Pavel Kirilin

The suggested implementation for non-prepared queries seems a little bit too complex for me. Because currently you can do it like: ```python from scyllapy.extra_types import Double, TinyInt scylla.execute(statement, (Double(5.0), 100000,...

I don't like the idea of another parameter in execute. Since you can wrap your values in extra_types. I've made an implemetation that stores information about types of query inside...

No. Because in batches we don't have access to queries. I wanted to implement it, but couldn't do it now.

Sure. But I'm not sure if it's possible. Anyway. Maybe we will be able to implement it later.

I guess we need to rebuild middleware stack by hand. ```python def startup(app: FastAPI) -> Callable[[], Awaitable[None]]: async def _startup() -> None: """Startup.""" app.middleware_stack = None # Do all startup...

I'm very glad you liked the project, and the issue seems to be very reasonable to me. Also, we can benefit from it by making support for custom types easier...

Regarding headers. Header copying happens only in the RabbitMQ. Not all brokers support headers for messages. The main reason we made it this way is because We wanted to give...

Okay. Now I cannot see how we can use serializers. For example, currently you can create a cbor formatter to make all your tasks serialized using this format. ```python class...

Hi @Kaelten, it took a long time, but this feature is finally coming in the next release! Currently we will only have a hybrid-s3 which works exactly as you said....

Hybrid-s3 storage is released in [0.6.0](https://github.com/s3rius/rustus/releases/tag/0.6.0). S3 storage without storing files locally will come later. Leaving this feature open until new storage is released.