Oliver Mannion

Thanks so much @jsibbison-square for this significant piece of work. I've pushed some refactoring to use pure functions rather than a class, as there's no permanent state to capture, and...

Thanks again @jsibbison-square - I've merged this and will address the above in a follow-up PR.

Glad to hear you'll be using this soon. I've raised #136

> For example, I like to use the uv environment manager for python projects, but many public python repositories do not use it, and if I want to make a...

Duplicate of https://github.com/ebkalderon/tower-lsp/issues/399

Retry backoff logic can be set by providing a `backoff_policy` on connection. The default backoff policy is exponential backoff: https://github.com/snowflakedb/snowflake-connector-python/blob/91aa0fcd23c1d87d069cd9989963c8fbcd1091cf/src/snowflake/connector/connection.py#L183 Retries will occur (with timing determined by the backoff policy)...

An even more minimal example that fails: ``` --- action_options: Append: max_actions: 2 --- action Init: history = [] action Append: history.append("A") ```

FYI - I'm still experiencing a deadlock when running with 1fd9b76 and setting max_actions, eg: ``` --- action_options: Append: max_actions: 2 --- action Init: history = [] action Append: history.append("A")...

To mimic SQS's reservation behaviour I've limited the concurrency: ``` action_options: ReceiveSQS: max_concurrent_actions: 1 ``` Although in practice the concurrency in my system is higher, I don't have interactions between...

I see... that would be a model that's closer to how SQS works, but I still have the same problem that if `ReceiveSQS` runs concurrently, and `process` is non-atomic, then...