Consensus Mechanisms

[forro]

Just finished watching this video, where The Crypto Show interviews Ryan after the Open House.

I'm much more comfortable with this discussion now. It is critical that there is a very long transition period. POW has a long track record of reliability. It must remain to some degree for a very long time, and we must remain the dominant x11 chain.
The transition period should be very gradual, starting with something like 1 out of 20 blocks going to POS and gradually increasing, incrementing to 2 out of 20 blocks, then 3 out of 20 blocks, on the order of every ~30 days, and so on as conditions permit. The stakers share of the mining reward would be directly proportional to that. I know that sounds extremely gradual, but it must prove itself.

I understand that the POS algorithm could be BLS based. The algorithm used for POS block production must be sufficiently different from that used by chainlocks in case a weakness in our BLS implementation is discovered. For that reason, I think it benefits the network if we stop at a 1:1 block alternation between POS and POW, with an even 22.5% split between them. I'm not a developer, but it may be a good idea to ensure that whatever BLS libraries are used for chainlocks are NOT reused for POS block production, in case a vulnerability exists, but that they are rewritten.

Will stakers need to be reliable? Will they need to be always on and subject to a proof of service score? Will they need to store a full copy of the blockchain? Will people really be able to stake from home individually, or will they end up giving their dash to POS pools, and eventually we end up with one or two large POS pools? How do we prevent POS pools from conglomerating? We could be creating a new world where the dash of our POS stakers is subject to a POS pool exit scam. Is that where a distributed staker list comes in? Same as a masternode where an address of exactly 100 dash, no more, no less, is registered on the network as collateral, and after a full payment cycle they are getting their rewards?

I agree with some here where it would be nice to have a low staking limit so as to distributed rewards to as many people as possible, but it would likely become unwieldy for the network to handle. Also, those with little dash to stake could be a less reliable, needing to withdraw it for one reason or another soon after. I imagine the lower we set the staking limit, the more volatile the population of stakers would be. The higher we set it, towards 100, we'd be tapping a set of people who would be more likely able to keep it there for a long period of time, reducing staker volatiliy, put another way, increasing staker stability, and reducing the load on the network.

I would appreciate some responses to these pretty fundamental questions on how this is really supposed to work and evolve.

Lastly, I myself do not consider collateralized mining a dead issue. I would like to hear someone explain to me why it should be considered dead. I think adding that to our arsenal could only further fortify the network.

 

[qwizzie]

What this needs to be viable is a user-friendly integration into the wallet, so people can stake inside a mobile wallet for example.

Something like this ? https://app.dashnexus.org/proposals...ndroid-wallet-powered-by-smart-valor/overview

Smart Valor + DashUp

• Smart Valor is the premier partner, publishing the Dash Staking app on the Google Play store.
• DashUp is developing the open source Masternode service marketplace Android code as developers.
• DCG will have the opportunity to review and merge into the core wallet if desired.

 

[toknormal]

Part 1...the issue with market share

Could I just say, I think there is a topic missing in the thread list of the "economics" group and that is "Market Share". So please consider this post as being in the that category.

We spend a lot of time examining Dash's own market when IMO we should be drawing our attention to the pressure gradients that pull market share away from us. Again in my opinion, one of these pressure gradients is not that we are paying too much for hashrate. In particular I would contribute the following points:

1. it would be instructive to analyse these markets systematically and avoid ambiguous pronouns such as "we" to cover a range of disparate market participants, all with distinct commercial roles, priorities and business models. These include miners, masternode operators, merchants, market buyers who may be in the market for proof of work, market buyers who may be in the market for proof of stake etc.
2. in that regard, we need to distinctly consider Dash's and its competitor's Primary Markets from their Secondary Markets (as Ryan has done to a limited extent by targeting mining supply for attention). This is important and a common way in which monetary asset markets are appraised. (See here for securities and here for monetary commodities and more background here)

Primary Market

To clarify, this is the market that buys the new supply from miners. The secondary market buys the pre-existing supply (coins that have already been supplied to holders). Clearly these are not distinct markets in terms of exchanges but the distinction is made in casual discussions, for example in Ryan's analyses where he delineates this market segment in terms of reward distribution.

Specifically, the definitions I'm using here are:

• if a miner places a sell order in an order book then that would represent the "primary market" supply
• if a holder of a coin which was bought or otherwise acquired without cost, places a sell order in an order book then that represents the "secondary market" supply

With this in mind, the first "market share" issue I'd like to highlight is one which is endemic in Dash and that is the effect of the reward split in creating a pressure gradient for market share away from us.

What does the phrase "We pay too much for hashrate" really mean ? (And how is this resolved)

I'd like to re-articulate Ryan's interpretation of "we are paying too much for hashrate as"...

• the primary market pays too much for hashrate compared with our competitors
• the reason is that the Dash protocol pays too much for masternode hosting costs and passes these costs on to the primary market buyer

Both are related and the problem is very definitely not resolved by restricting mining reward even further IMO. Rather it would be aggravated by such measure. The basis for these observations are respectively:

• as demonstrated below, the Dash's primary market has to pay the mining cost of 2 coins to receive 1. Moving to a 10:1 ratio sees it having to cover 10 coin's mining cost to receive 1
• the masternode network currently costs around $75k to host, yet draws $700k from the network revenue (coins sold into the secondary market). Dash's secondary market does not want to bear that cost and therefore moves to other coins

The Mechanics of our Loss of Market Share

This is demonstrated in the 2 illustrations below which compare the mining & market interactions of 2 identical coins - one with a split reward such as Dash's and one with 100% mining reward.

