Kostas Dermentzis

> We know when semantics of a command was invoked and when it give us back a response. That's all we care about from a black box perspective? That's not...

> We don't know the exact order, and that doesn't matter. We seem to agree that the exact order is unknown. > My understanding of the linearisability paper is that...

The dot graph can be indeed improved in the following way: If a `Response` of a command, is before the `Invocation` of another and the commands are executed from different...

@stevana my understanding is that this https://github.com/advancedtelematic/quickcheck-state-machine/issues/208 is a prerequisite for Jepsen-like tests right?

I think I found an improved way to write this, which doesn't use IORose and generalises IO ``` shrinking'' :: Testable prop => Monad m => (m Property -> Property)...

Is this necessarily added in the default schema, or can it be added in custom schemas for instances that use replication? A new schema file can be added eg at...

Alternatively an autoincremental primary key could be added in a similar way. This is not so trivial though, as it may affect the queries that db-sync performs. For some context,...

Some initial thoughts on this: > address and payment_cred tables We've made some [work](https://github.com/IntersectMBO/cardano-db-sync/pull/1396) on this. It slowed down syncing speed, so we decided to park it. It could become...

FYI DBSync now has this table https://github.com/IntersectMBO/cardano-db-sync/blob/master/doc/schema.md#pool_stat It's not populated by default only when "pool_stat": enabled is used in the db-sync insert config.

Easiest first step for this is to simply trim multiassets from outputs once every epoch. Then validate that it syncs properly on mainnet and compare the time and memory used...