Hi @timk11 if I look at the dashboard for mn2 I don’t see that right now and all nodes seem to be active. It could have been a (temporary) networking issue.
1 Like
There is an occasional connectivity issue between the mn2 DC and Frankfurt (possibly wider Europe). The connectivity possibly kicks in when some connections are overloaded, not clear, and not easy to troubleshoot.
Our networking engineers are in contact with the NP (over Matrix), who is trying to fix the issue in collaboration with the ISP. Networking is a challenge. Especially IPv6, since it’s still used less than IPv4. 
2 Likes
Hey @timk11 ,
I think it would be easiest to have a synchronous conversation on this topic, as it’s already taking up a lot of time and is cluttering this topic.
I’d like to invite you and anyone else interested to an open discussion around the pro’s and con’s of the different approaches that were discussed, to find the best way forward for all parties involved.
I know it’s difficult to find a time that works for everyone, so please vote on the 4 suggestions below and we’ll pick the one with the most votes, or let me know how you’d like to move forward.
- 12.12.24 – 10.00-11.00 CET
- 12.12.24 – 16.00-17.00 CET
- 13.12.24 – 10.00-11.00 CET
- 13.12.24 – 16.00-17.00 CET
0
voters
Thank you!
2 Likes
Thanks for the reply @cryptoschindler. Unfortunately those times are midnight and 4am for me so they wouldn’t work so well. Frankly I’d rather just have an answer to the one question that was highlighted on the forum than have a meeting about it. Other discussions on this thread have been quite a bit lengthier but this one never quite got finished. I don’t think we need to spend much more time discussing it but I’m interested to know what the opinions are as far as moving away from the subnet-specific threads.
1 Like
I’m afraid the proposed times are during work hours for me, so I won’t be able to attend. Given the diversity of time zones hopefully we can continue this discussion on this forum, as @timk11 mentioned above.
Now that the this new system has been used for a few weeks, I think it clearly fragments discussion regarding the proposals. I don’t think an example was ever given for how this new approach would benefit voters who are reviewing the proposals and how these relate to the state and recent history of the subnet.
Would anyone like to make a case for how the situation has been improved with this new approach?
I’m an early bird, but I can’t make 3AM CST meetings. I can make the 11AM CST meetings, but only on Friday this week. However, I’m the least important of the folks tagged here. I’ll defer to the consensus of the rest of the team. That said, I always thought from the beginning that each proposal would get it’s own post. I was a little surprised that the reviewers seem to prefer a single post per subnet. It is what it is. I tend to think this work process should be made easier for the reviewers than it is for DFINITY, but not if it is a real pain point for DFINITY. It has been working out ok to be able to search by proposal id. That does typically result in relevant hits when searching for information.
Hi @timk11,
Thanks for raising your concerns and continuing to engage in this discussion.
I proposed having a synchronous conversation about this topic because, in my opinion, the question you highlighted was already answered in the discussion. I thought a call could help us better get the point across and resolve any misunderstandings. However, let me try again here to address your concerns directly.
Clarifying the Intent of the New Approach
At first glance, it may seem like the shift to one topic per proposal is aimed at making life easier for the people at DFINITY who are currently reviewing proposals. However, this is not the primary objective. Instead, we are trying to articulate the needs of a group that is not yet actively participating in this discussion: independent community members who may want to review past proposal reviews to decide whether or not to follow specific reviewers or to educate themselves to make a informed decision on their own.
This group includes potential future reviewers or voters who may want to assess how thorough and consistent existing reviewers have been in their evaluations. This will become especially important as voting neuron grants are phased out, and we transition to an ecosystem where governance relies more heavily on independent and decentralized due diligence. Our goal is to create an environment where reviewing and assessing proposal reviews is as straightforward and transparent as possible.
