Consensus Mechanisms

[strophy]

Perceived Problems: PoW

• PoW is expensive
  • Unlike CPU and GPU hobbyists, which could retain their rewards in the past, miners today operate expensive equipment and operate as a business
  • Miner liquidation is now near 100% of miner revenue
  • The network spends ~$16M annually on miners at a price of $51/Dash
  • ASIC PoW is likely not a significant source of new users
• PoW contributes to long-term price volatility due to the economics exerted on miners
• PoW economics tend to concentrate hashrate toward the lowest cost/largest scale operations, leading to centralization and erosion of security
  • Just TWO POOLS control the Dash network without ChainLocks
• PoW miners’ interests diverge from that of users (e.g. fees)

Perceived Problems: Masternode Participation

• Dash users with less than 1,000 Dash are unable to collect rewards
• Rewards are viewed as important driver for adoption, especially for store of value use cases
• Calls from community to implement masternode shares or reduce masternode collateral requirements are focused on reward generation, but network operations may suffer from using these approaches:
  • More masternodes would slow the propagation of blocks, reducing the scalability of the network
  • Smaller ownership shares would likely provide a lower incentive to properly evaluate proposals OR to vote
  • Network would likely see a nearly equal number of masternode exits as a result of reward frequency dropping, affecting masternode ROI and causing masternodes to exit their investments

Proposed Solution(s)

• Consider moving to proof of stake
  • Less resources needed to stake → revenue is not dumped on the market - instead, it is collected by users with no need to sell
  • Eliminates $16M in annual expenses
  • Price (and budget) would be positively impacted immediately
  • Small Dash investors would be able to generate income for securing the network and producing blocks; in fact, it would initially be more profitable to stake than to operate a masternode
  • Voter turnout would improve - those less interested in voting would shift to staking (more flexible denominations, initially more profitable)
  • Masternode ROI would increase as masternodes shut down in favor of staking
  • The number of MNs would not decline over time - in fact absent a shift in the allocation, number of MNs would continue increasing (it would become a fixed proportion of the coins in circulation)
  • Vote quality would be maintained (since 1,000 Dash requirement would remain)
  • PoW contribution to price volatility would be eliminated
  • Block allocation would become less critical to “get right” - changing the allocation between staking and MN rewards would change the number of MNs, but not the ROI of either group
  • The only allocation that would “dilute” Dash holders would the portion allocated toward proposals
  • Increases security compared to two block producers constituting ~62% of PoW mining capacity
• Related Idea: ‘Deterministic Hodlers List’ / Savings Accounts
  • Users would be able to register to be rewarded for simply holding a certain minimum amount of Dash (say, 100 Dash) without having to run a node
  • Considerations:
    • List could be potentially huge - would need to ensure network could handle
    • Would require allocation of some % of block reward to pay for this new “job”
    • Addresses only the network participation issue, not the question of block production

Key Questions for Discussion

• How will we secure the network?
• How much money will this save?
• Allocation of savings: how will this money be distributed?

YouTube link: Why Was Dash PoW?
Lire en français

 

[qwizzie]

Key Questions for Discussion

• How will we secure the network?
• How much money will this save?
• Allocation of savings: how will this money be distributed?

Those are some tough key questions, i need to think about them some more.

I do have some questions with regards to 'Deterministic Hodlers List’ / Savings Accounts' :

* Who exactly will be our target group, when we place a barrier of 100 Dash on these Savings Accounts ?
* Assuming these Savings Accounts check Dash addresses on-chain, how do we avoid Masternode owners registering their collateral address for this as well ? (and getting interest twice)
Can these Savings Accounts cope with that ?
* How long do these Savings Accounts need to contain unspend amounts, before they receive interest ?
* These Savings Accounts will connect to cold wallets / hardware wallets, in full control of users .. correct ?
* Will these Savings Accounts run on our Mainchain or possibly on a sidechain ?

