Should Platform run on all nodes or should Platform run only on High Performance nodes ?

[GrandMasterDash]

@vazaki3 , don't be silly. If you are running a node of basically any blockchain, you likely already are hosting data related to porn/WikiLeaks/etc and have been for years.

No. He is saying, forced to go with hard forks that were never part of the Dash vision. In 7 years, when did DCG communicate to us this change in vision whereby some MNOs would be deemed more privileged than others?

Not forced, in the sense the code is open source, but if we're going to deviate from a 7 year vision, then let that deviation go their own way with a new name and team.

What if tomorrow, DCG asked for ALL collateral to be raised to 10K and that they would accept the longest straw regardless of super majority. Would you accept this too?

 

[GrandMasterDash]

I dont think so.
Many blockchains may facilitate hidden financial transactions (or may publish url links) related to porn/wikileaks/etc, but most of them dont have the database that contains the raw data of porn/wikileaks/etc. There is a big difference on this. If you facilitate the transactions or publish the links, you may argue in front of a court and claim plausible deniability. If you store the data themselves, the plausible deniability argument weakens, especially in case these illegal data are unencrypted.

Plausible deniability only applies to the largest / wealthiest companies (Apple, Microsoft etc) or to those licensed (telcos etc).

The company policy must also be consistent. It's not going to work if other parts of your business enforce moderation. You're either completely hands-off or you moderate. Middle ground would be the hardest to defend.

 

[vazaki3]

Plausible deniability only applies to the largest / wealthiest companies (Apple, Microsoft etc) or to those licensed (telcos etc).

The company policy must also be consistent. It's not going to work if other parts of your business enforce moderation. You're either completely hands-off or you moderate. Middle ground would be the hardest to defend.

When the agents are asking to the large companies to spy or to delete some data, these companies comply immediatly and censor the data or inject spying viruses (they call them "security updates") into the targeted hardware or software. There is no hands-off at all. Everyone is forced to censor or to spy.

I think this is what @QuantumExplorer implies, about open source developers compeled to insert back doors into their software, in fear of prosecution if they do otherwise.

QuantumExplorer said in discord.

1) How exactly is the Platform data stored on the masternodes? Are nodes aware whose data they contain and is that data at any point accessible in unencrypted form by the masternode?

Data is stored as nodes of data in a merkle-ized provable database the first of its kind, found here: https://github.com/dashpay/grovedb. Masternodes are aware as much as anyone of the contents of data if such data is not encrypted. Platform does have the ability to have users store encrypted data that isn't directly linked to them. However it can always be figured out who wrote whatever blob of encrypted data.

 

[GrandMasterDash]

You don't seem to be able to listen to reason.

The majority wants platform out. A decision needs to be made. How do you make a decision without going with the majority? Answer this question.

No they don't. The majority want DCG to meet the expectations they held for 7 years. Do tell us, when did DCG communicate to the network that our expectations couldn't or wouldn't be met and that an altcoin with it's own node requirements would be delivered instead?

If you want to start a new project with different node requirements and consensus, then by all means, please take your project and team elsewhere. But understand, I will not let you use the Dash name.

 

[kot]

This discussion is becoming pointless exchange of posts produced by ego-maniacs.

Let's be productive.

 

[vazaki3]

This discussion is becoming pointless exchange of posts produced by ego-maniacs.

Let's be productive.

Correct.
Although the real problem is not that we are ego-maniacs, the real problem is that we do not speak in a formal way, and thus the same things are repeated again and again.

Lets stop the endless discussion, and put a damned governance question.
Will you add a governane question in the dash budget system?
Shall I?

 

[GrandMasterDash]

Just wait for the heat when Sam actually submits the proposals. Am going to have a field day with this one. But hopefully the Trust Protectors will of stepped in by then.

 

[BoomStick]

For now we have 3 more ideas on the table (AFAIK):

1. Time-locked 1000 DASH collateral MNs by @rion

2. Time-locked 4000 DASH trustless MN shares by @QuantumExplorer

3. Kind of "sharding" by @krilen


Could devs please review viability of solution proposed by @krilen ? I'd like to have a better overview of alternatives

Cheers

 

[vazaki3]

For now we have 3 more ideas on the table (AFAIK):

1. Time-locked 1000 DASH collateral MNs by @rion
2. Time-locked 4000 DASH trustless MN shares by @QuantumExplorer
3. Kind of "sharding" by @krilen
Could devs please review viability of solution proposed by @krilen ? I'd like to have a better overview of alternatives

There is a 4rd option, but development is needed.

