Custom features in my own fork: SQLAlchemy, GH Action for PyPI. To PR or not to PR?

[wschuell]

Dear maintainers,
First of all thanks a lot for your work!

I am an engineer in a team of 15+ researchers creating each their own scientific software. I am very happy to benefit from your template, but for the specific things that I am promoting and regularly using (e.g. use of databases) I added features on my own fork, and from just a couple of details it starts to be more significant. I am regularly updating from the main repo but maybe others would want to benefit from that.

The main features so far:

• choice of using databases or not
• choice of which DBMS to support (for the moment duckdb,postgres,sqlite)
• whether to use sqlalchemy or not
• a github action worflow to automatically publish on pypi
• choice to limit the matrix OS/py versions in the tests: it can have a big impact on the GHA credits for private projects, especially when tests run into race conditions and lock processes

If you think there is a need (vs being too confusing for the general audience) I would gladly make a PR and coordinate with a maintainer

The changes are to be found on the 'databases' branch: https://github.com/wschuell/python-template/tree/databases

[sjvrijn]

Hi @wschuell, thanks for getting in touch, and glad to hear you're benefitting from the template!

The PyPI GitHub Action workflow and CI run limit are definitely welcome (see in particular issues #372, #373 and draft PR #376 for the former).

For the database related features, I'll be discussing with my colleagues on how relevant they are for our intended audience, so I'll get back to you on that.

[wschuell]

Hi @sjvrijn ,

• pypi: I wrote on the PR draft
• CI run limit: I'll make a small PR in the next days. One could just use the limit at the beginning (heavy dev/loads of commits) and retake the copier questions when the project reaches maturity; or to have the limit conditional on the branch name
• DBs: happy to at least have triggered your interest! My aim is to really lower the barrier for people to benefit from them/encourage DB vs custom files/separate or huge CSVs (less maintainable, more obscure -- not even talking about scaling and integrity issues for big projects)