Following up on this @nethermind, you admit that originally you suggested to reduce the bottom 30 validators solely because of profitability concerns but the $1k monthly costs you assumed is very incorrect which invalidates your suggestion, please confirm this and remove/cancel already this original incorrect suggestion.
The latency improvement suggestion was not done by you, but by us, the community. And the latency has already been improved by a lot in the latest upgrade with Optimistic Execution, as you see below.
To further improve latency, validators should be running from Japan/Tokyo. We would need to identify those validators not running from Tokyo (which are NOT the bottom 30) and target those specifically. Removing the bottom 30 validators was incorrectly suggested originally for your incorrect profitability analysis, it has nothing to do with latency. Extrapolating this bottom 30 validators for the latency argument makes no sense, because it is not like the top 30 are running from Japan and the bottom 30 no.
In summary: removing the bottom 30 validators for profitability concerns should be confirmed to be removed/cancelled by Nethermind since this was based on incorrect costs assumptions. Suggesting to remove the bottom 30 for latency improvement should also be cancelled because the validators not running from Japan is NOT equal to the bottom 30 validators. The only suggestion that could make any sense, made by the community, not Nethermind, is to identify the validators not running from Japan and either ask them to migrate to Japan for latency improvement and those not migrating could be removed
Moreover, what it is mentioned above it is very serious, recently both @karpatkey and @stride based their delegations decisions mostly on the assumed ‘top 30’ remaining validators assuming this proposal goes on-chain. This is serious for the following reasons:
-After the delegations from Karpatkey and Stride, the voting power is even more centralized and concentrated in the top 30, meaning even easier than before for the proposal to pass if the top 30 vote yes, which of course they will do to eliminate half of their competition, whether the arguments for the removal are flawed or not
-Even if Karpatkey and Stride don’t vote, now there is an even bigger bias by the top 30 validators to vote yes to remove the bottom 30 as to ‘align’ with what Stride and Karpatkey decided the top 30 to be
-Stride was the only supporter of the validator reduction together with Reverie, it seems a potential conflict of interest that they used their DYDX to further encourage their views by ‘pre-selecting’ a top 30 validator group and assuming and suggesting it is because this proposal could go ‘soon’ on-chain and might be approved
-They have centralized the stake so much that now the top 30 has almost 88% of the DYDX staked. This goes against the mission of both entities which is to delegate to increase the decentralization and security of dYdX, rather than the opposite, to centralize it even more
-They have created an artifical gap between the validator 30th and 31st, now not only they have greatly concentrated and centralized the stake in the top 30, but they also created a gap of around 1.5M DYDX between the validators 30th, and 31st. You can see that above 30th the DYDX increases smoothly, and below 31st also decreases smoothly, but between validators 30th and 31st there is a large artificial 1.5M DYDX gap
Validators have been since before v4 genesis supporting so much, trying to attract delegators. Now an entity comes with some flawed arguments to remove half of the validator set. And other entities with large DYDX allocations supposed to increase the decentralization and security of DYDX collude in a centralized way about who to ‘save’ and who to ‘eliminate’. Actually, this could be a potential attack vector for dYdX, a malicious actor could try in this way to keep centralizing the stake and eliminating validators, and then it would be easy to do a 1/3 attack to halt the dYdX Chain or even just take over the chain with a 2/3 attack. The whole point of dYdX v4 was about decentralization, 60 validators was already not too decentralized compared to most PoS networks, but now it is quite concerning concentrating so much the stake in 30 validators, putting the chain at higher risk and lowering a lot the security because of some ‘flawed proposal being discussed in the forum’. If this goes through, then nothing stops the process to be repeated and the stake centralizing even more and the security of the dYdX getting lower and lower. For the integrity of the dYdX chain and its security in the long term it is urgently needed to clarify all the above and put the situation in order