And by the way, there are alternative solutions that reside in between the "masternode solution" and the "High performance masternode solution".
For example, the megawhales that own many masternodes, should be allowed to maintain only ONE DashPlatform database. That way the databases' replication is reduced, and thus the fee is also reduced. This will result for Dash to have approximately 127 DashPlatform databases, a similar number to the 100 databases that the "High performance masternode solution" is planning to have.
But the 100 databases of the "High performance masternode solution" are not similarly decentralized as the 127 databases of my plan are. Because in my solution the decentralization is achieved due to the separate individuals that are holding these databases. Decentralization based on proved individuals is a real decentralization, in contrast to the fake decentralization based on collateral masternode addresses.
Why nobody proposed such a solution? Who insists of reducing decentralization or introducing fake decentralized solutions?
If you add a poll, please add my solution in the poll options.

 

[Bridgewater]

Those of you who really value "decentralization", do you value that primarily as a means to:

a) "fairness" (participation accessibility, not changing the rules as much, etc)
b) security (censorship resistance, attack resistance, etc)
c) some combination of both of the above.
d) some other reason.

I ask because some systems can be as (or even more) secure, despite (or even because of) some degree of strategic centralization:
- I consider my rough proposed solution both more "fair" and more secure than the high collateral systems.
- I also consider it better than the original plan of running platform on every node, even though it's more "centralized". The lower fees and other economic advantages far outweigh the minimal advantages of storing data on every single masternode.

I agree that there seems to be multiple ways to look at the decentralization/distribution issue. Maybe my vote would be for (c). I see Quantum/DCG looking at it consistently from the standpoint of the entire network's security--which is understandable because if the whole platform goes down or is compromised by a bad actor, then you don't have a Platform at all.

However, I think this topic of distributed storage has brought up an equally important property of decentralization and distribution: protecting the end user personally. I say equally important, because without people willing to use it, you won't have a network at all.

Take our privatesend security as an example. For the user, the protection is already one of obscurity and plausible deniability. That is actually our strength compared to the encryption-based competition. Maybe a similar approach could be taken with hosting unencrypted data. Is there any way to blind the nodes as to which client uploaded what data?

I don't even mind the ability to have a supermajority vote or an election of trusted custodians to monitor and remove certain files to keep things clean for good PR in certain jurisdictions. The main thing is to indemnify both the users and the nodes. The network as a whole can decide to do some spring cleaning if they want to after-the-fact. That's one way I could see the storage thing working without encryption.

 

[Bridgewater]

Basically the issue is that you need collaterals higher than the base collateral to incentivize only a portion of the network to run it, you need to play with a higher than base collateral and with the allocation to allow something desirable to happen.

Only incentivising a portion seems strage to me. If something is better, then everyone should aspire to do it. Otherwise it is not really a pure "incentive." So this actually means you want to DISincentivise or disuade people from running platform nodes because you only want only good actors with correct hardware running it, but don't have the decentralized means to enforce that. At least that's the reason I've come to understand so far.

Looks a bit like arbitrary central planning to me, and it might be fine for a stopgap measure due to lack of POSE but it might set a bad precedent and does represent a significant change to the project with the establishment of a new proof of stake node class that is similar to other projects--which I suspect receive more legislative scrutiny than we have historically received (privatesend notwithstanding).

Like some others here, I am disappointed that PoSe was not implemented for platform. I thought the whole creation of PoSe itself was because it was necessary for and would be used on platform.

 

[qwizzie]

Like some others here, I am disappointed that PoSe was not implemented for platform. I thought the whole creation of PoSe itself was because it was necessary for and would be used on platform.

To me this signals that Dash Platform in its most essential form (including PoSe for Platform) is just not ready for launch end of this year, it never was.
But instead of acknowledging this, DCG plans to proceed without PoSe for Platform due to feeling a strong pressure to release something end of this year, which now negatively affects its start options of Dash Platform.

