I’m excited about the PPC in general, and about all the great candidates in particular – a pleasure to be in such company! I like the AMA idea because the ecosystem-level problems we’re trying to fix cannot be solved by some small group in isolation; it’s pretty much for that exact reason that I’ve been active on DPO the last few years, in order to help nudge things in what I feel improves the overall user experience (e.g.[1] PEP 722 vs. 723).
IMO python packaging has historically not succeeded in having a good out-of-the-box packaging story, and that isn’t due to lack of interest of competence, but overwhelmingly because of the vast amount of hodge-podge complexity that python wraps around. C & C++ not having a standard way to build or package dependencies in particular has been a huge albatross around our collective necks.
A lot has changed from python packaging of yesteryear though, and if anything, I believe problems that were once unthinkably complex to consider for the volunteer-driven python packaging crew, are now beginning to crest the horizon of things we can actually resolve. In many ways, this kind of effort is – to my mind at least – what I’ve been participating in the last few years in conda-forge. Not to create an insular and distinct ecosystem, but to find ways to solve things structurally for a broad range of use cases, and let fixes & learnings flow back to the wider ecosystem.
For completeness, I’m part of the conda-forge core team, and quite involved on keeping things running smoothly on all kinds of infrastructure topics – from compilers to bots to CI to migrations.
by far my most controversial thread here, but I believe the outcome ultimately justified throwing the brakes on hastily adding yet another distinct config format. Irrespective of the outcome, I believe the decision process was better for it as well. ↩︎
The ecosystem-level focus here is valuable, especially because packaging problems often cross tool and platform boundaries. The conda-forge experience you describe seems like a strong foundation for turning those broad lessons into practical improvements.
Thank you for stepping up to serve on the PPC. Your background with conda brings a crucial perspective. Unfortunately for years I ran into problems (crashes) whenever I tried conda on Windows. Since Windows developers and end-users and non-conda Python users represent the majority of the Python user base, often facing problems with compiled dependencies and a lack of bundled compiler toolchains, can you promise to keep the users that avoid conda because it doesn’t work for them in mind?
The PPC will take on an important large amount of work. How much weekly time can you commit to for the in-depth review of proposed PEPs that don’t touch your preferred interests?
Obviously, the PPC is for all Python users! It’s true that my candidacy is aiming to bring some degree of representation of those Python-users coming from the conda side to the PPC, but the goal is of course not to ignore anyone (whether coming from conda, non-conda, distros, etc.).
I don’t think it’s useful or appropriate to make promises or predictions on hour per week; it will depend completely upon the proposals in flight, and the initiatives the PPC (as elected) wants to tackle or stimulate. I do plan to make time for reviewing and discussing proposals in depth, but this remains a volunteer position, not a job.
Picking up on an interesting aspect from an (almost entirely) unrelated thread
I have listed no less that 4 affiliations (including conda-forge, which never paid me anything ), but none of them have any say over how I do my FOSS work in general, or anything related to the PPC in particular.
In part, I have that many affiliations because I fiercely want to keep my independence, and being wholly dependent on one employer raises the difficulty of speaking one’s mind, should viewpoints ever diverge.
All that is to say, I can already state that my inputs & decisions – if elected – will be 100% my own, along the lines of how *I* perceive that the entire ecosystem would benefit the most.
I like the AMA idea because the ecosystem-level problems we’re trying to fix cannot be solved by some small group in isolation; it’s pretty much for that exact reason that I’ve been active on DPO the last few years, in order to help nudge things in what I feel improves the overall user experience (e.g.[1] PEP 722 vs. 723).
), but none of them have any say over how I do my FOSS work in general, or anything related to the PPC in particular.