You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We get most of the benefit from the verifier cache internally in the mempool. We do a lot of tx verification in various mempool operations. For example after aggregating txs together we verify the aggregated tx. Each verification leverages the rangeproof verifier cache as we do not want to verify the rangeproofs repeatedly each time.
But we may want to consider disabling the cache when we actually verify a new block.
Blocks are consensus critical, transactions are not.
We would need to take care to handle edge cases where a potentially invalid tx entered the mempool.
This makes me hesitate a little.
The text was updated successfully, but these errors were encountered:
I think we want to do the following initially -
DefaultHashable
for the full output (identifier + rangeproof)Additional possible improvements -
We get most of the benefit from the verifier cache internally in the mempool. We do a lot of tx verification in various mempool operations. For example after aggregating txs together we verify the aggregated tx. Each verification leverages the rangeproof verifier cache as we do not want to verify the rangeproofs repeatedly each time.
But we may want to consider disabling the cache when we actually verify a new block.
Blocks are consensus critical, transactions are not.
We would need to take care to handle edge cases where a potentially invalid tx entered the mempool.
This makes me hesitate a little.
The text was updated successfully, but these errors were encountered: