An open discussion about improving the node-runners cost-efficiency, decreasing sell pressure and increasing security by educating community on the importance of non-custodial node-running feature

[RawthiL (Ramiro | POKTscan)]

While I do think that non-custodial staking should be the way to go, I disagree with the reasons you mention for the system to be centralized and also with the proposals you cite to correct this problem.
Non-custodial should be revisited to make it more appealing. It should work better if:

1. Allows trust-less rev-share with the node runner address.
2. Allows using rewards to pay for claims before rev-share.
3. A stake-delegation feature exists to enable pools to exist in a non-custodial scenario.

Now, I don’t really know if this is possible in v0, but even so it would be a consensus breaking change, an we are trying to reach v1 without any stressful changes (this will be one). If anything should be done, I would think it as a V1 feature, which means post Q12024 (at least)

---

@RayBorg

PUP-27 should pass to give people an opportunity to consolidate further. Furthermore, PUP-27 passing would also force 15k servicers that are running their nodes at expensive providers to become even less profitable so all of the inefficient node-runners would be forced to shut down their nodes or switch to more efficient service providers. It would result with significant reduction of market sell pressure, ensuring more POKT goes to investors pockets rather than node-runners that are charging high fees.

This is not correct, as it was discussed in the PUP-27 thread, increasing the stake ceiling to the proposed valued does not lead to network reduced costs. The cost of running a Pocket Node is not the main cost of the network anymore since the mass adoption of LeanPOKT. Also the number of Pocket Nodes is not a problem on sell pressure, only on staked capital, but that’s another story.

@RayBorg

It’s massively important that PUP-31 gets implemented so that investors can switch from inefficient/expensive service providers to most efficient service providers. Introducing more competition among service providers serves as the most efficient tool for lowering the profit margins of service providers.
As of right now, many investors hesitate to change provider because they don’t want to wait for 21 days and lose all of the rewards during that time, remaining at inefficient service providers.

I agree that lower friction in changing providers is desireble, but PUP-31 in not the best conduct to implement that. The correct way to promote this is by implementing TransferStake, we were working on it but we were not able to put the hours needed on it and we are hesitant to do it if the community does not agrees that this work is required and should be rewarded (and it is also a consensus breaking change).

@RayBorg

The importance of non-custodial staking feature should stand out even above profitability itself. …

Most node-running service providers accept non-custodial, however it has some problems:

1. Pools cannot work with it.
2. The owner of the non-custodial node (the client) should be aware that s/he needs to pay for transactions, otherwise the node will not be able to claim (the client needs to work).
3. There is no trust-less way to make reward share, then the owner of the non-custodial node should pay a fixed sum in advance or transfer manually a given percentage of the earnings.

As you say this is a problem of educating the community not of the protocol itself. I think that many clients of node-running operations are comfortable with custodial staking since they don’t need to do anything, they just put some $POKT in and reeve some $POKT without needing to care about anything.

@RayBorg

All the suggestions above including PIP-31 are promoting the decentralization, which would be supported as much as possible. More unique people would hold their own private keys and being less reliant on centralized entities

Maybe I use the term “centralization” differently. The Pocket Network centralization wont change in the best case, it would turn more centralized if you keep pushing for node-running cost efficiency given the current network landscape. The most efficient will be the bigger one, as it once was the case in this network, the only reason for that to change was the creation of LeanPOKT and GeoMesh. This teach us that the only way to change the centralization of the Pocket Network is by changing the play field, not by stressing node-runners.