Proof of Membership for Blockchains

[alkan]

In a recent article I proposed a dual-token approach to blockchain consensus. The system aims to provide economic decentralization, fulfilling the paradigm: One entity, one vote! The model stands in a stark contrast to existing blockchains like Bitcoin, Ethereum or NXT in which the influence of a party is proportional (or even superlinear) to its resources or stake. My proposal can be broken down into the following key points:

- The blockchain makes use of two different tokens: minter accounts and currency units (coins)
- Blocks can only be built by minter accounts
- The creation rate of minter accounts is limited: With every block, only a predefined number of new minters accounts are created.
- Minter accounts can be sold once upon creation, while further sales are strongly discouraged as the seller will still know the private key (and could steal the funds back any time).
- No economic incentive to own multiple accounts

In my second post I relax some of the assumptions made in my seminal article and further explore the design space around Proof of Membership blockchains. The name “Proof of Membership” is chosen to express that every member of the system can contribute to the blockchain, while a member’s impact on the consensus is not based on stake nor on hash rate. Even though it is impossible to physically restrict the number of accounts held by a member, we can make it economically pointless to possess more than one (or maybe a few) account.

You can find my article here: https://hackernoon.com/proof-of-membership-for-blockchains-1534a3f9faba

Please let me know what you think!

[XbladeX]

***
Please let me know what you think!
***

Dash with insider masternode sytem is best circle jest and proof of membership of all systems.
Dash had best ponzi scheme of all ^^ it will be hard to you win with it.

[ttookk]

(…)

- No economic incentive to own multiple accounts

(…)

Interesting concept. Can you elaborate on the quoted part? How exactly can a single user have neither the ability nor the incentive to pose as multiple users (i.e. having multiple accounts)?

… nevermind, found it:

(…) The question arises if we can design a reward scheme for minting that is incentive-compatible with our one-entity-one-account approach. To solve the dilemma, we can combine both schemes and credit interests to every minter with every block, but lock the accrued interests until the minter actually builds a block himself. This model guarantees incentive-compatibility with regard to compound interests since the latter don’t depend on the distribution of one’s stake anymore. Depending on individual liquidity needs and costs, a minter may still decide to use several accounts for minting in order to get faster access to his interests. But the benefits of using multiple accounts would be rather limited. (…)

So, it still has advantages to have multiple accounts, they are just not as prominent as in other schemes.

[irukandji]

Sounds interesting but it is not clear what a Minter account is.

I think it could be useful if you explained the concept in 5 or 6 sentences, even if you miss important details. Something blunt

[ttookk]

Sounds interesting but it is not clear what a Minter account is.

I think it could be useful if you explained the concept in 5 or 6 sentences, even if you miss important details. Something blunt

Huh? I think the idea is pretty clear:

It basically works like Lisks/Arks DPoS system, with specific accounts, which are allowed to contribute to the blockchain (and mint coins in the process). Different to DPoS, who is eligible for minting is determined by a token which differs from the regular coin: If an account holds such a token, it is allowed to mint.

The specifics about how a token is accquired and what "minting" means exactly seems to be up for discussion, although the author gives some ideas, which are quite interesting. Overall, not a bad concept.

This looks like an idea, which looks interesting especially for semi-private blockchains, blockchain-as-a-service solutions and so on.

[alkan]

So, it still has advantages to have multiple accounts, they are just not as prominent as in other schemes.

Well, that entirely depends on the chosen incentive mechanism. If you opt for the "mine to be released from some duty" model, multiple accounts even become counter-productive. For example, if every member has to perform computional work to stay alive (e.g. by regularly submitting PoW-backed transactions that get included in the blockchain), and by mining a block you get released from that duty, then owning several accounts would result in a multiplication of the required work. In such a setting, it doesn't make any sense to have multiple account since the rewards are calculated based on your total balance, no matter how many accounts you use to store it.

It basically works like Lisks/Arks DPoS system, with specific accounts, which are allowed to contribute to the blockchain (and mint coins in the process). Different to DPoS, who is eligible for minting is determined by a token which differs from the regular coin: If an account holds such a token, it is allowed to mint.

Thanks for the hint. I will take a look at Lisks/Arks DPoS system.

The specifics about how a token is accquired and what "minting" means exactly seems to be up for discussion, although the author gives some ideas, which are quite interesting. Overall, not a bad concept.

