Discussion Post on the Proposed Changes to the 'Core Dev' Structure #147
Replies: 4 comments 3 replies
-
I am generally in favour of opening up core devs and strongly in favour of shifting the group away from governance. My main comment is that if we want the resulting open group ("open" as in "anyone can join without a voting process") to be a "technical advisory body", there need to be clear rules (e.g. with respect to the technical nature of the contributions) and requirements (e.g. with respect to activity) so that we can maintain a meaningful membership list. Moreover, these rules need to be applied across the board, both to "grandfathered" members and to newcomers. Without rules and requirements, we end up with a forum rather than a body. There's nothing wrong with the forum model; it just seems duplicative and of limited value, given venues for open interaction already exist, and makes me wonder whether we need core devs at all. Re the specific questions:
I'd avoid optimising early for a deluge of newcomers. My expectation is that separating governance from core devs will make this group boring (in a good way), so I'd expect those who join to be genuinely interested and for little gating to be required. I could be wrong, but always best to default to openness and adjust as necessary. |
Beta Was this translation helpful? Give feedback.
-
Openness makes sense if the group doesn't have governance responsibilities. The hard question seems to be how to achieve that (i.e. who else takes them). |
Beta Was this translation helpful? Give feedback.
-
Ideally, most discussion would happen async in FIPs and FIP discussions. Unfortunately, that's rarely sufficient when a FIP reaches the late stages. I wouldn't make the meetings completely open as they're intended to be a high-context forum for technical discussion on open FIPs and completely open meetings aren't really conducive to such discussions. However, as long as we find a way to separate technical discussions from the issue of governance, I'd be all for making these meetings "more open". That said,
I would caution against this as synchronous calls are expensive. I've seen OSS projects run into significant trouble because they were too open. This can be managed in an async setting, but gets very difficult in a synchronous meetings where loud voices can become unmanageable distractions. I.e., the reason the hacker news comments are such a cesspool: Those with the most time to pontificate are usually the least qualified; everyone qualified has has better things to do than to argue with random people on the internet. If the core dev calls become too open, the core devs will go elsewhere. This, e.g., happened with the IPFS calls way back in the day because they became too focused on "show and tell"/Q&A so many of the core devs stopped showing up because they had jobs to do. That's not to say that the community calls weren't wonderful and absolutely necessary, but there needs to be a space for high-context technical discussion. |
Beta Was this translation helpful? Give feedback.
-
I will echo the sentiments others have expressed (in this and other discussions) about wanting to logistically separate the core devs group from governance. I don't have the solution for how to handle governance outside of core devs or the process for getting there, but I do believe FF/PL are full of smart people who can figure that out given a mandate to do so. It's not a bad thing for core devs to weigh in on governance from time to time, but tasking core devs with governance is a burden that takes time and energy away from implementation, technical knowledge sharing, and coordination among teams. As far as openness, I think we are afraid of too many "what ifs" in regards to opening up core devs. I am in favor of following the Ethereum model and allowing self-nomination based on contributions to the Filecoin ecosystem. If we are worried about disruptors or detractors, we can put in place mechanisms for dealing with those "what ifs" (e.g., "voting off the island"); however, I think creating a hierarchical core devs structure (based on worries about events which may never even come to pass) creates an administrative overhead/burden which likely isn't really necessary and again, takes time and energy away from more productive efforts. I recall @jennijuju raising a valid concern about how we would address any potential network vulnerabilities that pop up in the future, and I do think we should have a subset of the core team (e.g., protocol PMs and team leads) identified to address such issues, but that could be something as simple as an ad hoc private channel on Filecoin Slack. |
Beta Was this translation helpful? Give feedback.
-
Introduction: Proposed Changes to the 'Core Dev' Structure at Filecoin
I am opening this discussion post to address two main objectives:
Introduction
As Filecoin has scaled and matured, reliance has been placed on existing Core Devs to function in various roles, including governance. In recent months, there has been active discussion about opening the Core Devs group and shifting away from it acting as a governance group. A reference to a previous related discussion is included at the end of this post.
Openness and Diversity
The overall preference appears to lean toward an open group, serving as a technical advisory body. The pros and cons of this approach are evident:
The suggestion is to allow open membership for anyone within the ecosystem who meets certain criteria, with the details of this to be openly deliberated by existing members.
The Way Forward: Proposed Changes
The way forward is outlined as follows:
The Ask: Open Questions and Community Input
Open questions:
Please weigh in with ideas and suggestions. Sufficient contributions will allow this topic to be added to an upcoming meeting's agenda.
Reference to other posts
For further insights into past discussions regarding the structure and function of Core Devs, review 'Core Dev' Purpose and Structure by Kaitlin Beegle.
Beta Was this translation helpful? Give feedback.
All reactions