Reshaping the network to allow for more contributors

Governance
#1

Abstract

The current system doesn’t make a lot of sense. P-reps are being payed based on the number of votes which clearly doesn’t work. Asking people to distrubute votes only goes so far, imagine if the ecosystem was 10x bigger. How centralized wouldn’t that be? Right now people may care about decentralization, but I can assure everyone that as things grow, centralization becomes a bigger issue- not a smaller one. Bitcoin is the prime example that now has 4 pools controlling > 51% of the network.

Identifying the issue

I have tried to think about how things could get better, but as long as the top P-rep teams are getting insane rewards, there is some serious incentive to run a staking service and give the rewards back to equity holders. Things look like they may be shaping for the better but I strongly don’t think so and the sole reason we don’t have more centralization is due to the ecosystem being small, and voting biases being pursused to make things look better than they are.

Voting Bias

Certain whales, unsure who, is selecting P-reps to ensure decentralization. Let’s say for a second that Min Kim is the one behind decentralization. Without people like him we are not even close to having 22 main P-reps, but rather 5 P-reps. This doesn’t scale. We don’t have 100 Min Kims that will vote for decentralization.

Solution

The P-reps are obtaining too much rewards. The voting system has important utilities, but comes with too much responsibility. The solution will be keeping the voting system, however letting P-reps work for their rewards. The more hashrate you produce, the more rewards you will receive. If you are in the top 22 you will receive extra rewards over the Sub-preps to prevent bad actors from having a voice.

To clarify: In order to prevent centralization, you get the same rewards being number 1 as well as being number 22. If you are number 23 you will make less because you’re not a main P-rep.

So instead of getting rewards purely based on your votes, you will have to be in the top 22 AND grinding rewards; the more power you put in the engine, the faster it goes (the more you make).

Sumarize

This solution helps secure the network with more power as long as there is incentive for it. P-reps will only have to run a node and can delegate rewards to EEP’s and DBP’s as their two sole responsibilities. They can submit their own proposals if they want to contribute as well, But Aren’t Getting The Same Rewards As Everyone For Doing Nothing But Running A Node.

After reading this again I felt a need to say this in different words: Instead of competing for votes as the primary drive, you are instead competing for hashrate similar to Bitcoin; However you still need to be in the top 22 in order to receive the rewards in the first place. This is mostly an idea to counter centralization. I don’t think there are many other good ways getting around the fact that P-reps are willing to make multiple nodes leading to centralization.

1 Like
BIG PROBLEM with P-Rep Governance Model! We need to solve this right now!
#2

I hate the idea of DQ’ing P-reps for those “only” securing the network. They should have the ability to just do that, and shouldn’t be obligied to contribute. Contributions should come in the form of EEP’s and DBP’s.

I am up for debating the solution, but I think we all can agree on the issue that we have.

Just to clarify once more the issue incase someone didn’t understand.

If a Staking service can get in the top 22 through giving all the block rewards of let’s say $30k a month ontop of the voting rewards, then those voters gets more than everyone else. This is a serious issue and overtime we will see more and more equity P-reps.

Asking P-reps to not Vote Buy is inefficient. Any P-rep that chooses to contribute will suffer a great loss if he doesn’t get votes. Let’s take Reliant Node as an example. They wrote a nice looking wallet for free and are ranked 40. They lose money by contributing. How is that a good system?

2 Likes
#3

I am willing to bump this post as it appears that more people have realized this being an issue. Oh and here is a lesson about capitalism that I find very helpful when evaluating these things.

If the job of a salesman is to make sales, then he should get compensated for making sales, not by gathering votes from people who are not in the sales business. If he is the former, he will have an incentive to become the best possible salesman, which is what we want, right?

#BringBackDBP

1 Like
#4

Aspected,

Apologies for letting this thread go unnoticed for so long. Going back to your original idea (not sure if you still feel this way), but switching our network to Proof of Work would be too immense of a change and take significant development resources.

As for the rest of your points, I tend to agree. I think a lot of this has been covered in recent discussions. The DBP and EEP were not removed, just rebranded as one item called a Contribution Proposal. Check out the Contribution Proposal System Paper if you have not already. It doesn’t rely on active participation of voters and is a budget based / results based system. I also think you should check out this thread as it relates to many of your points

Besides Proof of Work, I believe most of the high-level changes you were hoping to see are in the works. Thank you for your feedback and dedication.

#5

Hello, Scott.

