One concern: Some neurons exist for the sole purpose of using the manage_neuorn topic to control a big chunky neuron. So they maybe have one ICP in them and they mostly vote on manage neuron proposals. Half the configured neurons need to vote and they are more of a record-keeping operation than an investment seeing to influence governance and/or trying to yield farm ICP maturity.
This is only relevant to the extent that there is not just one reason to create a neuron. People may not want their management neurons indexed. Is there a good reason for this? I’m not sure. Your post clarified for me a couple of things about revealing neurons, so let me say a couple of things and you can tell me if they are true or not:
A neuron has an ID.
There is no way to tie a neuron ID to a Principal.
A neuron has an account(I send my top-ups to it) but there is no way to derive the account from the Neuron ID.
There is no way to see how a neuron voted(although I don’t think this is true because I’m pretty sure I know which neurons voted against the ICDevs proposal and someone gave me that info.) Where did they get it? Did they have to query all known ids to see what their votes were? So having this index would let you query all votes for a proposal?
It came unrequested! I was surprised it was queryable. I think it was only because some of the votes were the sequential seed neurons. It is actually maybe a great example of why one wouldn’t want them indexed.
The foundation is going to vote in favour of this proposal because this adds transparency on the NNS and addresses the current unsatisfactory situation where a lot of information about a neuron can be retrieved given the ID of a neuron. In particular, the foundation acknowledges the need for privacy and will make a follow-up motion proposal regarding which properties of a neuron should be private, which one should be public, and also about giving a neuron holder a choice on that. For instance, the voting history of a neuron should probably be private by default (which it is currently not), but visible for neurons that are listed as a followee option.
I think this is a fantastic idea. It seems reasonable that vote history should be private by default and that public and/or registered known neurons are the most likely neurons to want/need to make their voting history public.
I believe it to be true that every neuron has a unique id, although I would need Dfinity to veridy this.
I know of no way to tie a neuron id to its owners’ Principle ID for neurons created after Genesis. For Genesis neurons, I believe it is possible because I remember seeing them tied together on ic.rocks.
This is correct as far as I know. However a createdOn datetime is recorded when a neuron is created. I wonder if it’s possible to use that datetime and the ICP transaction history to figure out the account number feeding the ICP to the new neuron… seems like that might be possible.
Currently the governance canister contains the voting history for recent proposals for each neuron. You can see this on the governance tab of the ICA dashboard (also accessible via the API for the dashboard). Of course, this is only true for known neuron ids (which is mostly genesis neurons)
Probably a conversation for another day, but i think theres a good argument that recent voting history should be mandatory public. It would allow anyone to conduct a “recount” of any recent proposal and provide a second verification beyond the nns code audit. The real life example would be the equivalent of allowing hand recounts in democratic elections where the vote tally is close or there is reasonable doubt regarding the operation of a voting machine.
I think there are steps we could take to both protect privacy of neuron holders and increase transparency of neurons themselves.
I thought DFINITY was going to abstain from voting on governance proposals. DFINITY and the ICA have been voting on multiple governance proposals recently. What has changed? Did DFINITY/ICA announce this change?
The foundation considers a motion proposal by the community a mandate to work on a topic and feed that topic into the roadmap process. As part of that process, a topic is first scoped, meaning it is investigated what the topic actually entails and how it can be realised. Part of that scoping is a conversion on this forum and then eventually another motion proposal to agree on the scope. We are preparing a blog post on how the foundation aims to manage the roadmap and how to work with the community on the roadmap and individual features such as working groups etc. So please bear with us for a bit for a better answer. Thanks!
The work you’ve done producing visibility into DFINITY neurons and voting power is amazing, and I want to thank you for this.
Any update on how or when this Neuron Indexing Proposal work (which passed last month on the NNS) might fit into the DFINITY roadmap for the next few quarters?
Hi all, (for context, I am a researcher in the NNS team)
Please know that this feature is on the NNS team todo list.
As we are at the last mile of the SNS feature, which is the big feature scheduled for the Carbon milestone, the NNS team is focused on shipping the SNS. We might thus only get to the neuron list afterwards.
I hope that makes sense. Please let us know if you have more questions!
I want to bring up the importance of this feature, as I believe that both:
The release of all investor, seed and team neuron addresses (without doxxing individuals)
Adding neuron indexing
would dispel much of the concerns around decentralization and what happened at genesis.
To be fair, we are now at 9 months since the “SNS last mile” described above.
I understand this is a hard feature to prioritize, but recent events such as the Twitter space and public critiques have highlighted the lack of transparency around historical neuron data that is publicly accessible.
If the DFINITY team is too busy/priorities are elsewhere, maybe someone close to the code can outline a rough set features required to implement neuron indexing.
Also, I’d happily chip in some ICP or time to make this happen.