 

[GaryS]

I think that one of the only "interesting" things about Dash currently is the Masternode aspect. It still has the mantle as being the biggest Masternode project around, it still means something to finally rustle up enough Dash to get a MN.

Moving to a Proof of Stake seems really desperate and will just add Dash to the other ever-growing pile of POS coins out there - and I think will eventually tank the price more.

 

[p5yc071c]

Everyone believes that the model we're using for the MN layer works, right? We've combined PoS and PoSe....can't we just learn from that amazing innovation and just re-use it for a 'mini MN' layer...separate from the existing MN layer? Correct me if I'm wrong, but it seems a little simpler, and makes the network a whole lot more capable...I'm imagining like 10 Dash/Raspberry pi to start a mini mn

What's the largest issue here?

During Ryan's presentation he mentioned who was benefitting from the PoW model...imho this is what we need to highlight. If the 300,000 DASH is creating 18% inflation and is largely the people selling the metaphoric shovels (Mining rig manufacturers/power companies) and taking value from the holders of the currency.

How would a 'mini MN' layer affect the transaction price and the average mini MN operator? Could we build a separate test network and run some fail runs, or would that be cost prohibitive? ZEN Coin has a similar model, and they're running 28,000 nodes with a market cap that's 1/10th of ours...so would 100,000-300,000 nodes be enough to secure the network?

I made another related post in the 'quick win' category

 

[JGCMiner]

My thoughts (and feelings) on this issue. Skip down to the bold to get to the idea if you don’t want to read my justifications. Interesting enough it is similar to the post above:

Just made it through Ryan’s AMA on the podcast. Great work by @thedesertlynx and @Mark Mason. Also, kudos to Ryan. Say what you want about his ideas, but he definitely showed some backbone by standing up and answering community concerns for nearly 2 hours. Above and beyond the call IMO.

And, well -- after seeing that AMA, and seeing some of the community discussions since the open house. . . I have to say that I am much less enamored with the idea of moving to POS, in any capacity. I can list all the reasons, but they are much the same as others have posted on various chat platforms. My main concerns are that the POS security model is prone to centralization and there is real doubt if it is as solid as POW. And a custom chainlocks, POS solution would not have the years of hardening that comes with more established models. Very risky. Then you also have the fact that many believe that POW provides intrinsic value over just simply securing the network. I am not fully in this camp, but I do see their point. And so on. . . In short, messing with layer 1 block production might blowup in our faces.

By the same token, I also think that messing with the masternode protocol is also risky. I am with Ryan on this one. I can’t really get behind the idea of increasing the MN reward percentage or adding timelocks as a way to incentivize more whales to buy nodes (or current whales to hold tighter). Maybe I am wrong, but that just seems like it will cause another face-melting bubble, followed by another epic crash. Great if all us old-timers (like me) just want to dump everything on some poor minnow’s head. . . but, man, adjustments like this just feel so scammy. Maybe it is just me. Same with reducing the miners’ portion while keeping everything else the same. That would reduce the overall block reward and, as Ryan said on the podcast, that is just a bridge too far. IMO. Immutable currency supply and all that. Yes, I know we have changed the block reward in the past (I was there for all of it), and I know that our final coin supply floats because of the treasury, but I just don’t think overtly messing with the coin emission schedule will inspire confidence in investors. . .

So the idea: As a compromise of sorts, I would like to see a Deterministic Stakers List (DSL). Same idea as the Deterministic Masternode List (DML), but you would have a much more accessible minimum of 10 Dash rather than 1000. The mining percentage would then be reduced over some months from 45% to 22.5%. The remaining 22.5% goes to the stakers. The masternode protocol is unchanged. The security model is unchanged. We are just paying half as much to the utility companies (via the miners) for chain security – and there is no resulting change to the overall block reward. POS is not needed and we should keep enough hash to remain the dominate x11 POW chain.

