Splice 0.6.3

• The app_activity_record_store table has been modified to avoid unexpected DB performance issues. Upon upgrade to this version, the existing data in this table which has been ingested since the 0.5.18 release will be deleted in order to support that. This impacts the data being served via the experimental field app_activity_records on the /v0/events and /v0/events/{update_id} endpoints. Specifically the app_activity_records field will not contain the data which has been provided for the events which happened between the 0.5.18 and this release. Note that the app_activity_records data already provided for events during this period is correct and the network explorers who have ingested this data should keep a copy of it.
• Add /v1/holdings/summary endpoint that drops the accumulated_holding_fees_unlocked, accumulated_holding_fees_locked, accumulated_holding_fees_total, and total_available_coin response fields and the as_of_round request parameter, as those values are not meaningful aggregates. The same endpoint is also exposed on the validator scan-proxy. The /v0/holdings/summary endpoint is now deprecated but remains available.

• Bump the minimum DAR versions to the ones from splice 0.5.7 which introduced the development fund manager as downgrades to earlier versions already fail. SV app automation will unvet those on the SV nodes. The concrete versions are:

  name	version
  amulet	0.1.15
  amuletNameService	0.1.16
  dsoGovernance	0.1.21
  validatorLifecycle	0.1.6
  wallet	0.1.15
  walletPayments	0.1.15

• Fix an issue where onboarding a new SV could fail when importing the ACS snapshot due to a vetting issue.

• Update participantAddress in scan-values.yaml and sv-validator-values.yaml to use the participant address without a migration suffix. Ensure that you override this with the correct helm install name for the participant or reinstall the participant without a migration suffix.
• CometBFT: increased resource requests from 2 CPU and 5Gi to 3 CPUs and 7Gi, and the limit from 8Gi to 10Gi to better fit observed resource usage.