Join the PyPA?

[pradyunsg]

I've mentioned this on various channels before -- Poetry is an important project related to Python packaging. I think it would be a good idea to join the Poetry project to join the PyPA.

Other than needing to be under the pypa org on GitHub and needing to adopt the PSF Code of Conduct, joining the PyPA will bring basically no "enforced" changes to how the Poetry project works. If you're interested in the details, see https://peps.python.org/pep-0609/ (goals does a good job of summarising things).

Thoughts?

[neersighted]

Hi @pradyunsg, thanks for reaching out!

This would represent a major change in Poetry's governance and autonomy as a project, and thus will likely take a fair bit of time to discuss. So far there have been no outright objections, but there are also many major concerns and obstacles to any proposed migration to the PyPA, as well as a number of contributors who have not yet found the time to weigh in.

If you are interested in addressing concerns, objections, and questions as we formulate an official response, please feel free to join the Discord server, where I can set up a special-purpose room for this discussion. I'll try to herd cats so we can figure out what we want to do, and what our reasons are, as soon as is practical.

[pradyunsg]

I’m already on the Discord (same name). I’m very curious what “major change” to governance needed here is, or “major concerns” folks have around this. As I said, there’s not really anything that being in the PyPA affects in terms of governance/operation of the individual project (beyond the CoC).

[dimbleby]

Not that anyone asked me, but I reckon I'd be neutral on this.

Nothing in PEP609 strikes me as especially interesting and @pradyunsg is taking pains to say that it basically wouldn't make any difference, so I kinda don't see the point.

On the other hand, and for much the same reasons, I don't see harm in it.

[pradyunsg]

It wouldn't make a difference in how the project operates. That doesn't mean that it'll change nothing. Two relevant things that my sleep-deprived brain can think of at this moment are:

1. the perception of the project as well as the PyPA. ¹
2. provide the ability to this project, to more easily utilize PSF's capabilities to handle funds. This can be useful if the project gets grant-funding or substantial donations, and needs a fiscal body to help with handling it.²

Footnotes

1. As much as some folks want being the PyPA banner to mean nothing, it does -- there's a level of trust and "officialness" that comes by being under that umbrella. OTOH, not having a project as important as Poetry under the banner seems... inconsistent to many (rightfully so!). ↩

2. Just because you can doesn't mean you have to. :) ↩

[Secrus]

[update] The discussion is still ongoing on Discord.

[pradyunsg]

To be clear, the discussion on this is taking place on a private channel within the Poetry Discord. As far as I can tell, it's poetry maintainers/contributors + me in that channel. I've not had the bandwidth to sit down and write replies to the various concerns and things being discussed there -- but I'm hoping to come around to that later in the week.

[neersighted]

For the record, this is ongoing. The largest blocker is the availability of two of the custodians of the project and largest contributors, @sdispater and @abn, who have been occupied with life and life events and unable to weigh in. We will not move forward with any sort of decision until they are able to participate in it.

The objections that have been raised are technical and social. The technical obstacles can be overcome through some combination of trust, possible changes to PyPA bylaws (e.g. the Github org requirement), and support/sponsorship opportunities that we already know exist.

The social objections are harder -- Poetry has long enjoyed autonomy and has made a noticeable influence on the Python packaging world, doing so with code + user acceptance instead of direct lobbying. That being said, I think we have a lot of value to gain by sitting at the table with other projects and soliciting more input from contributors to the ecosystem who don't otherwise interact with Poetry.

Still, enfolding Poetry into a larger organization where we are less autonomous and distinct in identity is a hard sell, and what I expect to be the ultimate sticking point.

As I mentioned above, there are compelling benefits to joining the PyPA. While maintainers can certainly participate in the PyPA PEP process as outsiders, breaking down silos would benefit Poetry both in the standards process and in gaining new contributors. Many of the experts in the packaging system work almost exclusively in the PyPA, and we may be able to gain their input and/or contributions by joining the "big tent."

Most of all I ask that people be patient and respectful while we have these discussions -- we are all volunteers and as such have limited time, as well as understandable anxiety when considering a massive change to the project like this.

Thankfully, that has been the case so far, and I expect it to be so going forward. Whichever way we leap, I would like to gather and post at least minutes of the conversations had, if not verbatim logs, both for interested parties to consult as well as for posterity.

[neersighted]

Note: this could serve as a partial continuation of #4595 if we continue down the PyPA path