Stakers would get rewards proportional to the number of Dash they have registered to the DSL. Or maybe, if you want to slow centralization, you could taper the rewards for stakers with more than 1000 Dash. The particulars could be tweaked. Regardless, my main point is that with DSL we give back to the regular users, enticing more regular people (i.e. not whales) to join the ecosystem, without going to POS and without inflating the masternode count like you would with decentralized MN shares. Again, I feel like messing with the masternode protocol is risky even via something like shares. Decentralized shares might cause larger holders to dump as their MN ROI shrinks. DSL avoids this potential pitfall while allowing low commitment (10 Dash) staking. As a side benefit, since we already have a DML creating a DSL alongside should be fairly simple (though I am not a developer so who knows).

Anyway, I am sure people will have a field day poking holes in this and I know I am hardly the only one with their pet idea. But yeah, throwing this out there anyway. Cheers!

 

[JOL]

for those who have not directly read the reasons for using the bitcoin POW, I invite them to read the whitepapper.
https://bitcoin.org/bitcoin.pdf

this particular passage is highlighted POW reason

The proof-of-work also solves the problem of determining representation in majority decisionmaking. If the majority were based on one-IP-address-one-vote, it could be subverted by anyoneable to allocate many IPs. Proof-of-work is essentially one-CPU-one-vote. The majoritydecision is represented by the longest chain, which has the greatest proof-of-work effort investedin it. If a majority of CPU power is controlled by honest nodes, the honest chain will grow thefastest and outpace any competing chains. To modify a past block, an attacker would have toredo the proof-of-work of the block and all blocks after it and then catch up with and surpass thework of the honest nodes. We will show later that the probability of a slower attacker catching updiminishes exponentially as subsequent blocks are added.To compensate for increasing hardware speed and varying interest in running nodes over time,the proof-of-work difficulty is determined by a moving average targeting an average number ofblocks per hour. If they're generated too fast, the difficulty increases.

And

12.Conclusion
We have proposed a system for electronic transactions without relying on trust. We started withthe usual framework of coins made from digital signatures, which provides strong control ofownership, but is incomplete without a way to prevent double-spending. To solve this, weproposed a peer-to-peer network using proof-of-work to record a public history of transactionsthat quickly becomes computationally impractical for an attacker to change if honest nodescontrol a majority of CPU power. The network is robust in its unstructured simplicity. Nodeswork all at once with little coordination. They do not need to be identified, since messages arenot routed to any particular place and only need to be delivered on a best effort basis. Nodes canleave and rejoin the network at will, accepting the proof-of-work chain as proof of whathappened while they were gone. They vote with their CPU power, expressing their acceptance ofvalid blocks by working on extending them and rejecting invalid blocks by refusing to work onthem. Any needed rules and incentives can be enforced with this consensus mechanism


Bitcoin is about electronic transaction without relying on trust and he imutable ledger where consensus mecanism Proof-of-work is essentially one-CPU-one-vote

From my point of view, what we looking for, its the more efficient way to have transaction ledger security with a consensus mecanism base on the more largest numbers of participant. Where centralisation (as ASIC) is the oposit of the initiale requirement for "without relying on trust"

By hes whitepaper, Bitcoin is :
- A ledger
- A unit of account in this ledger
- And the cost of using this ledger service by the fees

 

[GaryS]

"As a compromise of sorts, I would like to see a Deterministic Stakers List (DSL). Same idea as the Deterministic Masternode List (DML), but you would have a much more accessible minimum of 10 Dash rather than 1000"

This just sounds like a slippery slope. 1000 too many? let's make it 10! Now everyone can get involved. I mean if you're gonna do this, might as well just make it a full staking network.

But this leads back to my earlier point, what will differentiate Dash if it just goes full POS, soon this fad (Cosmos / Tezos etc.) will fizzle out and we'll be left with another mediocre half privacy half payments "staking coin".

I dunno man, I feel uncomfortable about this. But it's good that people are talking at least. Bringing a bit of interest to otherwise dead community participation. Hope we can work sometyhing out.

