Hello - I need help understanding how to use a software defined asset in the old @repository type setup that has required resources.... the documents don't have an example where there are configured resources for an asset - it just sorta just lists the assets as something to be returned https://docs.dagster.io/concepts/repositories-workspaces/repositories#defining-a-repository. When I do it - it asks for the resource_defs... fair enough - when I try to run .to_job on the asset definition - it doesn't have that method like a graph does. So how do I give this software defined assets its resource configs? I cannot seem to use define_asset_job because that only works with code locations - not repositories.

hey @Matthew Karas you’ll want to use the

with_resources

function

@repository 
def my_repo():
    return with_resources([asset1, asset2], resource_defs={"resource_1": resource_1, "resource_2": resource_2})

here’s a link to some old docs https://docs.dagster.io/1.1.5/concepts/resources#providing-resources-to-software-defined-assets

Hi @jamie - getting closer -I got it up and running - but when I clicked the launchpad - where do I pass in the asset's configuration? Like when we use to_job on a graph - we can pass in the configuration.

you should be able to press Shift and click the Materialize button to open a launchpad to add config

Right - but can't we supply a default in the code?

how are you setting up the config schema for the assets? a code snippet would be helpful if you can share one! (for context, there are a couple ways to provide config schemas, so i want to make sure i give you an answer that matches your setup)

One sec let me sanitize it

ok since you are using new style config you can set defaults in the MyConfig class!

from pydantic import Field 

class MyConfig(Config):
    setting_a: str = Field(default="foo")
    setting_b: str = Field(default="bar")
    setting_c: str # no default

Thanks - I ended up getting stuff working by using workspace.yaml to have both Definition based and repository based configuration in two separate files. The pedantic defaults were great too.