Big questions for projects using ComposeDB

OrbisDB <> ComposeDB

• Where do they host their node?
  • Hirenodes?
    • If no: Details about their own server hosting
• Do they use ComposeDB libraries for write (probably yes)?
  • If yes: Mutation examples?
  • If no: How do they write data?
• Do they use ComposeDB GraphQL endpoint for reads (probably yes)?
  • Which GraphQL provider they are using?
  • If yes: Example of queries and relations?
  • If no: How do they query content?
• Do they have an existing Postgres database
• Details about their project
  • Are they building an SDK?
  • Are they building a front end app?
  • Are they building a back end app?

Simple

Action item:

• [ ] List eligible projects working with Hirenodes.

Flow:

• Write
  • Projects keep on writing data using ComposeDB library
  • Projects must update their node to support Data Feed API
    • Very easy if work with Hirenodes
• Indexing
  • By default will pick up new streams created on their node
  • Projects must have their own OrbisDB instance
    • Very easy if work with Hirenodes
  • OrbisDB Must support historical syncing
    • Using ComposeDB postgre
    • Using Recon / ComposeDB models indexed
  • OrbisDB support Relations written via ComposeDB
• Querying
  • Projects change the ApolloProvider endpoint to use OrbisDB one
  • Projects will probably have to change a few field names and slightly update their relations
  • OrbisDB must support historical relations from ComposeDB models

Advanced

Action item:

• [ ] Find a way to connect to ComposeDB to do historical syncing

Flow:

• Write
  • Projects migrate to DB SDK to write data
• Indexing
  • Same as Simple
  • For larger projects Gitcoin Passport and Oamo
    • OrbisDB create a ComposeDB migration plugin
      • Connect to CDB database using GraphQL exposed from ComposeDB libraries to sync everything to OrbisDB