 

[JGCMiner]

I am not sure what you mean by slippery slope. My idea doesn’t do anything to MNs or the required 1000 Dash collateral. That was kind of the point. MNs are working fine — let’s not touch that in any way.

This idea is just shifting some portion of mining to staking BUT NOT using proof of stake to do so. Miners would still make new blocks.

This is the basic question: If we are overpaying for security given Chainlocks, why let miners sell to play electric bills? Rich(er) utility companies don’t help Dash. Rather, lets give some of that money to regular users.

 

[qwizzie]

I am not sure what you mean by slippery slope. My idea doesn’t do anything to MNs or the required 1000 Dash collateral. That was kind of the point. MNs are working fine — let’s not touch that in any way.

This idea is just shifting some portion of mining to staking BUT NOT using proof of stake to do so. Miners would still make new blocks.

This is the basic question: If we are overpaying for security given Chainlocks, why let miners sell to play electric bills? Rich(er) utility companies don’t help Dash. Rather, lets give some of that money to regular users.

I do wonder what kind of barrier would need be set to avoid making the DSL (Deterministic Stakers List) / DHL (Deterministic Holdlers List) / Savings Accounts too taxing on our network.
Ryan propose 100 Dash, you propose 10 Dash.

• Users would be able to register to be rewarded for simply holding a certain minimum amount of Dash (say, 100 Dash) without having to run a node

I would prefer 10 Dash over 100 Dash to allow more Dash users to participate, but i have this suspicious feeling 10 Dash would be too taxing on our network.

What i do like about DSL / DHL / Savings Accounts, is that it can be implemented without much hassle. We can then take our time to investigate our available options for possibly switching to proof of stake in a later phase
(or not switching to proof of stake, if that investigation leads to that direction). It basically means postponing the proof of stake discussion and do more research first.

 

[GaryS]

I am not sure what you mean by slippery slope. My idea doesn’t do anything to MNs or the required 1000 Dash collateral. That was kind of the point. MNs are working fine — let’s not touch that in any way.

This idea is just shifting some portion of mining to staking BUT NOT using proof of stake to do so. Miners would still make new blocks.

So you're kind of proposing Masternodes + Mininodes + Mining (reduced)?

 

[JGCMiner]

I do wonder what kind of barrier would need be set to avoid making the DSL (Deterministic Stakers List) / DHL (Deterministic Holdlers List) / Savings Accounts too taxing on our network.
Ryan propose 100 Dash, you propose 10 Dash.

I would prefer 10 Dash over 100 Dash to allow more Dash users to participate, but i have this suspicious feeling 10 Dash would be too taxing on our network.

What i do like about DSL / DHL / Savings Accounts, is that it can be implemented without much hassle. We can then take our time to investigate our available options for possibly switching to proof of stake in a later phase
(or not switching to proof of stake, if that investigation leads to that direction).

Yeah, I am definitely not sure about network constraints. And admittedly, I was thinking more about accessibility to these saving accounts than network load.

I am not sure if an idea of this nature will get any traction in the community, but if it does we will need a dev’s input before going much further.

 

[JGCMiner]

So you're kind of proposing Masternodes + Mininodes + Mining (reduced)?

Yeah, Ryan had something similar in his presentation but with 100 Dash as @qwizzie pointed out.

These “mini-nodes” would just stake though. Like a saving amount. They would not vote or support IS, PS, or Evo — nor get those associated fees on Dash Platform. That stuff would be reserved for normal MNs.

 

[qwizzie]

So you're kind of proposing Masternodes + Mininodes + Mining (reduced)?

Yeah, Ryan had something similar in his presentation but with 100 Dash as @qwizzie pointed out.

These “mini-nodes” would just stake though. Like a saving amount. They would not vote or support IS, PS, or Evo — nor get those associated fees on Dash Platform. That stuff would be reserved for normal MNs.

It would form a nice, safe, not too drastic, compromis solution. A more conservative approach.

