- Moved all data processing into the data service
- Allows for all operations to be properly throttled to avoid errors from hot partition
- Reduce code duplication
- Puts the logic in the right domain
- Refactored how we write data for tokens and owners
- No more waiting for tokens to update data in crawler and loader to handle a few issues
- Contracts that are missed by the loader because they are not recognized (contracts making contracts)
- Contracts that never mint or burn but place transfers (OpenSea)
- Order dependence causing lots of extra queries
- Cleaned up commands/bin to be better contained and easier to follow
- Reduced code duplication for stats logging
- Standardized on naming for metadata_uri
- Updated parallel bus to allow for configuration option to raise exceptions caught by consumers
- Updated nft tail command to have bus raise exceptions, log errors from those exceptions,
and reprocess the block on which hte error occurred
- Fix bug in tailer where it would get ahead of the block height