Robert Craigie

This is actually independent from the use of custom output, this can happen in some situations regardless.

I will admit, going with Pydantic as the data representation was a naive choice. I did not thoroughly consider other options as I was much more familiar with Pydantic compared...

> I agree that certain options might be useful in the "runtime" layer, like async. However the data representation I'd keep in the generation layer since it might require additional...

@landeholt We already do this: https://github.com/RobertCraigie/prisma-client-py/blob/main/src/prisma/generator/generator.py#L213-L217 So the issue appears to be caused by something else :(

@caesariWilliam looks like something went wrong with Docker's caching, due to the error message including `4.11.0` when it should be `4.15.0` if you're on the latest version. Have you tried...

I haven't encountered this myself, what FastAPI version are you using? I'd currently recommend searching the FastAPI repository to see if anyone else is running into this if you haven't...

I can't reproduce this issue unfortunately, can you please share these details: - Operating system - Python version - Pydantic version (with `python -c "import pydantic.version; print(pydantic.version.version_info())"`)

This is likely because the Prisma version is outdated, I haven't published a new release for the last couple Prisma versions. Could you try installing prisma-client-py from GitHub? I'll hopefully...

Can you share more about the intended use case for this decorator? I'm wary of including something like this in the core library because connecting the client can be a...

Thanks @LucaTabone, those are all great points! Are you aware that we provide a context manager as well? I think that the majority of your points also apply to the...