Feedback from the 2nd Python Tooling Workshop at the ARC festival.

[samcunliffe]

Just noting down comments from attendees so we don't lose/forget them.

• could we do all of this in brew or something?
• should we make the gh command private by default (I think no)
• if no (above) should we add a warning about making code public?
• should we have a GUI for cookiecutter (I don't know of one already)
• See if it's possible to swap in pixi for uv #559
• Lots of red text when running tox on a fresh cookie. #560
• should we add an option to Dockerise the package?
• should we remove the gh install from the tutorial?

[samcunliffe]

About this one:

should we have a GUI for cookiecutter (I don't know of one already)

... my feeling is almost certainly "no" unless someone makes a cookiecutter GUI. Googling around for "cookiecutter GUI" --perhaps obviously-- leads to a lot of templates for making a GUI, but no actual GUI for the cookiecutter tool.

The main points of opinion voiced in the workshop: researchers and research software developers are going to need to become familiar with the terminal anyway, it's not necessarily "simpler" than just asking questions at a prompt.

If we manage to get

• Make a template from the generated cookiecutter project so that people can one-click use #533

I reckon that would satisfy this.

[Saransh-cpp]

Re: GUI

It would be nice to keep an eye on https://github.com/aryamanjeendgar/CopyCuTTer (TUI for the cookiecutter and copier projects). The project was created as an IRIS-HEP (Scikit-HEP is an IRIS-HEP project) fellow project, but I guess it went unmaintained once the fellow moved on. I see some recent revival activity/efforts in the repository by Henry (scientific-python/cookie's lead maintainer).

It will still be a terminal-based thing, but at least it will be more closer to a GUI.

[matt-graham]

@all-contributors
please add @Saransh-cpp for talk tutorial eventOrganizing.
please add @samcunliffe for tutorial.
please add @matt-graham for tutorial.