Aside: I recommend studying the origin and fortunes of Peercoin which had a hybrid POW / POS protocol​

Example 1: Split Reward Model

Example 2: Unsplit Reward Model

 

[toknormal]

Part 2....How to Optimally Address This Issue

While I see this as one of Dash's problems (in terms of competitiveness) it's not a show-stopper and can be fixed. The fix is to get rid of the supernormal profit in the masternode sector before the market does it for us. (...is doing for us). There are 2 ways to do that:

1. increase the productive service provision that the masternode network makes (and consequently its running costs so that the redundant revenue the network makes is deployed to the network's benefit instead of wasted)
2. reduce the masternode reward ratio (increase the mining reward ratio) until such time as [1] is possible (i.e. there is demand for services provided by the MN network which can absorb their revenue)

If we do not do either 1, or 2, my prediction (and current observation) is that the market will bring the masternode margin and mining margins into parity. That means a coin price where masternode revenues are moderately above hosting costs.

Clearly switching reward ratio away from nodes and back to miners is not a popular option amongst (some) operators. But they've already made enormous unrealised losses in $USD terms so they face the choice of 2 evils. I have promoted here what I think is the "healthier" one all round.

Other Observations

1. our market share loss has gone hand-in-hand with a growing masternode population, this would seem to re-inforce the basis that the shared reward model is impacting it adversly as described
2. publicly posted anecdotal observations also provide evidence of this explanation over the one that mining supply is the issue

Hashrate and/or POS Considerations

Firstly, in my opinion, Dash's buyers are in the market for a Proof of Work imbued commodity asset. They are not in the market for POS-based monetary transfer services on a blockchain since Ripple practically dominates that market.

They are also not in the market for an efficiently secured network (as demonstrated by the fact that chainlocks, while a useful innovation, did nothing to add to our valuation). They take for granted that their blockchain holdings are secure in most of the "top coins".

Aside. I made a crude, initial measure of the amount of "Cumulative Proof of Work" (CPOW) in Dash's blockchain by taking the difficulty for each day (which is a non-dimensional property and therefore notionally comparable between chains), multiplying it by each day's coin inflation and summing the results for the duration of Dash's lifetime. Then taking the current market cap and dividing it by the CPOW, comparing the result for Bitcoin, Litecoin and Dash. By this measure Dash gets many times more marketcap for each unit of CPOW than Bitcoin. So it would seem (by this crude measure at least) that we are not paying "too much for hashrate". This may be a flawed measure (for example it should take into account the effect of ASICS and changes in mining efficiency) but it would be an illuminating area of research

There is otherwise very little liquidity available for pure monetary, POS digital assets according to the current market state. The ones that spring to mind are Peercoin (hybrid POW/POS), Cinnicoin, Blackcoin et al. Dash's on-chain features only make sense when complimented with a thoroughly proof of work algo, not even a hybrid.

Secondly, whether the assertion that "we are paying too much for hashrate" is true or not, we already have only half the mining reward ratio our competitors do anyway, yet we still lose market share to them. So it's hard to see a credible basis for further diminishing it.

 

[toknormal]

DASH needs an honest, ethical stable coin.

You can't be a commodity-based store of value and a stablecoin at the same time (unless you're Bitshares). The two are in conflict. Nobody is going to make any gains by investing in stable coins because (by definition), to work, a stable coin needs to increase its supply when there is demand for liquidity so that its own value doesn't increase.

If you want a blockchain where changes in the stablecoin supply are transmitted to the associated blockchain collateralising token as a revaluation then you need to invest in either Bitshares or one of the platforms that supports the creation of meta tokens backed by non-native blockchain collateral.

Otherwise feel free to post the detailed mechanics by which you see such a "stable-coin" working in Dash so that they can be appraised accordingly. I've never seen any advocate of a Dash stablecoin do that yet.

 

[bhkien]

Part 2....How to Optimally Address This Issue

While I see this as one of Dash's problems (in terms of competitiveness) it's not a show-stopper and can be fixed. The fix is to get rid of the supernormal profit in the masternode sector before the market does it for us. (...is doing for us). There are 2 ways to do that:

1. increase the productive service provision that the masternode network makes (and consequently its running costs so that the redundant revenue the network makes is deployed to the network's benefit instead of wasted)
2. reduce the masternode reward ratio (increase the mining reward ratio) until such time as [1] is possible (i.e. there is demand for services provided by the MN network which can absorb their revenue)

If we do not do either 1, or 2, my prediction (and current observation) is that the market will bring the masternode margin and mining margins into parity. That means a coin price where masternode revenues are moderately above hosting costs.

Clearly switching reward ratio away from nodes and back to miners is not a popular option amongst (some) operators. But they've already made enormous unrealised losses in $USD terms so they face the choice of 2 evils. I have promoted here what I think is the "healthier" one all round.

Other Observations

1. our market share loss has gone hand-in-hand with a growing masternode population, this would seem to re-inforce the basis that the shared reward model is impacting it adversly as described
2. publicly posted anecdotal observations also provide evidence of this explanation over the one that mining supply is the issue

Hashrate and/or POS Considerations

Firstly, in my opinion, Dash's buyers are in the market for a Proof of Work imbued commodity asset. They are not in the market for POS-based monetary transfer services on a blockchain since Ripple practically dominates that market.