Which explains why we get the current three DCG poor quality Platform start options, which either leads to Platform network centralization (bye bye Dash Platform Vision) or leads to a general Dash network safety issue (which makes that option without a PoSe solution for Platform in place, a very difficult option to vote yes on).

If there was a PoSe solution for Platform developed by devs which fixes the Dash network safety issue with regards to Platform on all nodes, i wonder if there would still be consensus among devs for the HPM option with regards to starting Dash Platform.

 

[Bridgewater]

Can someone explain again why one of the possible options isn't just to have Platform be optional for any masternode who wants to participate for a chance at some extra revenue through platform fees?

Is the problem that nobody will want to use platform at first so the nodes won't have much payment from fees for running it? That doesn't seem wrong or abnormal to me--it sounds like the way any worthwhile system begins. I would even run the optional platform code on my node for free for a while just to make sure the darn thing works...then when people start using it and the fees start rolling in, more people will run the platform code as it becomes profitable to do so.

I see wide distribution as a strength not a weakness, and if it means you pay more for this "quality" then it just means the demand isn't high enough yet. Yes it will be more expensive and slower than centralized systems, and that is to be expected--there's no free lunch.

So again what exactly is the danger in simply allowing our current nodes to run platform if they want to? Is it that Platform won't be a good experience at first because people use slow hardware? I can personally live with that while we work on the PoSe solution.

What else, high fees? I can live with that too (as stated above). Security? What exactly is at risk if Platform gets taken over by a bad actor due to too few nodes and low-spec hardware?

Dash users' $millions (formerly Billion*cry*) of investment are on the POW+Masternodes layer. The Evo/Platform was always an optional extra thing to make the base layer more user-friendly and expand its potential. If it goes down, the real money is still safe. I think I heard/read QE say that the identity connections made through Platform will still be valid and connected users can still send to one another even if Platform is down. Sounds acceptable to me for initial release.

I guess I just don't understand the desire to tinker with the economic fundamentals--where some people literally have their life savings parked--just to ensure a good experience on this optional interface so it compares more favorably to other projects that have different concepts and vision. I don't want to see us throw the baby out with the bathwater, as Toknormal might say.

 

[Semarg]

Can someone explain again why one of the possible options isn't just to have Platform be optional for any masternode who wants to participate for a chance at some extra revenue through platform fees?

Is the problem that nobody will want to use platform at first so the nodes won't have much payment from fees for running it? That doesn't seem wrong or abnormal to me--it sounds like the way any worthwhile system begins. I would even run the optional platform code on my node for free for a while just to make sure the darn thing works...then when people start using it and the fees start rolling in, more people will run the platform code as it becomes profitable to do so.

I see wide distribution as a strength not a weakness, and if it means you pay more for this "quality" then it just means the demand isn't high enough yet. Yes it will be more expensive and slower than centralized systems, and that is to be expected--there's no free lunch.

So again what exactly is the danger in simply allowing our current nodes to run platform if they want to? Is it that Platform won't be a good experience at first because people use slow hardware? I can personally live with that while we work on the PoSe solution.

What else, high fees? I can live with that too (as stated above). Security? What exactly is at risk if Platform gets taken over by a bad actor due to too few nodes and low-spec hardware?

Dash users' $millions (formerly Billion*cry*) of investment are on the POW+Masternodes layer. The Evo/Platform was always an optional extra thing to make the base layer more user-friendly and expand its potential. If it goes down, the real money is still safe. I think I heard/read QE say that the identity connections made through Platform will still be valid and connected users can still send to one another even if Platform is down. Sounds acceptable to me for initial release.

I guess I just don't understand the desire to tinker with the economic fundamentals--where some people literally have their life savings parked--just to ensure a good experience on this optional interface so it compares more favorably to other projects that have different concepts and vision. I don't want to see us throw the baby out with the bathwater, as Toknormal might say.

I couldn't agree more.

 

[GrandMasterDash]

For reference. Sia coin creates a free market for data contracts. The host sets the price and the customers consider their options. Each file upload is broken up into 30 pieces and you only need 10 pieces to restore everything. It uses Proof of Storage and this project goes back to 2015. It's main competitors are Filecoin and STORJ.
How to rent storage on Sia
Is my data secure?

Now, the common problem for online data storage is cutthroat competition. At the start of this conversation, Sam made it clear that end user price was important to him. But price is not how you get ahead when you're cutting the fat and operating on razor thin margins. You make money by differentiating yourself and upselling.