Maybe, you can come up with some more ideas/alternative incentive models? Let's have a discussion.  

This looks like an idea, which looks interesting especially for semi-private blockchains, blockchain-as-a-service solutions and so on.

I think, Proof of Membership could be used with both permissioned (closed) and permissionless models as it only limits the number (creation rate) of accounts, whereas everyone (who can afford to make the highest bid) could become a member.

[alkan]

I'm not sure if I understand the way how ARK tries to combat centralization risks.

Here's a quote from their article on Medium:

User Kevin has 20 million tokens. Kevin can nominate his delegate fairly easily. He may even be able to vote in 2 of his own delegates, but rightfully so, he purchased 20 Million tokens. What he can’t do is be guaranteed to nominate more than 8 delegates before his weight is too diluted to maintain his position.

What makes Kevin's weight diluted before he can nominate the next 8 delegates? Can't he just create several accounts and spread his 20 million tokens between them?

I didn't find any clues on that in ARK's whitepaper, just this:

The 51 forging nodes with the highest number of votes are eligible to Forge ARK
blocks. This design eliminates the possibility that any single large ARK holder or
an organization holding large percentages of ARK are able to gain control over
the entire network by voting for all of their nodes into forging positions, thus
effectively taking complete control over that DPoS Blockchain. Votes from ARK
Tokens held by ARK Crew may be used at ARK Crew's discretion.

[ttookk]

Ok, please be aware that i'm just playing around here; thinking loudly. Tis is not to be taken too seriously and not aimed at a specific usecase now, but could evolve in that direction.

What if the token can also be seen as a negative thing? Holding a token would be an obligation, not an advantage. A possible scenario could be like this:

The tokens are auctioned off as described, but it's a negative auction: users pay fees not to get the token.

For every block or number of blocks, a token is sent to the account which paid the smallest fees. Token holders are obligated to forge blocks, otherwise they are penalized with a negative interest: If the account hasn't been online within a certain number of blocks, the token is burned and a percentage of the holdings is substracted from the account. Additionally, they are not allowed to spend coins held in an account with a token. Accounts may lose the token in three ways:

- After a certain number of blocks, the token is burned automatically and the account is penalized by losing a percentage of its holdings.

- The user may burn the token for a fee

- The user forges a block.

Forgers/minters/miners earn the fees, penalties and so on a block holds, but also pass the token to the adress which received the transaction with the smallest fees within the block they forged.

In this scenario, some users might want to get the token to earn fees, while others don't want it, because they don't want to keep their wallet online.

To prevent tokens endoing up in empty accounts, the token is sent to the receiver of a transaction, not the sender. While this may seem like a penalty for something the user didn't do at first glance, the penalty is rather small and can be seen as part of the transaction fee.

Now, an obvious problem may be, that those who want to earn interest spam the network with small transactions without fees to earn as many tokens as possible. There are some remedies against that, though:

- A minimum transaction fee may be set in place.

- Those forging blocks still have an interest to earn fees, so they would probably not choose transactions without fees.

- depending on the system, spamming the network with small transaction may actually be a good thing. If you look at Iotas tangle, It works better the more transactions are made. I'm not too familiar with Iotas structure, but you could imagine a system, in which different kinds of transactions exist. Token holders/forgers may pose as witnesses, similar to Byteballs system, rather than actual forgers, maybe.

Anyway, this are just some thoughts for now, may have more later.

[ttookk]

… damn, didn't think of the most obvious problem:

if token holders want to keep the token, they can just send a transaction with low fees in their block.

randomly assigning the token to a transaction doesn't work either, because a token holder could just increase their odds by having lots of own transactions in the block, so…

…nevermind

[joeydangerous]

lol the proof of ownership reminded me of the time there was a guy posting in the Altcoin section doing proof of dev.

[ttookk]

I'm not sure if I understand the way how ARK tries to combat centralization risks.

Here's a quote from their article on Medium:

User Kevin has 20 million tokens. Kevin can nominate his delegate fairly easily. He may even be able to vote in 2 of his own delegates, but rightfully so, he purchased 20 Million tokens. What he can’t do is be guaranteed to nominate more than 8 delegates before his weight is too diluted to maintain his position.