They are also not in the market for an efficiently secured network (as demonstrated by the fact that chainlocks, while a useful innovation, did nothing to add to our valuation). They take for granted that their blockchain holdings are secure in most of the "top coins".

Aside. I made a crude, initial measure of the amount of "Cumulative Proof of Work" (CPOW) in Dash's blockchain by taking the difficulty for each day (which is a non-dimensional property and therefore notionally comparable between chains), multiplying it by each day's coin inflation and summing the results for the duration of Dash's lifetime. Then taking the current market cap and dividing it by the CPOW, comparing the result for Bitcoin, Litecoin and Dash. By this measure Dash gets many times more marketcap for each unit of CPOW than Bitcoin. So it would seem (by this crude measure at least) that we are not paying "too much for hashrate". This may be a flawed measure (for example it should take into account the effect of ASICS and changes in mining efficiency) but it would be an illuminating area of research

There is otherwise very little liquidity available for pure monetary, POS digital assets according to the current market state. The ones that spring to mind are Peercoin (hybrid POW/POS), Cinnicoin, Blackcoin et al. Dash's on-chain features only make sense when complimented with a thoroughly proof of work algo, not even a hybrid.

Secondly, whether the assertion that "we are paying too much for hashrate" is true or not, we already have only half the mining reward ratio our competitors do anyway, yet we still lose market share to them. So it's hard to see a credible basis for further diminishing it.

What do you think if we reduce coin emission. For example, instead of each year emission rate reduces 7.5%, we change to let each year the emission rate will reduce 10% or 14%.

 

[toknormal]

What do you think if we reduce coin emission. For example, instead of each year emission rate reduces 7.5%, we change to let each year the emission rate will reduce 10% or 14%.

I think that if the coin emission reduces according to a scheduled reduction in the original protocol, then that can lead to a revaluation upwards. (Because the market interprets that as a healthy coin which has reached an advanced stage in its growth successfully and, all else being equal, will anticipate a reduction in supply for a consistent demand)

If, on the other hand , the coin emission reduces according to an unscheduled reduction (i.e. a revision to the original protocol emission schedule) then that can lead to a revaluation downwards because the market could interpret that the coin thinks it's in trouble and is not otherwise able to generate demand.

The real problem with Dash at the moment isn't the protocol inflation, it's the disparity between masternode and mining margin. This leads to an unstable core business model in terms of the liquidity flows between market and mining costs. Having to charge the market twice the mining cost means that either hashrate can be stable or coin price can be, but not both at the same time (as with our competitors). Further, whether the mining cost is market price driven or whether market price is mining cost driven, the both lead to a reduction in coin price where the equilibrium is only reached at the point when mining margin and masternode margin are at parity.

If masternode hosting cost is $20 per month and reward ratio is 50/50, then you can work out the equilibrium coin price from that. To protect ourselves from this problem we need to restore the reward ratio back to miners by a fairly large measure (because masternode expenses are so low).

IMO, changing to Proof of Stake will just kill the coin completely since, for its money, the primary market then would receive no Proof of Work and only $20 a month worth of services per masternode. Market share would instead migrate to 2 places:

1. POS service chains with a far higher activity level than Dash such as one of top ranking ones here: https://blocktivity.info
2. POW monetary assets where the primary market gets near 100% return on its money-for-mined-coin rather than the 50% it currently gets with Dash

 

[Ryan Taylor]

I have been reviewing the comments here and in discord, as well as working with our developers to assess various options for the consensus mechanism. There are clearly some very good reasons that both pure PoS and hybrid PoW / PoS would be problematic. There are very good reasons why Ethereum is struggling to deploy PoS. My conclusion is that if we chose that route, the project would run the risk of being a monumental effort... at least if you want a highly secure network. That said, there may be a solution leveraging BLS "magic" that could enable a safe implementation of PoS at some point in the future, but I believe there is much we could do now to improve economics even if we stay with PoW for the time being.

Therefore, if we assume staying with PoW is the most direct path to a better economic profile, it begs the question "how should our PoW allocation be adjusted in light of the enhanced security from ChainLocks?". If you enhance security through ChainLocks, it logically follows that you are less dependent on PoW than you were before. What is clear is that we would want to remain the dominant X11 PoW chain. Therefore, we calculated the hashrate of the other X11 chains. There wasn't much out there... Happycoin, Axe, Imagecoin, Magnet, Cannabiscoin, MonetaryUnit, Startcoin, and PACcoin were the biggest. There are others, but collectively, they are a rounding error. I then multiplied the sum of these other chains by 10x. To buy that much hashrate, we would only need about 5-7% of our block reward. Therefore, even if you rounded up to 10%, there is very little incremental security from spending more than 10% of our block reward on mining.

This is great news, because it means we have the ability to safely reduce mining from 45% of our block reward to 10% (which represents about 80% reduction in mining expense), and that is the vast majority of the gains we could be making with pure PoS. And we could capture that efficiency gain without a lengthy development process.

I believe the best approach to distributing the gains is to ensure as wide of participation as possible. This means one of two paths... either a deterministic holders list, or masternode shares. There are pros and cons with each approach.

DHL:
+ Likely simpler to implement... similar process to DML
- Because DHL participants wouldn't really be doing "work" for the network, this would make Dash more closely resemble an investment... I believe that we would nonetheless have a compelling case why Dash is not a security, but it might nudge Dash in a direction that regulators would view less favorably. Not a reason not to do it, but it is a consideration.
- Not that scalable. We might need a minimum of 100 Dash to prevent the list from becoming too long for nodes to maintain. Or we would need to rearchitect the way the DML list works to make it more scalable. Either way, this is less than ideal.