Perspective of Independent Community Reviewers
For community reviewers—especially those outside of formal organizations—the previous approach of consolidating all discussions in a single thread per subnet may present challenges:
- Navigational Difficulties: Subnet-specific threads can become unwieldy over time, making it hard to connect discussions and reviews with specific proposals, especially when multiple proposals for the same subnet are submitted during a small interval.
- Diluted Focus: Discussions about individual proposals can get lost amid broader subnet discussions.
By transitioning to a structure where each proposal or multiple closely related proposals have their own dedicated topic and are crossposted to subnet specific topics, we aim to:
- Provide clarity and focus for discussions about individual proposals.
- Make it easier for participants to find and engage with specific proposals.
- Ensure that the history of reviews for every proposal remains transparent and accessible for current and future stakeholders.
Alternatives and Moving Forward
That said, we recognize that the new approach isn’t perfect as it requires a little work with crossposting and may sacrifice some of the overarching context that subnet-specific threads provided, if this is not done.
There are a few alternatives we could explore:
- Subnet-Specific Tags: Introducing tags for each subnet would allow proposals to be filtered and sorted by subnet while keeping discussions organized on a per-proposal basis. This approach preserves the broader context while maintaining focus, especially when combined with crossposting.
- Appending Posts to Subnet Threads: Another option is appending a new post to the existing subnet management thread whenever a proposal is created. The post could contain a collapsed summary of the proposal. The NNS proposal would then contain a link to this specific post, rather than the entire topic. This approach was followed for a little while in the past. Reviewers could then reply to this specific post with their evaluations, keeping discussions accessible and connected. While this approach would improve navigability, long threads could still become difficult to manage over time and proposals submitted in the same timespan might influence the experience.
A Collaborative Path Forward
Ultimately, our goal is to ensure that the process meets the needs of all stakeholders—whether they’re current reviewers, ecosystem participants, or future community members assessing past reviews. We don’t want to make unilateral decisions but rather find a solution that strikes the right balance between focus, context, and usability.
Your concerns and suggestions, along with input from others, are crucial to finding this balance. Let me know your thoughts or if you have additional ideas to contribute. Together, we can refine this process to serve both present and future needs effectively.
Below my personal assessment of the different approaches summarised in a table.
| Link to subnet topic from proposal | Link to specific post in subnet topic from proposal | Link to proposal topic from proposal | |
|---|---|---|---|
| Preserves context | yes | yes | yes, if linked back to subnet topic |
| Difficulty to assess proposal reviews | difficult | easy, if all reviews are replies to post linked from the NNS proposal and there are no other proposals for this subnet submitted during the same time | easy |
| Effort for voting neurons to follow process | difficult | easy | easy |
5 Likes
Thanks @cryptoschindler for this very helpful and thoughtful response. I might have missed some of the points earlier in the discussion, but in any case this pulls together the thinking around this question very clearly.
I’m interested to hear other’s responses to this and I’m certainly willing to adapt my approach to whatever seems to work best overall.
I agree that linking from a proposal to a subnet topic generally (rather than a specific post) made things difficult to follow when this was happening a while ago. I’d personally like to see every proposal contain a link to a specific forum post. (I think this should apply to all the funded-review proposal topics, and probably most of the rest.) Whether this links to a post within an existing thread or to a new thread / topic is largely what we’re discussing here.
For subnet proposals I’ve found that if a new forum topic is started for each proposal then a lot of doubling up can result. If it’s a straightforward “adopt” then I’ve usually just posted once, but if there’s more discussion involved I’ve felt that I ought to at least cross-link between the two relevant threads. From the reviewers’ point of view this can get very time-consuming when a lot of proposals come out at once, but I’d be very interested to hear from any non-reviewers who are reading these posts as to how this has been been affecting their experience.
5 Likes
Thanks @cryptoschindler
This approach makes the most sense. Detailed examples illustrating why have already been given above.
I haven’t seen any counter examples, but that would be useful to move this conversation forward if there’s disagreement.
