Topic: Better than POS. Proof of Virtual Work. (Read 343 times)

Wasn't a similar system (albeit with a PoW component) used in Decred? As far as I remember, in this coin you had to buy "minting tickets" or something similar. However, I don't remember how it was implemented exactly - would need to read the whitepaper.

The proposal has also some similarity with "deposit-based PoS" - where you must deposit a certain amount of coins to participate in minting (you could view "deposited coins" as the "mining tokens"). It's also a bit similar to Ardor, which has a currency (Ignis) and a "minting token" (Ardor). There are two differences though: in Ardor there is no punishment for validators that don't find blocks when it's their turn, and you receive the reward in mining tokens. The latter has also led to the problem that the minting token is used currently as the main currency, while Ignis hasn't got much traction.

Overall, however, I think the proposal is interesting.

I think monsterer2's attack vector exists. But if at least a significant fraction of the mining token holders are honest (=include all token buy transactions) then most "mining token buy transactions" should eventually be included in a block. But an attacker with a significant fraction  - let's say 20 or 30% - can imo disrupt the on-chain market, if some actors only try to buy mining tokens for a short period of time; so the attacker could try to buy up all sell orders with attractive prices.

This problem should be present in all kinds of "DEX" on PoS-based chains - only that with mining tokens the attacker can manipulate the validator set.

I think this would lead to the "mining token" being rather cheap (only worth as much as the PoS returns are) or the transaction costs being rather high (to maintain a high value for the mining coin). This is a very interessting thought though, I dont know how you could solve the problem I just stated though.

Agreed. But how to ensure that mining tokens distributed equally in the very beginning? Just give them away to the bunch of people or gradually switch from POW (after all mining tokens + some actual coins have been mined) to virtual POW?
And also that system assumes 2 blockchains (mining tokens and actual coins) or a single one with 2 embedded blockchains in it?

Lots of ways. You could use the same system ETH uses for ERC20 tokens, or a more custom native system.


Sure. But UNLIKE pure POS the total value of Mining tokens may actually be MORE than the total value of the Currency, making it impossible to acquire 51% (or it may be less ).

The price would be market driven, since depending on the fees per block, you could calculate a yield per mining token.

You could give out the mining tokens 1:1 at the initial Genesis distribution along with the coins, then be in a situation where stake holders sell their mining tokens, have more money, but less V-POWer on the network, or vice versa. This cannot happen on pure POS.

I am trying to ascertain whether this is a desirable quality - as having the richest people not necessarily in charge seems like a good thing.

The idea itself sounds intriguing, but I didn't get how is on-chain mining token trading happening? Can someone buyout majority of mining tokens and allocate it all to his own nodes pretending to be different people?

Not coins. Specifically mining tokens - because if you start with ~0 you've got the problem I outlined. If you decide to start the chain with, say a large number like 33% mining tokens allocated, then this is basically exactly the same as plain PoS.

Fair enough.


The $64,000 question.. you need some way of distributing the coins.

-------------------

Seeing all the current goo-goo-love for EOS, made me think this is a Version of DPOS, with analogue delegates - no fixed number or strength, where you can buy and sell your 'delegate powers' at a market rate.


edit: ( .. I'm really not a big fan of DPOS in it's current forms.. )

Perhaps if you'd have described this with more than one paragraph it might have been easier to analyse?



And what do you do at the currency's inception when no one has any mining tokens?

Mate - I'm sorry - but you're still not quite getting it..

.....

I have 33% of the mining tokens.
You have 33% of the tokens.
Anony has 33% of the tokens.

That means we each get to build 33% of the blocks. Like a normal chain with normal mining. No one needs to buy or sell any tokens at this point. You don't lose tokens when you build a block. You can of course do what you want with your blocks, and keep it empty, but you'll just be hurting yourself (hence my link to the empty block fallacy). I'll just add those juicy transactions to my block. Unless you have 51% of the mining tokens, you can't stop me creating an accepted valid block.

There is an on-chain market for mining tokens, and their price will be determined by the fees made by those mining per mining token. The number of mining tokens is fixed.

All we've done is virtualize the POW mining scene - and decoupled your Stake and your Mining prowess. POS issues still remain, absolutely.

What I really want to know is what in pure POS is better than VPOW ? (since I feel VPOW has a couple of real advantages - where are it's disadvantages)
 

Why?

If I have a mining token to build a block, I can exclude all other purchases (as discussed), but my own, so I buy another token in the block I produce using the block reward, which I then use to repeat the process forever.

In fact, it is in everybody's best interests to do this, since it maximises their rewards - they can all build private infinite length chains and fork the coin into infinity for maximum rewards.

You'd need 51% of the mining tokens to pull that off, and you'd be turning down huge amounts of legitimate fee paying transactions. You'd also be destroying your initial investment in aforementioned mining tokens by undermining the chain you mine.. I don't see it.

..
 
https://github.com/libbitcoin/libbitcoin/wiki/Empty-Block-Fallacy

Fair enough... But presumably, you buy them with a transaction? So this will devolve into plain POS - what will happen is that the person who buys a token can produce a block in which no one else bought any tokens, so he is the only one who can produce a block, rinse and repeat into infinity.

No no.. I think you're making it too complicated.

You don't spend mining tokens. They are the same as your stake in POS.

More tokens more chance you find a block. That's it.

I don't think they are. How do you buy a mining token? Presumably via some kind of on chain transaction? How is that transaction finalised? By 'spending' that token and then mining a block?

So, what can happen is that an attacker will just undo the spend of his purchased mining token in a private chain - he can do this forever if he has sufficient tokens, forking the chain into infinity.

No - no real POW component.

Virtual POW. POW generated by the mining tokens you own. More Tokens - more VPOW. Zero energy - like POS.
 
It still has some intrinsic issues that POS has - I agree.

But some POS issues, large issues IMHO, are dealt with in this incarnation.

It isn't clear from your OP that your proposal actually has a PoW component? I assumed not. In which case you'd have the illustrated problem.

Sorry - not quite sure I get you.

Consensus is reached by the chain with the most VM (Virtual Mining).

Are you referring to the Boot-Strap Period, when you are starting the chain ?  Some one needs to own _some_ mining tokens to start the chain ?

You've got the same chicken and egg problem here that PoB has; how can you come to a consensus on the valid set of transactions by sending more transactions?

I did have a thread about this about 2-3 years ago..  can't find it..

You have a Dual Coin chain. One is the currency, and one is the mining tokens.

You buy/sell mining tokens on chain with the currency and those give you the right to mine blocks, more mining tokens, more blocks you find. Block fees are paid in the currency.

The issues with POS that this fixes..

1) All those who own mining token WILL mine (or be penalised if they fail to find a block when it's their turn), so you will have ~100% staking power, rather than the usual 15-20% you get on a POS coin. Small holders of POS don't mine, and as they make up a majority (or should), this leads to low percentage of stake holders mining - thus making attacks cheaper. (I know exchanges have most of the coins anyway, but at some point this shouldn't be the case)

2) Decoupling mining power and stake.. means you don't end up with any entity having an unassailable lead, as a 51% POS holder has. You don't need to be the top stake holder to be the top miner - or vice versa.

You still have nothing at stake.

The lure of low energy consensus is strong, and I do hope that someday someone fixes the NAS issue. (Reverse Stellar Based Check-Point Holes?)