Protocol level shared MNs:
+ Not difficult to implement, but there are a few minor complexities we could work through quickly (e.g., the engineers have practical solutions in mind)
+ Can scale very well. Completely feasible to stake 10 Dash or less, which mean the benefits are extended to the most number of users.
+ Provides incremental decentralization and security to the network... because it provides benefit to the network, this would make Dash no more susceptible to being labeled a security than it is today.
+ Ensures that shared MN services are trustless, which eliminates the need to take risks that people currently take to earn rewards on smaller balances
- Might require time-locked inputs to properly function, which may be less attractive to users
- Doesn't fully solve the fundamental issues with the network economics, but it can defer those issues to a later time, when they are less impactful (e.g., several years from now, the inflation rate would be much much lower)

I believe by combining reduced mining expenses with greater access to the MN rewards (and increasing the share of block rewards that go to this larger group of recipients) we can successfully mitigate MOST of the economic issue with a relatively simple solution. My current hypothesis is the following (assuming no change to the proposal funding share, which I view as a fully independent discussion):
1) Introduce shared MNs to enable greater access to staking rewards
2) SLOWLY increase the share of block rewards from 45/45/10 to 80/10/10 over a very long period of time. Our previous reallocation was far too fast. If we reallocated about 0.5% per month, it would take nearly 6 years to fully transition.

One other benefit of this approach is "fairness" to miners, as they would retain the ability to mine dash with the full life of their existing equipment. In fact, I need to model it, but I believe even they would be financially better off. How is this possible if we are cutting the amount of Dash allocated there? First, they would be mining slightly less Dash the next 18 months, but at an otherwise higher price, which would help. They also would face less competition from new mining capacity because miners would be unlikely to invest in new X11 equipment knowing reward reductions were baked into the protocol. Lower competition would in turn extend the life of existing equipment and allow miners the opportunity to mine for longer without replacing their miners. Most of a miner's profit takes place in the first 12-18 months under normal circumstances, so this could create a unusually long runway of profitable mining if new equipment is not brought to market at the same pace it otherwise would.

Another benefit... low risk of implementation. The only complex piece is masternode shares, which wouldn't be overly complex as it would simply require a different DML registration type, a way for users to signal their inputs are available to stake, and the ability to distribute the rewards differently as part of the coinbase transaction. All of that is based on technology we already have, and would simply need to modify.

Would like to hear reactions to this model from the community.

 

[Ryan Taylor]

Part 2....How to Optimally Address This Issue

While I see this as one of Dash's problems (in terms of competitiveness) it's not a show-stopper and can be fixed. The fix is to get rid of the supernormal profit in the masternode sector before the market does it for us. (...is doing for us). There are 2 ways to do that:

1. increase the productive service provision that the masternode network makes (and consequently its running costs so that the redundant revenue the network makes is deployed to the network's benefit instead of wasted)
2. reduce the masternode reward ratio (increase the mining reward ratio) until such time as [1] is possible (i.e. there is demand for services provided by the MN network which can absorb their revenue)

If we do not do either 1, or 2, my prediction (and current observation) is that the market will bring the masternode margin and mining margins into parity. That means a coin price where masternode revenues are moderately above hosting costs.

Clearly switching reward ratio away from nodes and back to miners is not a popular option amongst (some) operators. But they've already made enormous unrealised losses in $USD terms so they face the choice of 2 evils. I have promoted here what I think is the "healthier" one all round.

Other Observations

1. our market share loss has gone hand-in-hand with a growing masternode population, this would seem to re-inforce the basis that the shared reward model is impacting it adversly as described
2. publicly posted anecdotal observations also provide evidence of this explanation over the one that mining supply is the issue

Hashrate and/or POS Considerations

Firstly, in my opinion, Dash's buyers are in the market for a Proof of Work imbued commodity asset. They are not in the market for POS-based monetary transfer services on a blockchain since Ripple practically dominates that market.

They are also not in the market for an efficiently secured network (as demonstrated by the fact that chainlocks, while a useful innovation, did nothing to add to our valuation). They take for granted that their blockchain holdings are secure in most of the "top coins".

Aside. I made a crude, initial measure of the amount of "Cumulative Proof of Work" (CPOW) in Dash's blockchain by taking the difficulty for each day (which is a non-dimensional property and therefore notionally comparable between chains), multiplying it by each day's coin inflation and summing the results for the duration of Dash's lifetime. Then taking the current market cap and dividing it by the CPOW, comparing the result for Bitcoin, Litecoin and Dash. By this measure Dash gets many times more marketcap for each unit of CPOW than Bitcoin. So it would seem (by this crude measure at least) that we are not paying "too much for hashrate". This may be a flawed measure (for example it should take into account the effect of ASICS and changes in mining efficiency) but it would be an illuminating area of research

There is otherwise very little liquidity available for pure monetary, POS digital assets according to the current market state. The ones that spring to mind are Peercoin (hybrid POW/POS), Cinnicoin, Blackcoin et al. Dash's on-chain features only make sense when complimented with a thoroughly proof of work algo, not even a hybrid.

Secondly, whether the assertion that "we are paying too much for hashrate" is true or not, we already have only half the mining reward ratio our competitors do anyway, yet we still lose market share to them. So it's hard to see a credible basis for further diminishing it.