 

[Geert]

Here is a thought experiment for a solution where LLMQ quorums emit every other block with POW backup:

There is a new quorum type, one designed to emit blocks. There are three of these quorums active at any one time. Every other block, the network will only accept a valid block as having been signed by one of these three quorums (for the first five minutes).

After 2.5 minutes, quorum one signs and publishes the new block. If after three minutes no block, quorums two signs and emits the new block. If 3.5 minutes no block, quorum three signs and emits the next block. If there is no new block after five minutes (all quorums have failed), miners may now also emit the new block (POW).

The new LLMQ signed blocks would be be subsequently chainlocked just like a traditional POW block.

 

[forro]

I enjoyed the open house and the dash news q&a. They seemed well-reasoned. However, and I heard this mentioned elsewhere, and I agree, that Dash's slide has less to do with selling pressure from miners, and more to do with the market's skepticism on DCG's ability to deliver evolution/platform. It has been a very long time, and DCG needs to be honest with itself and admit it has done a terrible job interfacing with the public (quarterly Q&As cut short every time, infrequent 'weekly' updates, consisting of only a few one-liners, only a handful of blog entries over the last few years, etc). The enemy here is loss of confidence in DCG, not the miners. However, DCG is in the home stretch of this very big project. It will mark a turning point when the world sees that Dash delivers on it's promises no matter how difficult nor how long it takes, due to it's superior governance and design. All dash needs to do is stay the course and make good on it's promises. Once confidence is restored due to the useful nature of the platform the developers, users, and investors will come and the price will reflect that.

--

Miners and masternodes receive rewards because they provide important, valuable services to the network. What services would stakers actually provide? If they are to take away part of the miners reward, then they need to be contributing somehow to block production and network security. In that case, they would need to be always available, and have a full copy of the blockchain.

It is the nature of mining that it attracts new participants until margins are razor thin, and only the most efficient can survive the competitive environment. It will always be the case that miners will need to sell nearly everything they have to stay in business over the long term. Unless... Why not make miners stakeholders? Whatever happened to the collateralized mining discussion? Would it cause a reduction in the number of miners? Would it artificially cap the number of miners, changing the economics? Would they have more breathing room and rather than dump dash as fast as they can to keep the lights on, they can help set a floor on price? Perhaps long term the same situation would emerge: while holding their 10 dash or 100 dash to mine, they would still need to sell whatever they could produce to survive against the competition. It seems the difficulty algorithm would ensure that is the case. Even if so, could other effects from it benefit the network? Has a game theory discussion been made publicly available? What are the pros and cons? Would it help decentralization of mining? Would it help break up mining pools? In any case, I do like the idea of miners becoming stakeholders, as it presumably further aligns their interests toward the benefit of the network, chainlocks or no chainlocks. The barriers to entry should be extremely low for users, but quite high for the guardians of the network (miners and masternodes). Can you be double bullet-proof?

As mentioned elsewhere BLS can fail under stress, so it's not a good idea to have the masternodes doing it all. However, if stakers were able to participate in block production in a separate and independent block-producing quorum, maybe at first every twentieth block could be selected from the stakers. Over a very long period of time and repeated stress tests bugs are worked out it works so well that eventually staker blocks could be selected more frequently than minerblocks until they're alternating, with a 22.5 split. At that point it could even be possible to open up block production so miners and stakers are directly competing for every block.

Whatever we do, we should do it gradually. Ryan previously carried a much more conservative viewpoint on dash, and it's still appropriate.

 

[thephez]

* Assuming these Savings Accounts check Dash addresses on-chain, how do we avoid Masternode owners registering their collateral address for this as well ? (and getting interest twice)

The list of addresses used when registering a MN are stored by Dash Core - for example, the operator key must be unique (you get an error if you try to use the same one for multiple MNs). That would allow checking if an address was already used in a MN to prevent it from signing up for both rewards.

* How long do these Savings Accounts need to contain unspend amounts, before they receive interest ?

