Do we need thegraph.com on IC? Indexing, querying and caching of advanced queries needed

  • The standardised log format: Exists via ICRC-3 and each new standard should be defining it’s block types
  • The pubsub mechanism - We are defining this and building implementations in the Events Utility WG. We have a meeting today! Technical Working Group: Inter-canister Event Utility Working Group - #18 by skilesare
  • The db to index the data - Some work has been done. cc: @ferMartz and @icme . I think the remaining work is to figure out how to make the index creation dynamic according to the ICRC3 block structure.
  • The graphql (or other) engine to process incoming queries - CanDB has a bit of a query language, but it would be great if there was a kind of graphql query → canDB client execution component.

An external web2 indexing would certainly be easier as you could just keep querying the icr3_get_blocks endpoint, but then if you need to get that stuff from inside the IC you’d have to use HTTP outcalls.

1 Like