Masternode costs have nothing to do with the value they provide. Costs are not an input to value. The fact that they have high profit margins provides them with the option to not have to liquidate most of their earnings if they don't desire to.

Increasing the masternode's cost of operating does not result in the services they provide be worth any more or less, in the same way that if I'm manufacturing shoes, and I buy laces from a more expensive distributor, the value of the shoes does not increase.

 

[forro]

I have been reviewing the comments here and in discord, as well as working with our developers to assess various options for the consensus mechanism. There are clearly some very good reasons that both pure PoS and hybrid PoW / PoS would be problematic. There are very good reasons why Ethereum is struggling to deploy PoS. My conclusion is that if we chose that route, the project would run the risk of being a monumental effort... at least if you want a highly secure network. That said, there may be a solution leveraging BLS "magic" that could enable a safe implementation of PoS at some point in the future, but I believe there is much we could do now to improve economics even if we stay with PoW for the time being.

Therefore, if we assume staying with PoW is the most direct path to a better economic profile, it begs the question "how should our PoW allocation be adjusted in light of the enhanced security from ChainLocks?". If you enhance security through ChainLocks, it logically follows that you are less dependent on PoW than you were before. What is clear is that we would want to remain the dominant X11 PoW chain. Therefore, we calculated the hashrate of the other X11 chains. There wasn't much out there... Happycoin, Axe, Imagecoin, Magnet, Cannabiscoin, MonetaryUnit, Startcoin, and PACcoin were the biggest. There are others, but collectively, they are a rounding error. I then multiplied the sum of these other chains by 10x. To buy that much hashrate, we would only need about 5-7% of our block reward. Therefore, even if you rounded up to 10%, there is very little incremental security from spending more than 10% of our block reward on mining.

This is great news, because it means we have the ability to safely reduce mining from 45% of our block reward to 10% (which represents about 80% reduction in mining expense), and that is the vast majority of the gains we could be making with pure PoS. And we could capture that efficiency gain without a lengthy development process.

I believe the best approach to distributing the gains is to ensure as wide of participation as possible. This means one of two paths... either a deterministic holders list, or masternode shares. There are pros and cons with each approach.

DHL:
+ Likely simpler to implement... similar process to DML
- Because DHL participants wouldn't really be doing "work" for the network, this would make Dash more closely resemble an investment... I believe that we would nonetheless have a compelling case why Dash is not a security, but it might nudge Dash in a direction that regulators would view less favorably. Not a reason not to do it, but it is a consideration.
- Not that scalable. We might need a minimum of 100 Dash to prevent the list from becoming too long for nodes to maintain. Or we would need to rearchitect the way the DML list works to make it more scalable. Either way, this is less than ideal.

Protocol level shared MNs:
+ Not difficult to implement, but there are a few minor complexities we could work through quickly (e.g., the engineers have practical solutions in mind)
+ Can scale very well. Completely feasible to stake 10 Dash or less, which mean the benefits are extended to the most number of users.
+ Provides incremental decentralization and security to the network... because it provides benefit to the network, this would make Dash no more susceptible to being labeled a security than it is today.
+ Ensures that shared MN services are trustless, which eliminates the need to take risks that people currently take to earn rewards on smaller balances
- Might require time-locked inputs to properly function, which may be less attractive to users
- Doesn't fully solve the fundamental issues with the network economics, but it can defer those issues to a later time, when they are less impactful (e.g., several years from now, the inflation rate would be much much lower)

I believe by combining reduced mining expenses with greater access to the MN rewards (and increasing the share of block rewards that go to this larger group of recipients) we can successfully mitigate MOST of the economic issue with a relatively simple solution. My current hypothesis is the following (assuming no change to the proposal funding share, which I view as a fully independent discussion):
1) Introduce shared MNs to enable greater access to staking rewards
2) SLOWLY increase the share of block rewards from 45/45/10 to 80/10/10 over a very long period of time. Our previous reallocation was far too fast. If we reallocated about 0.5% per month, it would take nearly 6 years to fully transition.

One other benefit of this approach is "fairness" to miners, as they would retain the ability to mine dash with the full life of their existing equipment. In fact, I need to model it, but I believe even they would be financially better off. How is this possible if we are cutting the amount of Dash allocated there? First, they would be mining slightly less Dash the next 18 months, but at an otherwise higher price, which would help. They also would face less competition from new mining capacity because miners would be unlikely to invest in new X11 equipment knowing reward reductions were baked into the protocol. Lower competition would in turn extend the life of existing equipment and allow miners the opportunity to mine for longer without replacing their miners. Most of a miner's profit takes place in the first 12-18 months under normal circumstances, so this could create a unusually long runway of profitable mining if new equipment is not brought to market at the same pace it otherwise would.

Another benefit... low risk of implementation. The only complex piece is masternode shares, which wouldn't be overly complex as it would simply require a different DML registration type, a way for users to signal their inputs are available to stake, and the ability to distribute the rewards differently as part of the coinbase transaction. All of that is based on technology we already have, and would simply need to modify.

Would like to hear reactions to this model from the community.

This seems much more sensible, especially the long transition period. It also fulfills one of the long awaited promises.

Would MN shares have voting rights?
Would it be better to start off with a higher entry level like 200-500 dash at first, to see how it effects the MN economy, before opening it up to lower amounts? I suspect the lower the amount, the more MN volatility we can expect.

To me this is not off topic, so I'll ask again here, could you explain the likely dynamics of collateralized mining? Would it lead to more centralization or less, or have no effect?