A MN must be in the list for one full cycle before being eligible for payment. Presumably a similar thing could be part of such a solution for "savings accounts". This also probably limits how quickly people would want to move back and forth between MNs and savings to chase ROI. If you go to the back of the line every time you switch, the potential ROI would have to be large enough to risk loss of current rewards (or you at least might wait until you receive your next payment).

 

[drkrooster]

So the idea: As a compromise of sorts, I would like to see a Deterministic Stakers List (DSL). Same idea as the Deterministic Masternode List (DML), but you would have a much more accessible minimum of 10 Dash rather than 1000. The mining percentage would then be reduced over some months from 45% to 22.5%. The remaining 22.5% goes to the stakers. The masternode protocol is unchanged. The security model is unchanged. We are just paying half as much to the utility companies (via the miners) for chain security – and there is no resulting change to the overall block reward. POS is not needed and we should keep enough hash to remain the dominate x11 POW chain.

as a POW chain, I don't think we can reduce the block reward to just 22.5% for miners.
as only the miners participate in the actual consensus and adding blocks to the chain, MNs do not.

 

[drkrooster]

there are arguments to switch to POS, but I don't agree that miner is the only factor that causes the price issue for Dash.
to me, strong arguments to change to POS would be POS can better handle the network transactions that we need to have on the Dash Evo platform.

Miners can play an important role in the ecosystem too. If the Miners think Dash is valuable, there will be more than 2 pools.
Miners will re-invest to build on Dash platform and keep some of the DASH they mined and etc. These are not happening, because they think Dash is not useful.

We need to create the demand, and image that Dash is more than useful. Changing to POS due to miners' selling all their mined coins is not a strong enough argument to change.

 

[JGCMiner]

as a POW chain, I don't think we can reduce the block reward to just 22.5% for miners.
as only the miners participate in the actual consensus and adding blocks to the chain, MNs do not.

Those numbers were meant to be illustrative, not final. The actual numbers can, and should, be adjusted. In fact, I would welcome letting research and data speak when it comes to setting any new reward split.

 

[Dandy]

So the idea: As a compromise of sorts, I would like to see a Deterministic Stakers List (DSL). Same idea as the Deterministic Masternode List (DML), but you would have a much more accessible minimum of 10 Dash rather than 1000. The mining percentage would then be reduced over some months from 45% to 22.5%. The remaining 22.5% goes to the stakers. The masternode protocol is unchanged. The security model is unchanged. We are just paying half as much to the utility companies (via the miners) for chain security – and there is no resulting change to the overall block reward. POS is not needed and we should keep enough hash to remain the dominate x11 POW chain.

Stakers would get rewards proportional to the number of Dash they have registered to the DSL. Or maybe, if you want to slow centralization, you could taper the rewards for stakers with more than 1000 Dash. The particulars could be tweaked. Regardless, my main point is that with DSL we give back to the regular users, enticing more regular people (i.e. not whales) to join the ecosystem, without going to POS and without inflating the masternode count like you would with decentralized MN shares. Again, I feel like messing with the masternode protocol is risky even via something like shares. Decentralized shares might cause larger holders to dump as their MN ROI shrinks. DSL avoids this potential pitfall while allowing low commitment (10 Dash) staking. As a side benefit, since we already have a DML creating a DSL alongside should be fairly simple (though I am not a developer so who knows).

Anyway, I am sure people will have a field day poking holes in this and I know I am hardly the only one with their pet idea. But yeah, throwing this out there anyway. Cheers!

I really like this idea. Also, 10 Dash seems like a good point of entry.
What this needs to be viable is a user-friendly integration into the wallet, so people can stake inside a mobile wallet for example.
Also, I would propose that the "unit" of staking be 10 Dash, so you can only stake in increments of 10 Dash, no more, no less. Then we could relatively easily implement a reward system which would go down the list (like for MNs) and for example reward 100 of 10 Dash holding addresses in each block.
When you start staking, you go to the end of the queue (like MNs do). That would lead to a minimum time between two payouts of a week or more, so anyone that want to get rewards must not move collateral from that address for at least that time.