Proof of Work might give the wrong idea, although that’s essentially what it would be. The distribution of selected entities (Selected with votes) is a great idea, as it ensures scalability without any real sacrifice of decentralization & censorship resistance. My idea was once you’re in the top 22, only then are you competing with computational resources (Proof of Work), or alternatively availability - Which ever has the biggest impact on redundancy is what I would go for. I think that P-reps should be responsible for uptime mainly - Not building/producing stuff (Unless they really want to). You’re probably correct either way though, it’s quite a dramatic change at this point, not only development wise, but also politically.

Anyways,

I’m personally not a fan of indirect funding which is what you linked looks to be. If I understood the paper correctly, it’s based on me asking for funding and hopefully building something neat in the future. I wouldn’t say that I’m against others using it, but I’m also not a fan of using such a system myself, as that aggravates mental pressure to perform, not to mention the ‘begging’ aspect which I prefer not to do.

My ideal philosophy is to create something great, and be incentivized for it afterwards, as that leads to pure innovation better than any other system that I know of. I think the ideal solution is to give let’s say 50% of the transaction fees generated back to the Dapp operators, and also bringing back the Dapp Booster Program term along with it.

#6

In your proposal here to give 50% of transaction fees to Dapps, how do you suppose enforcing what qualifies as a DApp and what does not? How would you stop people, for example, to deploy a smart contract and say it’s a dapp just to get rewards? Curious for more specifics on your proposal and how you would want to get dapps more interested in ICON.

Obviously we would like more DApps on the network. Right now the generally low transaction fees, fee 2.0 (fee sharing and virtual step), and python smart contract language are some of the protocol level advantages to building on ICON.

#7

It wouldn’t be a centralized program similar to the Dapp contest last year, I completely understand the problems with that, the focus is on decentralization. There’s actually multiple implementations of this system. I’ll list two that I really like.

The first, more simple idea is to just allow a 50% or even 75% fee reduction for the first 12 months (Options for more if P-reps agree to it). No need to keep it fancy. I don’t believe it would be the most efficient way of hoarding development, but I can see it making ever so attractive regardless. I believe there is a commitment obstacle from a developer perspective that you might have to pay thousands of ICX if your project ever takes off. It’s still pretty easy to implement and would be maintained by P-reps so spammers didn’t abuse this feature and clogged up the blockchain with useless data.

The second alternative, which is a little bit more interesting: P-reps would vote on Dapps that truly generated activity - How do they decide on that? I’ll explain (Sorry for the incoming rant): Let’s take ICONBET, which is probably the most successful Dapp on ICON right now if you ask me: Why do P-reps not reward them further so that they can afford more developers and what not? If these people are able to create something robust enough to gain users, then shouldn’t they get extra compensated for pulling it off? Instead they have to apply for a P-rep, which by the way gives them less time to develop ICONBET. The big issue having to apply for a P-rep however is that they’re competing with far less successful P-reps who may or may not have done anything. Although they’re doing reasonable being ranked #19, I just find it unsustainable at scale to see it being a reality that quality P-reps are like needles in a haystack.

Anyways, The idea is if you recieve an adequate amount of votes, you get 50% extra fees to fund future development (Up to a certain point a day). P-reps can verify themselves through data analytics on the blockchain to make sure no wash-trade is being done, although I find it unlikely that a successful and prominent DAO like ICONBET would even put their reputation on the line knowing that all transactions are public. I think they would be very keen to actual get compensated, and be happy to further fund development to keep themselves going with good momentum.

#8

So I would say that the Contribution Proposal System is what you describe as the second alternative. For ICONbet, they could apply to the CPS to build a new game, then P-Reps vote on it, and if it passes they will get the funding they requested to build the new game. After that game is complete and if they prove themselves as trustworthy I would think it would be even easier going forward to get funding through the CPS.

As for transaction fee promotions, I would say that fees are already quite low and a 50% reduction would not garner much interest. Alternatively, Virtual Step is built into the protocol already and offers tx fee credits to developers that stake ICX in their SCORE. I know a few ICON DApps are using it currently.

#9

Arguably so, but my description of it suggests that only Dapps that already are successful in generating traffic are eligible for future funding. I see plenty of proposals that have little to do with ICON for instance - not to mention them lacking a track record.

In general I do realize that there is a lack of specified solutions for the DBP. I originally envisioned a program that was meant to be independent of anyone (Such as P-reps) to proceed the funding. The base pitch is still: “Make a Dapp on ICON and get funded for based on how successful it and nothing else” - No need for anything that may cause a headache in terms of being denied funding, publicizing your identity etc…

I do see that the above vision suffers from spam one way or another. I initially thought that the competition last year was designed poorly, but perhaps there is no better way around it. I’ll think about a more specified way of funding regardless because if it is possible, then that’s a significant step towards more developers adoptiong the ICON public chain.