Collateralized miners would have to be insane to risk their skin in the game, centralized or not, chainlocks or no. It may even help decentralization.

 

[toknormal]

Masternode costs have nothing to do with the value they provide.

True. But they do have a great deal to do with the value that miners provide because they devalue Dash's hashrate to the market by 50%. Note that:

Investors in Dash have invested in:

• a high energy budget coin
• with on-chain services that are competitive with other high energy budget coins

Not...

• a low energy budget coin
• with services that are uncompetitive against other low energy budget coins

...which would be the case if Dash moves even partially to a POS model.

@BabyGiraffe, I urge you to revisit the appraisals in previous posts from that perspective. While you may not value Dash's high energy budget, many of its core investors do. Its primary market also does (otherwise Dash miners would not find any liquidty). In that regard, the "value" that masternodes provide doesn't matter. The margin does because that margin has to be paid for by the primary market buying newly mined coins.

Given those constraints, the only solution I see that's sustainable is one of:

1. find services (other than holiday cruises, and salaries) that absorb the masternode margin to the extent that it makes the coin attractive again, or
2. reduce the masternode margin to bring it close to parity with mining margin to allow time for [1] to develop or
3. have the market do it for us (which it can do by impacting masternode margin disproportionately compared to mining margin by reducing coin price) or
4. switch to POS and watch the last of us disinvest

4 would clearly be my least preferred option

 

[Ryan Taylor]

This seems much more sensible, especially the long transition period. It also fulfills one of the long awaited promises.

Would MN shares have voting rights?
Would it be better to start off with a higher entry level like 200-500 dash at first, to see how it effects the MN economy, before opening it up to lower amounts? I suspect the lower the amount, the more MN volatility we can expect.

To me this is not off topic, so I'll ask again here, could you explain the likely dynamics of collateralized mining? Would it lead to more centralization or less, or have no effect?

Collateralized miners would have to be insane to risk their skin in the game, centralized or not, chainlocks or no. It may even help decentralization.

At least as far as delivering an initial version, I would not anticipate including voting rights. I would be reluctant to include voting rights AND continue requiring 10% net votes for proposals to pass, because I think it's unlikely that a population of 10 Dash collateral holders would vote with the same frequency as those with 1,000 Dash. It runs the risk of stagnating the voting system and cutting off project funding. Perhaps there is a solution to this risk, but it would require more time to solve.

I actually don't think MN volatility would result... if anything I believe it would be lower for several reasons. First, I presume you would have more MNs, so a whale shutting down their MNs would have less impact on a % basis. Also, if shared nodes required locked inputs they couldn't be shut off at will by the individual collateral providers.

Collateralized mining basically reduces the hashrate / expense of mining by requiring miners to hold a certain amount of Dash to compete for blocks. There are several ways to implement it, so I'll describe one scenario to give you an idea. In normal mining, miners need to provide equipment, labor, and electricity. Collateralized mining requires holding Dash so that holding Dash either affects the "difficulty" of the hash (and therefore affects the odds of finding a block), or holding a certain amount of Dash enables a certain number of blocks to be found (e.g., low levels of collateral could limit the number of blocks you are permitted to find out of the last 100 blocks for example). The added "cost" of the collateral prevents any one miner from easily controlling the mining layer just as MN collateral prevents any one person from controlling the MN layer of the network. The result is 1) Increased demand for Dash (for the collateral), 2) better alignment of incentives for miners (because they are Dash holders), 3) more capital is directed toward holding Dash, which leaves less capital for hashrate. Basically, you get the same "cost" to attack the network with less expense of equipment and electricity with collateralized mining. That is the idea behind it.

 

[Ryan Taylor]

True. But they do have a great deal to do with the value that miners provide because they devalue Dash's hashrate to the market by 50%. Note that:

Investors in Dash have invested in:

• a high energy budget coin
• with on-chain services that are competitive with other high energy budget coins

Not...

• a low energy budget coin
• with services that are uncompetitive against other low energy budget coins

...which would be the case if Dash moves even partially to a POS model.

@BabyGiraffe, I urge you to revisit the appraisals in previous posts from that perspective. While you may not value Dash's high energy budget, many of its core investors do. Its primary market also does (otherwise Dash miners would not find any liquidty). In that regard, the "value" that masternodes provide doesn't matter. The margin does because that margin has to be paid for by the primary market buying newly mined coins.

Given those constraints, the only solution I see that's sustainable is one of:

1. find services (other than holiday cruises, and salaries) that absorb the masternode margin to the extent that it makes the coin attractive again, or
2. reduce the masternode margin to bring it close to parity with mining margin to allow time for [1] to develop or
3. have the market do it for us (which it can do by impacting masternode margin disproportionately compared to mining margin by reducing coin price) or
4. switch to POS and watch the last of us disinvest

4 would clearly be my least preferred option

If they happened to base an investment decision on that basis, that is simply a purchase made on the basis of poor rationale. Dash does not accrue value because of the costs it incurs. It is valuable despite the costs it incurs. The costs are necessary to provide security (which is necessary to be valued). But, if the network can be equally secure for less cost, that is a very good thing. I think we can keep having the same debate here, in Discord, and elsewhere. I simply do not agree with you that costs are in any way, nor in any free market, an input into value except to the extent that it provides security (a necessary component no doubt). Unfortunately, the relationship between cost and the value of security you get is not linear, so we would be foolish to spend more than is needed to secure the network as hashrate beyond what we need is waste. ChainLocks provide immutable security within 4-6 seconds using clever code (delivered by one-time coding expense and some small overhead for masternode bandwidth and CPU), rather than expensive hashrate. It is a real example proof that we can deliver value with lower expense.

