Thanks; I appreciate you taking the time to respond here. As much as there are things that could have gone a lot better (at least in hindsight), I don’t think anyone here would envy being in your position here, given the overwhelming breath of the problem space, and the sometimes deafening cacophony of cats to herd. You’re also stuck inn the role of the one project manager in a room full of programmers and scientists, and as others noted we can speak very difficult languages sometimes with different expectations and ways of doing things.
It’s understandable that there’s a lot of healthy skepticism toward a relative outsider coming in with a somewhat unconventional plan to accomplish what many veterans and newcomers alike have wanted to do since forever but from experience accept as nearly impossible—unify the packaging landscape and develop a single coherent high-level strategy. However, maybe just maybe the right balanced of fresh-faced enthusiasm and out of the box thinking combined with battle-hardened experience and proven strategies, catalyzed by a healthy infusion of financial resources, will finally succeed where countless others have not. I certainly hope so…
Thanks for the clarifications. As it seems you picked up on, I think some of what might have raised people’s concerns was not only the length of time of the process, but also the apparent complexity/number of steps involved as well as the perception of it being more rigid/water-fally and less flexible/agile than you may have actually intended it.
Sure. Some here have suggested at least trying one or a couple virtual meetings to complement the threads and seeing how well-attended and productive they are and how well they are received, given they were a close second in your previous poll and the Discourse threads have met a mixed reception so far. We could also give this topic (or multiple related ones that this has organically spawned) a slot at the upcoming packaging summit, or even schedule a dedicated BoF session, meetup or sprint at PyCon.
Beyond that, as mentioned above, deep into the first strategy thread (and before I think people were aware of what you shared with us now, that there was a whole plan for working grounds, workstreams etc) there was some considerable discussion and some amount of rough agreement around an overall path forward to gather community consensus and decide on a course of action on these topics, that is somewhat of a hybrid of the accepted PEP/PyPA process and the sort of thing you propose:
- Discuss and agree on the overall process and document that in a PEP (i.e. the equivalent of your strategy document)
- Determine a “steering council” of ≈3-5 ish PEP delegates to oversee and steward the process—the PEP in step 1 would determine how and by whom these are selected
- Invite community members or teams of them to propose PEPs with different ways of addressing the overall topic (equivalent to your workgroup technical proposals)
- The community discuss, revises and iterates on the proposals, via Discourse threads and other venues (equivalent to your consensus building stage)
- A final decision is made on which PEP(s) to accept and implement based on the community consensus and the input of the affected maintainers, as determined by the council of PEP delegates
- The accepted proposals are implemented (with funding as needed/available)
This was originally proposed by @dstufft in post 225, followed up by me in post 228 and further discussed by others in the immediate posts following. There were also some other approaches proposed there, but that one is probably the closest to the one you were thinking of, and the one that seemed to garner the most support.