Q:A simpler way to use it

[Euraxluo]

I am very interested in this project and can feel that the code and architecture are quite complex

[matiasdahl]

Thank you! Definitely agree with you on "the code and architecture are quite complex". Now there is eg low level stuff that leak up to the user (like eg compiling the front end). This introduce both complexity and also make pipelines run slowly.

---

To make this easier, I am working towards the following smaller goals (to both simplify and speed up usage):

• Offer pip package with precompiled front end
• Offer pip package for main pynb-dag-runner library

These are initial steps and there are likely many other way to further simplify things.

As a goal: If one have a repo with some notebooks, one should be able to convert them into a pipeline w. reporting by adding a few yaml:s and maybe a driver script. I am not completely sure about options here, but this should be possible with Github actions.

---

I have only recently moved to public project planning and to write documentation. So please ask if anything is unclear, or if something can be expanded.

Project planning: https://github.com/orgs/pynb-dag-runner/projects/2/views/3

• topics "Speedup for faster feedback" and "Refactor for simplicity" might be most relevant

Documentation: https://pynb-dag-runner.github.io/pynb-dag-runner/home/

---

Please also let me know if are interested in contributing (also later), or if you have some particular use case in mind. All ideas are welcome. We can also have a call/email in profile.

[Euraxluo]

I want to use him to do automated testing and test report.I have already used MLFlow to success