Unlike Sia, Platform isn't intended for binary blobs. Platform = database + version control + notarization + uncensorable. What killer apps will be born from this and what price are people willing to pay for this? - that, imo, is the hard part and the reason we should not get ahead of ourselves with too many assumptions.

 

[QuantumExplorer]

Can someone explain again why one of the possible options isn't just to have Platform be optional for any masternode who wants to participate for a chance at some extra revenue through platform fees?

Is the problem that nobody will want to use platform at first so the nodes won't have much payment from fees for running it? That doesn't seem wrong or abnormal to me--it sounds like the way any worthwhile system begins. I would even run the optional platform code on my node for free for a while just to make sure the darn thing works...then when people start using it and the fees start rolling in, more people will run the platform code as it becomes profitable to do so.

I see wide distribution as a strength not a weakness, and if it means you pay more for this "quality" then it just means the demand isn't high enough yet. Yes it will be more expensive and slower than centralized systems, and that is to be expected--there's no free lunch.

So again what exactly is the danger in simply allowing our current nodes to run platform if they want to? Is it that Platform won't be a good experience at first because people use slow hardware? I can personally live with that while we work on the PoSe solution.

What else, high fees? I can live with that too (as stated above). Security? What exactly is at risk if Platform gets taken over by a bad actor due to too few nodes and low-spec hardware?

Dash users' $millions (formerly Billion*cry*) of investment are on the POW+Masternodes layer. The Evo/Platform was always an optional extra thing to make the base layer more user-friendly and expand its potential. If it goes down, the real money is still safe. I think I heard/read QE say that the identity connections made through Platform will still be valid and connected users can still send to one another even if Platform is down. Sounds acceptable to me for initial release.

I guess I just don't understand the desire to tinker with the economic fundamentals--where some people literally have their life savings parked--just to ensure a good experience on this optional interface so it compares more favorably to other projects that have different concepts and vision. I don't want to see us throw the baby out with the bathwater, as Toknormal might say.

Sure I'll explain. The whole way proof of stake works is that you need to have something at stake. If you have nothing at stake then the system just doesn't work.

Now you might say: "Well I have my masternode at stake", but if the system is completely optional without heavy economic rewards you still don't really have anything at stake.

Having nothing at stake causes two issues:
*First: it can allow very highly centralized situations. Maybe one entity really likes platform so starts their 200 nodes, but most MNs don't seem to care -> Those 200 nodes can control waaaaaaaay too much of the network. If they get remotely close to 1/3rd they can stop it, if they get remotely close to 2/3rds they can take control of it.
*Second: there is no incentive to do a good job, because if you don't, well, you didn't really lose anything. This drags the system downwards and could lead to platform just not properly starting. Most systems solve the chicken and egg issue by rewarding miners/stake holders with block rewards. If we have block rewards, it means we need to split block rewards from core as we can't create them from nothing without changing the coin supply. When we split block rewards from core there needs to be an equilibrium that can form. Let's say we give 20% to platform, and 80% to core. Now you are running a core node, it's in your interest to run platform because you want the fees. So the equilibrium would only happen if everyone runs platform.

You asked what is the risk if Platform gets taken over: All credits locked in platform could be either destroyed, or worse withdrawn into Dash.

I care very much about people who have parked their life savings into Dash. That's why I'm doing my best to explain the various solutions we have.

 

[QuantumExplorer]

To me this signals that Dash Platform in its most essential form (including PoSe for Platform) is just not ready for launch end of this year, it never was.
But instead of acknowledging this, DCG plans to proceed without PoSe for Platform due to feeling a strong pressure to release something end of this year, which now negatively affects its start options of Dash Platform.

Which explains why we get the current three DCG poor quality Platform start options, which either leads to Platform network centralization (bye bye Dash Platform Vision) or leads to a general Dash network safety issue (which makes that option without a PoSe solution for Platform in place, a very difficult option to vote yes on).

If there was a PoSe solution for Platform developed by devs which fixes the Dash network safety issue with regards to Platform on all nodes, i wonder if there would still be consensus among devs for the HPM option with regards to starting Dash Platform.

There is no type of PoSe that you want in a blockchain in production. I want to make this clear. Please confirm that you understand this, and are expecting DCG with 20 or so devs to do what other projects with 200+ can not.