I'm sorry, but I don't agree with your general philosophy that costs must increase to add to the value of Dash. Quite the opposite.

 

[JGCMiner]

I have been reviewing the comments here and in discord, as well as working with our developers to assess various options for the consensus mechanism. There are clearly some very good reasons that both pure PoS and hybrid PoW / PoS would be problematic. There are very good reasons why Ethereum is struggling to deploy PoS. My conclusion is that if we chose that route, the project would run the risk of being a monumental effort... at least if you want a highly secure network. That said, there may be a solution leveraging BLS "magic" that could enable a safe implementation of PoS at some point in the future, but I believe there is much we could do now to improve economics even if we stay with PoW for the time being.

Therefore, if we assume staying with PoW is the most direct path to a better economic profile, it begs the question "how should our PoW allocation be adjusted in light of the enhanced security from ChainLocks?". If you enhance security through ChainLocks, it logically follows that you are less dependent on PoW than you were before. What is clear is that we would want to remain the dominant X11 PoW chain. Therefore, we calculated the hashrate of the other X11 chains. There wasn't much out there... Happycoin, Axe, Imagecoin, Magnet, Cannabiscoin, MonetaryUnit, Startcoin, and PACcoin were the biggest. There are others, but collectively, they are a rounding error. I then multiplied the sum of these other chains by 10x. To buy that much hashrate, we would only need about 5-7% of our block reward. Therefore, even if you rounded up to 10%, there is very little incremental security from spending more than 10% of our block reward on mining.

This is great news, because it means we have the ability to safely reduce mining from 45% of our block reward to 10% (which represents about 80% reduction in mining expense), and that is the vast majority of the gains we could be making with pure PoS. And we could capture that efficiency gain without a lengthy development process.

I believe the best approach to distributing the gains is to ensure as wide of participation as possible. This means one of two paths... either a deterministic holders list, or masternode shares. There are pros and cons with each approach.

DHL:
+ Likely simpler to implement... similar process to DML
- Because DHL participants wouldn't really be doing "work" for the network, this would make Dash more closely resemble an investment... I believe that we would nonetheless have a compelling case why Dash is not a security, but it might nudge Dash in a direction that regulators would view less favorably. Not a reason not to do it, but it is a consideration.
- Not that scalable. We might need a minimum of 100 Dash to prevent the list from becoming too long for nodes to maintain. Or we would need to rearchitect the way the DML list works to make it more scalable. Either way, this is less than ideal.

Protocol level shared MNs:
+ Not difficult to implement, but there are a few minor complexities we could work through quickly (e.g., the engineers have practical solutions in mind)
+ Can scale very well. Completely feasible to stake 10 Dash or less, which mean the benefits are extended to the most number of users.
+ Provides incremental decentralization and security to the network... because it provides benefit to the network, this would make Dash no more susceptible to being labeled a security than it is today.
+ Ensures that shared MN services are trustless, which eliminates the need to take risks that people currently take to earn rewards on smaller balances
- Might require time-locked inputs to properly function, which may be less attractive to users
- Doesn't fully solve the fundamental issues with the network economics, but it can defer those issues to a later time, when they are less impactful (e.g., several years from now, the inflation rate would be much much lower)

I believe by combining reduced mining expenses with greater access to the MN rewards (and increasing the share of block rewards that go to this larger group of recipients) we can successfully mitigate MOST of the economic issue with a relatively simple solution. My current hypothesis is the following (assuming no change to the proposal funding share, which I view as a fully independent discussion):
1) Introduce shared MNs to enable greater access to staking rewards
2) SLOWLY increase the share of block rewards from 45/45/10 to 80/10/10 over a very long period of time. Our previous reallocation was far too fast. If we reallocated about 0.5% per month, it would take nearly 6 years to fully transition.

One other benefit of this approach is "fairness" to miners, as they would retain the ability to mine dash with the full life of their existing equipment. In fact, I need to model it, but I believe even they would be financially better off. How is this possible if we are cutting the amount of Dash allocated there? First, they would be mining slightly less Dash the next 18 months, but at an otherwise higher price, which would help. They also would face less competition from new mining capacity because miners would be unlikely to invest in new X11 equipment knowing reward reductions were baked into the protocol. Lower competition would in turn extend the life of existing equipment and allow miners the opportunity to mine for longer without replacing their miners. Most of a miner's profit takes place in the first 12-18 months under normal circumstances, so this could create a unusually long runway of profitable mining if new equipment is not brought to market at the same pace it otherwise would.

Another benefit... low risk of implementation. The only complex piece is masternode shares, which wouldn't be overly complex as it would simply require a different DML registration type, a way for users to signal their inputs are available to stake, and the ability to distribute the rewards differently as part of the coinbase transaction. All of that is based on technology we already have, and would simply need to modify.

Would like to hear reactions to this model from the community.

Very interesting and an approach that I am likely to support.

One question though, why is the DML scalable down to 10 Dash and the DHL only scalable down to 100 Dash. Aren’t they basically the same thing?

 

[Ryan Taylor]

Very interesting and an approach that I am likely to support.

One question though, why is the DML scalable down to 10 Dash and the DHL only scalable down to 100 Dash. Aren’t they basically the same thing?

