This is a nice idea. But in my experience the tag system requires lots of clicking around to eventually find the topic that you’re after. I’ve used this a few times before to revisit IC OS election history, and the results don’t really provide you with much context (so you end up having to drill into each topic and scroll around).
In the examples I provided above, simply scrolling up gives you the opportunity to conveniently realise things you may not have otherwise known (and/or bothered searching for if it wasn’t presented to you).
I’d like to get a better understanding of the problem this new approach is solving. Is there a good example that can be pointed to where the current system would be inclined to hinder rather than help voters?
Well I guess it was after the fact. But it’s fine.
This sounds doable, we could do it on a “first saw it first did it” basis so it doesn’t necessarily has to be only Alex who does this, specially on weeks when there are many proposals.
Over all I liked the “old” way of doing it with each proposal directly in it’s own “subnet” thread that it is related to. My main concern was the high number of proposals that we had over the last 2 months. Also since we provide a link to each vote/review in the spread sheet I thought it was easy to follow, (unless ofc we screwed up the link ) but going forward I don’t mind posting in each individual proposal topic if this simplifies things, given the fact that so far at least the actual forum link was specified in the proposal summary and does not require additional forum search.
yeah the first to provide a review could take over this task, agree
IMO this is sth. we need to identify whether this helps the broader community or not. I meanwhile understand that you as active proposal reviewers prefer the “old way”.
one disadvantage of the old way as you said was/is, that the forum link in the proposal always pointed to the start of the general subnet thread. so one would first need to perform additional forum search as you said.
if it turns out that we all agree we do not need specific threads and that the “old way” it would be great if the proposal would directly point to the correct post in the general subnet thread as @timk11 suggested above:
my personal point here is that we cannot expect newcomers to know the whole forum structure, so it would be great if they would be pointed directly to the right place of the conversation when they look at the proposal details on the ICP dashboard. the perspective of such users should not be ignored.
that’s all from my side to this topic for now. looking forward the comments of @cryptoschindler to this change and all of your feedback so far
Thanks @marc0olo, I would just add that the original old way was to announce the proposal on the appropriate history thread, and then the NNS proposal would link to that specific post (rather than the top of the thread). I’m not sure why this stopped happening in more recent proposals.
I don’t expect it would be a big change to the new way to just post on an existing thread instead of creating a brand new thread - but @sat would be the authority on how practical that would be
Agree and it makes sense since the Grant’s will end and the main focus should be on the rest of the community and “newcomers” with the ease for anyone to be able to review and make a decision on the proposals, to which I may add that great progress has been made (*hats of to the team btw) in including a very comprehensive summary including links as how is Decentralization calculated and howto perform additional analysis.
These are DCs that are anyway rarely being picked by the DRE tooling since they do not contribute a lot to the network decentralization, so IMHO this is a positive change.
Hopefully some other NPs will pick up the offboarded nodes and continue operating them from another country (although that needs to be discussed elsewhere).
I see that a fresh set of 18 proposals has just gone up, each with a new forum thread. Cross-posting between threads for 3 proposals is easy enough but for 18 will take a lot of extra time.
@sat What procedures or rules are there around offboarding data centres after a set period of time?
@ZackDS In general case one thing that I could recommend you to do to confirm such proposals is join the NP matrix channel and ask NPs directly, to double check that they are indeed okay with the proposal. Or reach out to them elsewhere and get a confirmation. You shouldn’t trust a since source of information - the more sources the better.
However in this case the NPs from these DCs are already not in the ecosystem (de facto) so you will have a tough time reaching out to them. But feel free to prove me wrong
We are just doing this for the first time, so no real procedures and rules. Nodes should generally clean up all sensitive data as soon as they are removed from a subnet - so basically that’s the only thing that really needs to be done – we need to have these nodes unassigned at the time of offboarding. And of course they shouldn’t be added to another subnet.
Regarding what is the “set period of time” - that depends. The initial NPs were onboarded for 4 years. The recent renewals are done for 2 years. We’ll see how it goes in the future. Although I’m not sure if that was what you were asking.
That’s fair, was asking in general for upcoming similar events if it would/should be good to have a heads up post in NP or other forum thread. I am in the NP Matrix so that would also work.
Yes, that’s more or less what I was asking. Did the same timeframe apply to DCs? Is there anything you can point to in earlier proposals, in the forum or somewhere else just to back this up? Or has there been some direct communication from DC owners that they intend to offboard? Just something so that we know the context behind these proposals. Thanks!
I just realised - the Internet Computer has only been running since May 2021 so there shouldn’t yet be any NPs or DCs that have been operating on it for 48 months. Have I missed or misunderstood something?
Yes, a NP is typically closely associated with a DC. There are some NPs that/who share a DC and these NPs are to the best of my knowledge closely in the story together, and make and leave a project (such as the IC) together.
I sadly can’t point to any public communication about this offboarding. There are some NPs with which I had very difficult time communicating (with all my messages likely ending up in /dev/null). With Gen2 NPs I wanted to make communication very transparent and open, and I’m happy how the communication with most of them runs today. But in contrast I never managed to motivate some of these early Gen1 investors to join these public channels. They never cared enough to spend any time on this, or they cared about their privacy too much to do it. But again… feel free to prove me wrong.
Thanks @sat for the further information. I had a hunt around for some further detail but it’s not that easy to find and it looks like some of it might have been lost. Also tagging @cryptoschindler and @katiep who I think were involved at the time. If there’s any documentation about when these data centres were onboarded and the timeframe that they agreed to, or failing that even just some more information about how the agreements were reached, this would help a lot with being able to vote in favour of these proposals.
(Rome wasn’t built in a day), some of which are offboarding now.