 

[GrandMasterDash]

Sure I'll explain. The whole way proof of stake works is that you need to have something at stake. If you have nothing at stake then the system just doesn't work.

Is it fair to say, if you hadn't chosen PoS you wouldn't be in this sticky situation?

 

[vazaki3]

Sure I'll explain. The whole way proof of stake works is that you need to have something at stake. If you have nothing at stake then the system just doesn't work. Now you might say: "Well I have my masternode at stake", but if the system is completely optional without heavy economic rewards you still don't really have anything at stake.

Having nothing at stake causes two issues:
*First: it can allow very highly centralized situations. Maybe one entity really likes platform so starts their 200 nodes, but most MNs don't seem to care -> Those 200 nodes can control waaaaaaaay too much of the network. If they get remotely close to 1/3rd they can stop it, if they get remotely close to 2/3rds they can take control of it.

You can use the same above argument, when arguing in favor of the "proof of individuality".
Why you choose the road of the proof of stake, instead of the road of the proof of individuality?

*Second: there is no incentive to do a good job, because if you don't, well, you didn't really lose anything. This drags the system downwards and could lead to platform just not properly starting. Most systems solve the chicken and egg issue by rewarding miners/stake holders with block rewards. If we have block rewards, it means we need to split block rewards from core as we can't create them from nothing without changing the coin supply. When we split block rewards from core there needs to be an equilibrium that can form. Let's say we give 20% to platform, and 80% to core. Now you are running a core node, it's in your interest to run platform because you want the fees. So the equilibrium would only happen if everyone runs platform.

There is a great incentive to do a good job, if you use "proof of work", "proof of service" or "proof of usefull work" (and by the way, the forum thread that mentions PouW is deleted. why??????). So you can use the above argument, when arguing in favor for the proof of work/service.
Why, once again, you choose the road of the proof of stake, instead of the road of proof of work/service?

I will tell you why. Because the government's agents like neither the proof of individuality, nor the proof of work/service, nor a combination of them. They only like the proof of stake, because they have unlimited dollar money. Either you are an agent, or (more probable) the agents devoured all your mind with their proof of stake propaganda.

Actually I think the one who devoured your mind with the proof of stake propaganda, is your predecessor CTO. The same CTO devoured dash incubator, so that dash incubator was used mainly for the development of a (proof of stake based) Dashplatform. Maybe this is one of the reasons why dash incubator is currently in danger not to be voted in the budget.

 

[qwizzie]

There is no type of PoSe that you want in a blockchain in production. I want to make this clear. Please confirm that you understand this, and are expecting DCG with 20 or so devs to do what other projects with 200+ can not.

Then why use a fork of Tendermint in the first place if you already knew as developer this would lead to a problem with starting Dash Platform on all nodes (due to a flaw with PoS blockchains), which requires a PoSe solution on Platform to work safely ?

The way i see it, this is what happened so far :

1 DCG researched options what to use for a sidechain for Dash Platform
2 DCG settled for forking Tendermint, knowing full well its advantages and its disadvantages
3 DCG knows one of the disadvantages has to do with a flaw in PoS blockchain, directly causing a safety issue for starting Dash Platform on all nodes
4 DCG puts Proof of Service on their internal roadmap since Jan 2022 as a fix (most likely this was a necessity even before Jan 2022), but ended up never developing it, never allocating devs to it, never asking additionally funding for it, never subcontracted it. Basically it was put in a freeze state.
5 DCG announced this topic (starting Platform on all nodes or just starting Platform on some HPM's)
6 DCG released information about the safety issue with regards to starting Platform on all nodes to the Dash community, in this specfic forum.
7 DCG announced three upcoming decision proposals (Platform on all nodes, 4K HPM, 10K HPM)
8 DCG stated Proof of Service was not planned for release of Dash Platform due to lack of funding, manpower, pressure to release.

Where did it go wrong ? I would say right after number 3

 

[QuantumExplorer]

Is it fair to say, if you hadn't chosen PoS you wouldn't be in this sticky situation?

First, I didn't personally chose PoS as Dash has always been built on a PoW PoS hybrid system. Masternodes and quorums are inherently PoS. You put away a stake and then provide services for the network.