The deterministic masternode list (DML) currently only records less than 5,000 masternodes and was designed for that type of scale. A deterministic holders list (DHL) would probably need to hold hundreds of thousands of records if we had no restrictions on it, a scale at which the system couldn't operate. So we would need some kind of minimum to prevent the DHL from becoming too large. Perhaps that would be 100 Dash or 50 Dash, but it would exclude a lot of people, or we would need to engineer something completely different to facilitate it.

 

[toknormal]

Don't mistake your own motivations for the motivations of others

I'm not. I and others invested in a Proof of Work asset when Proof of Stake / Hybrid Proof of Stake has been available for the last 7 years. Even pre-dating Dash.

We all made the same choice. It turned out to be the correct choice because POW now dominates store of value while POS only serves utility tokens. That has resulted in demand for hashrate. A demand which Dash devalues by supplying half a coin for the price of one compared to our competitors.

I think if you really are so convinced that revising Dash's fundamental algo will appeal more to the market than the approaches I've presented above, you should test it out first by adding masternodes to Peercoin rather than adding POS to Dash. At least the former option would be seen as extending the feature set rather than bining the operating system and starting again.

 

[JGCMiner]

The deterministic masternode list (DML) currently only records less than 5,000 masternodes and was designed for that type of scale. A deterministic holders list (DHL) would probably need to hold hundreds of thousands of records if we had no restrictions on it, a scale at which the system couldn't operate. So we would need some kind of minimum to prevent the DHL from becoming too large. Perhaps that would be 100 Dash or 50 Dash, but it would exclude a lot of people, or we would need to engineer something completely different to facilitate it.

Ok. That makes sense. No matter how many distinct pieces make up a masternode, the DML only keeps track of complete masternodes so it won’t grow to be unwieldy. Got it.

One more thing if you don’t mind, how would the hosting work for these trustless masternode shares? Has a hosting provider offered to run the remote nodes created from these trustless shares? Or is hosting something DCG will get into?

 

[Ryan Taylor]

I'm not. I and others invested in a Proof of Work asset when Proof of Stake / Hybrid Proof of Stake assets have been available for the last 7 years. We all made the same choice. It turned out to be the correct choice because POW now dominates store of value while POS only serves utility tokens. That has resulted in demand for hashrate. A demand which Dash devalues by supplying half a coin for the price of one compared to our competitors.

I think if you really are so convinced that revising Dash's fundamental algo will appeal more to the market than the approaches I've presented above, you should test it out first by adding masternodes to Peercoin rather than adding POS to Dash. At least the former option would be seen as extending the feature set rather than bining the operating system and starting again.

To be clear, I'm not proposing PoS. We think it will be a long time before PoS can be made adequately secure. I tried to articulate in the presentation that I gave that PoS was included because it illustrates the maximum extent to which PoW could be reduced (to zero) to demonstrate the maximum potential economic impact. I was not endorsing PoS as the solution. There are tradeoffs with any of the solutions discussed, and we're discussing a range of possibilities. It's clear that PoS or even hybrid PoW / PoS would be very difficult to implement, and we also know that we could reduce PoW while maintaining a large hashrate compared to all other X11 chains combined.

 

[Ryan Taylor]

Ok. That makes sense. No matter how many distinct pieces make up a masternode, the DML only keeps track of complete masternodes so it won’t grow to be unwieldy. Got it.

One more thing if you don’t mind, how would the hosting work for these trustless masternode shares? Has a hosting provider offered to run the remote nodes created from these trustless shares? Or is hosting something DCG will get into?

Likely, an MNO wishing to use other's collateral it would simply be part of the ProRegTx. They would probably need to provide 10% or 20% of the collateral themselves (to align their incentives with the other participants to keep the node operating / earning, and to prevent them from controlling an unlimited number), and identify the other inputs that will constitute the remaining collateral totaling 1,000 Dash. No central entity would operate these nodes. They would continue to be run by MNOs in a similar fashion.

 

[toknormal]

To be clear, I'm not proposing PoS.

But you are advocating a further revision of the reward ratio away from miners. This is where our perspectives diverge because your model of where the value is is not consistent with what Dash investors have bought.

At the moment the priorities you've identified focus on securing the network and you've asserted that by doing so at a cheaper cost than the protocol currently supports we'd be delivering more value to the market.

However, if you actually observe where the (pure monetary asset) market has parked its capital, it isn't in cheaply secured blockchains since they've been available for a very long time and have ended up valueless. Rather it's in high energy-budget coins. i.e. coins that have required large amounts of natural energy to produce on a cumulative basis.

In that context, Dash has a great advantage in being able to provide on-chain services in that group. It's the only leading Proof of Work coin that can (because to do so you need to decouple the service provision aspect of the protocol so it can work independently of the mining process which clocks in slow blocktime. For that you need masternodes).

Utility coins on the other hand are only concerned with.......utility ! That's all they're there for. They don't need any kind of store of value monetary attributes which is why they use POS and we are now starting to see all kinds of technology-driven specialisation which blows Dash's utility features away.

IMO therefore, Dash has to consolidate its strength which is providing utility in the POW sector where it's competitive. That means:

• selling hashrate for a high price
• giving primary buyers a competitive amount of coin for their money (not 1 in every 2)
• only subsidising the service layer to the extent that it adds value to the coin (but growing it because that's where we're strong)

This sector is also diminishing in terms of players but growing in terms of value so it would be a highly rewarded challenge to be able to stay in the pure POW race IMO.