so, i created this blog post with a step-by-step tutorial and my personal real-world-use case which brought me here. i don’t want to bash, yet i still think the docs are incredibly bad to get started, so this is my short-term contribution to anyone who is trying to get up and running as well. https://www.flypenguin.de/2023/04/30/getting-started-with-dagster/

Hi Axel, thanks for sharing! #dagster-showcase as well as #dagster-feedback might be a better place to surface content like this! cc @Tim Castillo / @Odette Harary who are actively thinking of our docs

Thanks for the feedback @Axel Bock . This inspired me to write a few docs PRs: • https://github.com/dagster-io/dagster/pull/14034 • https://github.com/dagster-io/dagster/pull/14030 • https://github.com/dagster-io/dagster/pull/14027 Your post led me to believe that our docs over-emphasize ops and graphs for these kinds of use cases. If it were me, I would implement your use case using dynamically partitioned assets instead of ops and jobs. Here's an example of a pipeline that uses dynamically partitioned assets, to create a file for every object observed by a sensor: https://github.com/dagster-io/dagster/tree/master/examples/assets_dynamic_partitions