[WIP] Reboot of CEP7 development #296
Conversation
|
@gonuke Thank you for putting this together! If I understand correctly, there is now going to be a process to become a Cyclus developer? |
|
@jbae11 - This is not about a process to become a developer - we don't need an more obstacles there. This is about a process to govern a community of those who effectively volunteer their effort to that community. |
Fix typos - thans @jbae11
| governance within this organization. Membership in this organization is | ||
| governed by a process defined below. | ||
|
|
||
| Council of Principal Investigators |
There was a problem hiding this comment.
Some thoughts about the council:
- If a community manager exists, their perspective would be really useful in the council. Maybe it would be helpful to reframe this council as a "leadership" council that is encoded to have a set of PIs on it
- The perspective of a person who is exclusively a user, and not a developer, can be helpful at a top level council making decisions. Would it be helpful to add in a role for this council that fills this representation?
There was a problem hiding this comment.
Two thoughts on these two good thoughts.
- With respect to the community manager, I think this makes sense in principle. But, in practicality: No community manager exists at this time, and it's hard for me to imagine a Cyclus community manager that isn't PI (or Co-PI) of any funded work. So, if a Cyclus community manager exists, that person is probably also already included as a PI. My tendency would be to leave this as-is until there is, someday, a non-PI community manager.
- With respect to the person who is exclusively a user -- again, great idea in principle, but in practice, I see some issues. A single user is never representative of all users. And, a user who is exclusively a user but also sufficiently interested to be part of a council of PIs, is very likely, at least in the Cyclus universe, to become a developer fairly quickly. I would suggest that the council should simply be expected to to gather population-level input from the whole exclusive-user population on a routine basis, especially when that input is useful for decisions.
There was a problem hiding this comment.
It seems like we still need to define how membership on the Council of PIs is formally defined (given the ???? below :) ). I know that we once discussed it being PIs who are explicitly funded to work on Cyclus, but I wonder if that's complete enough? We could always start with that and go from there.
| technical grounds. | ||
| 6. Changes are adopted when at least one member of the Organization has | ||
| approved the changes and no members of the Organization are blocking the | ||
| changes. |
There was a problem hiding this comment.
For more substantive changes (new features maybe?) it might be worth adding in a requirement that a PR is open for a period of time (perhaps two weeks?) before merging is allowed. That way community members have time to weigh in even if they aren't able to attend a development meeting where new features are getting pushed forwards.
There was a problem hiding this comment.
I'm a little concerned about this 2 week timeout. Sometimes event big changes need to happen quickly. I'm sure that most PRs stay open at least that long anyway, but when one needs to be merged quickly, it would be a shame for this kind of rule to stand in its way. There's nothing fatal about merging something that someone may wish fixed later -- it's not ideal, but it doesn't become set in stone. If someone's PR creates a sufficiently serious issue for someone that they would have otherwise stood in the way of it being pulled -- they can always create an issue and PR to fix it (or even revert it). There's no deadline one that.
|
|
||
| :CEP: 7 | ||
| :Title: Cyclus Community Constitution | ||
| :Last-Modified: 2015-08-18 |
There was a problem hiding this comment.
If this CEP gets updated as a result of this meeting (or the hackathon) then we should update this (or do we have a bot that handles it?)
| :Last-Modified: 2015-08-18 | |
| :Last-Modified: 2024-03-21 |
There was a problem hiding this comment.
I don't think there's a bot... We probably need to do that manually and the clock is still ticking. Maybe I should look into a bot...
| :CEP: 7 | ||
| :Title: Cyclus Community Constitution | ||
| :Last-Modified: 2015-08-18 | ||
| :Author: Paul Wilson |
There was a problem hiding this comment.
Do reviewers count as authors for CEPs?
|
It appears I've been ignoring this review request for 5 years. Madicken (@munkm ) seems to have done a recent review. Is my review still needed? |
|
Since CEP7 is about governance, I think your input would still be valuable |
| Best Practices | ||
| --------------- | ||
|
|
||
| (should these be in their own CEP) | ||
|
|
||
| 1. Substantive changes to the work of others should require those others to | ||
| comment during the review process. | ||
| 2. Individual pull requests should be as small as possible to facilitate | ||
| timely review. | ||
| 3. Before embarking on long time-intensive changes, it is wise to collect the | ||
| opinion of the |Cyclus| community on the value of such changes, generally | ||
| through the mailing list. |
There was a problem hiding this comment.
I like the idea of having a development best practices CEP (I think this is what is commented on in l135?). This section isn't specific to governance and broad decision making, but it is something we want to encourage (or require?) community members to practice.
|
I think this looks good. |
|
Alas... the document is actually incomplete, as written, and I'd welcome PR's or other modes to suggest improvements, particularly to those incomplete places. For example:
|
In light of possible expansions and changes to the Cyclus community, I am rebooting the consideration of CEP7.