is a tracking database to facilitate replicating that data out to the Filecoin network in a variety of ways.

INPUTS

• Data that's flowing in to the system by various means, for example:
• EDGE-UR nodes (new Edge-URID architecture)
• Ptolemy / versioned static datasets
• Static import

OUTPUTS

• deals,deals,deals
• closed-circuit (known SPs)
• public market (any SP)
• Delta, Spade

[jcace]

[jcace]

Edge URID will be the first major INPUTS component https://github.com/application-research/delta-edge/issues/51

[jcace]

Closing as this issue is in the wrong repo - continued: https://github.com/application-research/delta-dm/issues/83