Introduce ACD-G (All Core Devs – Governance)

[abcoathup]

ACD improvements

After listening to All Core Devs Execution #227 - Forkcast I am still strongly opposed to EIPIP meetings being rebranded as ACDG.

It comes across (whether intended or not) as an attempt for EIPIP to place themselves as a governing body above ACD. This is an overreach and outside the remit of EIPIP.

@poojaranjan

For many years, EIPIP meetings have served as the primary venue for EIP and network upgrade process improvements and governance discussions.

At least in recent years, network upgrade process improvements and governance discussions have primarily occurred on Eth Magicians and then adopted by ACD.

• AllCoreDevs, Network Upgrade & EthMagicians Process Improvements
• Community Consensus, Fork Headliners & ACD Working Groups
• Reconfiguring AllCoreDevs
• EIP-7723: Network Upgrade Inclusion Stages

Even the EIP/ERC split appeared by be driven by ACD rather than EIPIP.

@poojaranjan

This proposal aims to establish a single, dedicated forum for “process governance” discussions while saving ACD meeting time by removing process governance topics and allowing ACD to focus solely on technical governance.

Whilst I would be ok with a working group which makes proposals for ACD to accept/reject.

I am not ok with a self appointed ACDG making decisions that are imposed on ACD. This would be capture.

To improve ACD we should continue to do an upgrade-retro after each network upgrade (with an async component) and incorporate those findings into planning for future upgrades.

EIP process

Repeating again: the EIP process should work for all core devs, not the other way around.

There are currently 354 open PRs in the EIP repo. The problem appears to be a lack of resourcing. EF should have a dedicated person who’s only responsibility is making the EIP process smoother, getting PRs merged or closed, updating the website and making life as easy for EIP authors and ACD coordinators as possible.

Rebranding (but not to ACDG) may help with more exposure, but the problem is lack of resources. Note: I spend (unpaid) time every weekday assigning EIP/ERC numbers (I’ve assigned 2/3rds of the numbers since we split EIP from ERC).

EIPIP should be narrowing their remit to focus on improving the EIP process, rather than trying to expand their remit as a way to get more exposure.

ERC process

ERC process should be completely split from EIPs.

Given that there is a lack of resources, stretching the existing editors across two layers makes the processes worse and doesn’t allow for innovation to meet the needs of the respective layer.

There are currently 174 open PRs in the ERC repo. When authors ask how long to get a draft ERC merged, my answer would be weeks to months (unless you are sharp elbowed, but the process should be blind, as well as fast for every author).

---

Finally

Just wanted to add that improving ACD process, EIP process & ERC process are all really important and I am keen that they get the appropriate attention and resourcing to make that happen.

EIP editing and ERC editing are thankless tasks, so is process improvement. So thanks to everyone involved.