What makes Kevin's weight diluted before he can nominate the next 8 delegates? Can't he just create several accounts and spread his 20 million tokens between them?

I didn't find any clues on that in ARK's whitepaper, just this:

The 51 forging nodes with the highest number of votes are eligible to Forge ARK
blocks. This design eliminates the possibility that any single large ARK holder or
an organization holding large percentages of ARK are able to gain control over
the entire network by voting for all of their nodes into forging positions, thus
effectively taking complete control over that DPoS Blockchain. Votes from ARK
Tokens held by ARK Crew may be used at ARK Crew's discretion.

I'm not too familiar with Arks specific structure, but 1 Ark represents 1 vote. Thus, it doesn't matter from which account someone voted, the voting weight stays the same. What the Kevin example tries to say, I guess, is, that if Kevin wants to use his Ark to vote his own delegate into a forging position, he shouldn't vote for too many other delegates, because then, his votes are spread over multiple delegates, lowering the overall approval, putting him at risk of losing his spot.

In Lisk, one account can vote for up to 101 delegates and for all of them, the voting weight stays the same. when I have 20 million Lisk, every of the 101 votes counts as 20 million Lisk. The Ark team sees a problem here, because in the Lisk system, big accounts are more likely to be king makers. In the above example, 20 million Lisk would be a little less than 20% of all existing Lisk. At the current level of approval of delegates, such an account could put 99 of 101 delegates in forging positions. Would Poloniex use their cold wallet to vote (currently holding 28 million Lisk), they could decide all 101 delegate spots, unless the rest of Lisk holders could come together to give even more approval to other delegats. The Ark team may have a point there.

You could create a hybrid between proof of membership and DPoS, where token holders can decide to give a token to an account. This account applies for a token, not so much in an auction way, but more in a social way, by somehow convincing those who already hold tokens, that it's a good idea to give them minting privileges.


By the way, I come back to the idea of the minting token either being something negative, to prevent users from hoarding them. For example, what if a newly created address would have a token at inception and only after a certain amount of blocks were forged, the token would be burned? The advantege could be lower fees or something along those lines.

I know, it's far away from proof of membership, just a thought

[alkan]

Ok, please be aware that i'm just playing around here; thinking loudly. Tis is not to be taken too seriously and not aimed at a specific usecase now, but could evolve in that direction.

Thanks for your thoughts!

What if the token can also be seen as a negative thing? Holding a token would be an obligation, not an advantage. A possible scenario could be like this:

The tokens are auctioned off as described, but it's a negative auction: users pay fees not to get the token.

For every block or number of blocks, a token is sent to the account which paid the smallest fees.

I'm not sure if I understand your idea. In Proof of Membership, accounts act as the (second) token. So, how can they be sent to an account and penalize its owner?

(Sorry, I didn't have the time to fully grasp the details of your idea... I may come back later)

[ttookk]

Ok, please be aware that i'm just playing around here; thinking loudly. Tis is not to be taken too seriously and not aimed at a specific usecase now, but could evolve in that direction.

Thanks for your thoughts!

What if the token can also be seen as a negative thing? Holding a token would be an obligation, not an advantage. A possible scenario could be like this:

The tokens are auctioned off as described, but it's a negative auction: users pay fees not to get the token.

For every block or number of blocks, a token is sent to the account which paid the smallest fees.

I'm not sure if I understand your idea. In Proof of Membership, accounts act as the (second) token. So, how can they be sent to an account and penalize its owner?

(Sorry, I didn't have the time to fully grasp the details of your idea... I may come back later)

Yeah the idea wasn't well thought out. Penalizing an account would result in the user setting up a new one. Unless:

A newly created Account automatically has a "penalty" on it, possibly by requiring relatively high transaction fees. To lower the penalty, a user can either mine/stake blocks with the account, where the more blocks are mined, the lower the penalty gets, or pay a (pretty high) one time fee to lower the penalty.

Obviously, this system works best if the block rewards are extremely low, consisting only of penalties and fees.

Penalties should probably "grow back" over time. Additionally, it would make sense to have a system in place, which penalizes miners who only confirm transactions with high transaction fees, since being able to pay low transaction fees would be a reward, that should be protected.

[NewbCoins]

Alkan, do you know of any good proof-of-membership coins in development? Are you still interested in this? I'm interest to discuss this more.

Thanks.