I'd like to nominate myself to serve on the initial Packaging Council.
I am currently the PEP delegate for interoperability-related packaging standards, as well as a core Python developer and a pip maintainer. If you have been involved in any recent packaging PEP discussions, you will probably be aware of my views, and my approach - if elected, I plan on carrying that over into my role on the Packaging Council.
In particular, I feel that it is important for the new council to adopt a hands-on approach to standards discussions and proposals, actively working with PEP authors to help them to achieve the best possible outcome. Many of our standards are complex, and getting as much input as possible throughout their development is critical to achieving the best results. I do not believe the council should stay silent during PEP discussions, reserving their views for when they are asked to make a decision. Instead, they should be open about their priorities and views, and take the time to be involved in the whole process.
Having said that, I'm very aware that the current standards process - both the PEP process itself and the community discussion - can feel very confrontational, often acting as a deterrent for people who want to propose a standard. I think it's critical that the Packaging Council looks at ways to improve the process, making it more open, inclusive and welcoming. It's still important that we ensure that our standards are precise, robust and effective, but that can be done in ways that avoid authors feeling like their proposal is being blocked or criticised, and we need to make sure that happens.
I'd also like to see the Packaging Council take a wider view of what "packaging" involves. Historically, the community has mostly focused on projects that are intended to be built as wheels, for publication on an index like PyPI. I'd like to see us look at other ways in which people use Python, and how we can help them:
- People building and distributing standalone applications using Python.
- People working in more adhoc environments - scripting, or using tools like Jupyter.
- Distributors and repackagers
One of the big problems here (and indeed with anything around packaging) is collecting feedback from our users. There are significant portions of Python's user base who never interact with the packaging community, and who have no way of communicating what matters to them. I'd like to see the council look into ways of addressing that - it's a hard problem, but I think it's one we need to tackle, and the first step is acknowledging that it's something worth doing.
Finally, I'd like to say that I'm very proud of what the packaging community has achieved so far. If I look back to my early days with Python, before distutils/setuptools, PyPI and binary distributions via wheels, we have made huge advances in usability. It's hard to express how much of an impact that has had. But there are still significant challenges ahead of us. Python is being used in more and more diverse and specialised areas, and by more and more users who don't have a programming background. There's lots still to be done, but I'm confident that we can continue to improve the packaging experience for everyone.
Thanks for your consideration.