I just read this thread tonight; well, I pretty much agree that SC’s decision should have as much continuity as possible, even if all SC members are changed for the long-term roadmap. (e.g free-threading is quite long-term project from the view of whether we will promote this project as one of the official distributions or not, I hope that we will have the same consistent decision even after few years later )
some voters might be more conservative in their vote, and would prefer to vote those who has experience being in SC, which basically is “problem #2” you mentioned above.
I agree that people have preferences when we have to vote on committee members (even for me), but the problem is there are no natural-born SC members.
This issue is already happening not only for the SC but also for the real world where we live, most of tech manager has experience of early failure, every tech interviewer has failure experience when deciding candidates should be up or not. so I think that we should observe how outside models solves this problem.
Here are some ideas without considering it is realistic or not.
For decision consistency, some companies have kinds of leadership principles that will help people make decisions without reference. So maybe we may need kinds of SC principles; if the principle is outdated, we can revise the principle when we think that is needed.
For the experience issue, the shadowing program will help people who want to get experience of the SC decision process. I am not sure how many people want to participate in the shadowing program, but if it is possible, it will be helpful to people who want to get confident when they become SC members one day.
(Of course, those participating in the shadowing program cannot directly or indirectly influence decision-making, and if necessary, may be required to sign an NDA not to disclose information outside of internal discussions. So, because of this, it is questionable whether it is realistically possible in the current situation. Because I also have no SC experience…)
Resigning from responsibilities is generally a difficult decision because you’re disrupting other volunteers’ work, especially if the team is only 5 people nominally.
Edit: sorry, I realise I’m responsding to an old post. I’ll shamelessly blame Discourse for that!
I don’t have time to make a more substantial contribution to the discussion, but I’m not sure there’s a real problem to solve here. Rather there is worry that it might be a problem, and I think we could safely wait until the problem manifests before solving it. (YAGNI.)
Agreed, and looking at Hugo’s nice graphics, it’s pretty obvious that we have always had enough overlap to guarantee continuity.
Note that more important than keeping people in power, is having documentation of the processes they use and proper handover. In my experience, this is far better than trying to achieve continuity by having people do longer terms.
fwiw, in the year and a half since starting this discussion, my own thinking has evolved, and I tend to agree with the comments that we don’t need to solve a theoretical problem that we haven’t actually saw in practice, and that if we end up with elections where the entire presiding council is replaced - maybe there was a good reason for that.
I agree that it doesn’t matter if we have 2 roll off one year, and 3 the next.
I’ve served on volunteer boards that do this, with 3 year terms. The important thing to specify in the document describing the setup is that the “seats” have these terms, and people fill the seats until that seat is up for election. That is, seats 1, 2, and 3 are elected in even years, and 4 and 5 in odd years. If, 1 year into it, someone in seat 1 quits, then that seat is still elected in the next cycle with seats 2 and 3. Otherwise you’d have seat 1 elected with 4 and 5. Not only is this hard to keep track of, you might end up with 4 or 5 seats being elected one year, and 0 or 1 the next year.
Specifying it as seats being elected in specific years addresses these issues.
I don’t recall if we have a mechanism to fill vacant seats. This becomes more important the longer the terms are. Every board I’ve been on handles this by the board itself filling the seat, until the next election.
regardless, this discussion should ultimately turn into a PEP-13 modification that we all get to vote on per PEP 13 changing this doc - sounds like we’re getting there in terms of what could be proposed.
another topic that was floated was the idea of term limits. that belongs in its own discussion and change vote though. it’ll likely go more smoothly if each individual characteristic of pep-13 change is its own vote.
This is how PEP 772 defines the terms of Packaging Council members. Cohort A has two members, Cohort B has three, and only one of the cohorts is up for re-election every year. We also explain how the initial election results fill the cohorts. We deliberately modeled these terms on how we[1] envision the SC terms could be amended for continuity.
Coming back to the original proposal, I’m with Raymond on this one.
Unlike in some other forms of governance, core devs will not suddenly become unavailable to ask questions, so there isn’t much risk of losing institutional knowledge.
At the same time, having longer fixed terms can result in people not stepping up as SC members (after all, this is a lot of work), promoting change on the SC becomes harder and the risk of having SC members becoming inactive during their term increases (which we cannot afford at the SC’s small size).
It is also important to strike a good balance between size of the SC and term length. Fewer people in a council means more power for each of them. By extending the term length you increase this power even more and limit the possibility of the voting membership to have influence on those powers.
I feel the same about the current packaging council proposal.
To make an analogy: What we need is a doc fix, not a PR to change the code.
Early on in the SC terms, it was very helpful to have some continuity to bring the Developer In Residence program to fruition. Having some SC knowledge transfer from strategy to funding to implementation was very helpful in contributing to the success of this effort.
I’m fine with either the current 1 year terms or staggered 2 year terms (with no shame if someone needs to step down for any reason).
)