← In a positive way. If they interleave each other, they’re likely to be related, the order that the proposals are adopted in may be significant, and side effects or interdependencies could well be relevant between proposals submitted around the same time.
2 Likes
Let’s conclude on how to move forward by voting on the two prominent options. I’d ask anyone who has an opinion on this to participate in the vote, but at least the current voting neuron grant recipients for the Subnet Management Topic. Does anyone have any concerns about this approach?
- Link to specific post in subnet topic from proposal
- Link to proposal topic from proposal
0
voters
3 Likes
After reading this very detailed response and after the last weeks proposals I agree on going the “new topic per proposal” way instead of using the subnet thread. My main concern was about spamming the forum with single new proposals but after all with at least 4 posts from current reviewers for each change at least for a couple of more months the subnet thread would be hard to follow and would end up ignored.
Edit: it goes without saying that in case of majority decision I don’t have a problem with either approach.
1 Like
I frequently use the subnet threads to scroll up to find discussion that I know has taken place previously, but can’t quite remember when. Out of interest do you ever do this @ZackDS? If I know the subnet it relates to, I know exactly what forum topic it will be on. This would be significantly more difficult if I also have no idea which forum topic to look at.
I don’t understand that argument that the more posts a topic has, the less likely it is to be used. I think it’s the opposite (that’s evidently the case for me at least).
I’d really appreciate an example to demonstrate the problems other’s see with subnet history threads.
1 Like
Yes I do, but wouldn’t it be cleaner to have discussions related to said subnet, that are not disrupted by post that say voted yes or no ?
But the discussions are always about a proposal + the subnet the proposal affects (points come to light by discussing the proposal)
Let’s se what the rest will vote, afaik I am minority in this, but still stand by my opinion as in opening the proposal and posting my decision on the forum link from the summary that is a new topic and only engage in discussions on the thread when it’s needed feels a bit cleaner approach to me personally, (others may have different workflow like say ending up on the actual proposal from the forum post, or from OC) many times all of us agree on the vote, so than a simple yes or no would suffice (on the topic, and not the thread). But again this is just me and hopefully at least a few users could give a vote on this that are outside the grant program, just as Tim said.
1 Like
Thanks @cryptoschindler for the detailed summary on the discussions and approaches that have been thought out.
In my experience it has been very helpful to have the context of past proposals on a subnet when reviewing a proposal. With a link to a forum post in a subnet topic this can be done with a simple scroll up and re reading past reviews that may be relevant.
2 Likes
Thank you for your feedback. It seems like a majority is in favor of “Link to specific post in subnet topic from proposal”, so we will move forward like this.
6 Likes
Public notice to all:
I have edited this post, for easy reference to all those following proposal links to that message, with this update:
- DFINITY has regained control of the SJ1, FM1, TP1, and AT1 servers. (As explained above, we had no access until this month, as we were loaning the hardware but not operating them.)
- As of today, Dec 13, we have removed the servers in SJ1, FM1, and TP1 from those DCs to repurpose them for infrastructure. This is reflected in the dashboard showing them all as “offline.”
- The AT1 servers were sold (auction in this thread), so those are not yet removed from the DC. This is being scheduled with the buyers.
2 Likes
New proposal to update the list of authorized subnets. Follow the post above for more info
1 Like
I’ve been thinking about this on and off for a while, and it’s started making me wonder how easy it could be for a node provider to own multipe nodes under different identities/entities, with the intention of eventually having full control over a subnet (controlling > 2/3 of the nodes), i.e. a sybil attack.
I’ve not looked much into what it takes to become an NP. I’d expect node providers to need to put down a significant amount of staked ICP (in order for it to be slashed if they misbehave, and to make it unlikely to be in their interests to attempt to assert control over a subnet). Perhaps this is why it’s only accessible to the rich, by design?
Asking out of curiousity and to learn a thing or two from those who know more about this 
1 Like