Bobby Calderwood
Bobby Calderwood
In addition to schema references as strings and namespaced keywords, it would be very useful to refer to schemas via UUID.
There are several reasons why we'd want to start a CausalBase with empty leaf collections, including initializing a new data structure. Currently we're unable to do so. Repro: ``` (require...
In Datomic, the [transaction result](https://docs.datomic.com/on-prem/transactions.html) includes `tx-data`, which is a fully-resolved and complete version of the transaction originally submitted (i.e. with tempids resolved, transaction functions applied, maps flattened, etc.). It...
Currently, it is possible to view the EDN current-state of a CausalBase. However, there is no way to serialize the entire CausalBase (including history, etc.) for, e.g., putting it in...
Extracting and abstracting the main components in the API: * Event Log (producer and consumer) * Database/Index (writer and reader) Then, implementing AWS versions of these components: * Event Log...
i.e. don't leave pagination as exercise to the client. We now need to support both SQL/JDBC-style pagination (limit, offset) and DynamoDB-style pagination (limit, exclusive last key/value), so as not to...
My team ran into some issues when using KafkaStreams with Micronaut on a current project: * Returning `KStream` from factory methods seemed odd and arbitrary (what if my topology only...
Ultimately, this library could provide a data DSL for defining processing topologies that provides equivalent functionality to the Java Builder DSL. See [James Reeves' talk](https://www.youtube.com/watch?v=zznwKCifC1A) for notes about implementing the...
The OpSet backend currently supports atomic moves of subtrees per section 5.2 of the OpSets paper. However, atomic moves of values/leaves don't work the same (possible duplication). We must store...