 

[QuantumExplorer]

Then why use a fork of Tendermint in the first place if you already knew as developer this would lead to a problem with starting Dash Platform on all nodes (due to a flaw with PoS blockchains), which requires a PoSe solution on Platform to work safely ?

You constantly also keep on talking about PoSe as if it's a magical solution that solves all our problems. Mind you I like PoSe, but we need to be truthful about what it is and what it isn't.

Next you seem to imply that Tendermint is the root of our problem. Can you explain what system(s) in your mind works better?

I would also like to add that the decision to use Tendermint was done by the previous CTO. I had wanted a different solution. At this point though there is no turning back though. Please understand that even though we chose Tendermint as a starting point, Tenderdash is very different and quite close to my original plan. Still I don't think it would have changed the situation for the issues you care about.

 

[stan.distortion]

Proof of service is a key component, ideally it's also payments, the block rewards are distributed based on the amount of verifiable work done.

 

[qwizzie]

You constantly also keep on talking about PoSe as if it's a magical solution that solves all our problems. Mind you I like PoSe, but we need to be truthful about what it is and what it isn't.

Would a robust PoSe scoring solution help with the safety issue that is inherent to the Platform on all nodes option ?
If yes, then i think PoSe is not a magical solution that solves all our problems, it is a solution that fix the safety issue with the option to start Platform on all nodes.
It is a solution that requires time, manpower, development and do i dare say .. foresight.

Also i would like to remind you that till very recent the Dash community was only communicated one way to start Dash Platform, namely to start Platform on all nodes. This was communicated endless times to us through DCG quarterly calls and DCG presentations. It is the default way to start Dash Platform. Now we suddenly get alternative ways to start Dash Platform presented (by introducing higher collateral) and we get informed about a safety issue with the default way to start Dash Platform. A safety issue that would not exist, if devs developed Proof of Service.

Next you seem to imply that Tendermint is the root of our problem. Can you explain what system(s) in your mind works better?

Forking Tendermint would not be the root of our problem, if Proof of Service was developed alongside of it.
Now that it turns out that Proof of Service was never developed for Platform, it does form a problem.

Which raises the question : why was Proof of Service never developed alongside Platform over the years ? Why was it given such low priority by devs ?
With devs knowing full well how this would directly affect the start of Dash Platform.

 

[vazaki3]

Which raises the question : why was Proof of Service never developed alongside Platform over the years ? Why was it given such low priority by devs ?
With devs knowing full well how this would directly affect the start of Dash Platform.

Because of the government agents (proponents of a pure "proof of stake" solution and sworn enemies of any alternative proof solution) who control DCG and(or/xor) threaten DCG.

 

[GrandMasterDash]

https://twitter.com/i/web/status/1583809706799222785

 

[vazaki3]

https://twitter.com/i/web/status/1583809706799222785

The whole cosmos is controled by agents, thats why it is doomed!

Neo - 'The One' | The Matrix [Open Matte] - YouTube

David Ulansey - The Cosmic Mysteries of Mithras - YouTube

 

[qwizzie]

How to Measure Decentralization! Top L1's and Regulatory Risk
www.youtube.com/watch?v=7niqkGspx9s

I wonder how low the Nakamoto Coefficient for Dash would end up, with a centralized Dash Platform (that as we know is running on a PoS sidechain).
Pretty low i guess.

 

[TheNetworkIsBiggerThanYou]

I don't even mind the ability to have a supermajority vote or an election of trusted custodians to monitor and remove certain files to keep things clean for good PR in certain jurisdictions.

Fuck this completely.

 

[QuantumExplorer]

Would a robust PoSe scoring solution help with the safety issue that is inherent to the Platform on all nodes option ?

So there are a few safety issues that could happen. Some of them PoSe helps, some PoSe doesn't really do anything.


So let's list them out in the situation that everyone is forced to run Platform.

* 1 Entity could have too much Power - very low risk on non optional Platform - PoSe doesn't do anything.
* Nodes could propose blocks, but not respond to queries - PoSe helps a lot - this is what it is designed for.
* Platform having a bug that could take down the network - PoSe doesn't do anything.
* Not enough nodes will actually start strong enough Platform - PoSe helps a little.

* Sharding - PoSe required
* Sharding security - PoSe doesn't really help all that much.