• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

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

Status
Not open for further replies.
xkcd said:
No, my post was not speculation, it was informative, your post is FUD! Crowdnode is developing trustless MN shares right now.
Click to expand...
Knipsel.JPG

Source : https://knowledge.crowdnode.io/en/articles/6531845-intro-to-crowdnode-trustless-masternodes

This is not trustless !! You are required to trust Crowdnode during the setup process. The Dash masternodes operating on the Dash network and not being setup and operated by centralized hosting service providers on the other hand, are trustless. Know the difference, understand the difference, understand the risks.

You seem to still have this nasty habit of calling anyone that disagrees with your posts, of making FUD posts. Such a weak way of participathing in a discussion.
 
Last edited:
A two-tier masternode network would have to deny all voting powers to the super nodes, or risk being corruptible like any other PoS network. Even now, companies like crowdnode are gradually edging closer to this eventual outcome and it makes me uncomfortable.

The potential high price for storing data is of little consequence. The customer is paying for clever and structured access from a fully functional DAO. Otherwise, there would always be cheaper options at the expense of censorship. Let's just say, for example, we add AI contracts to these decentralized databases. In such circumstance the customer is paying for added value, not to fiercely compete with all other centralized storage providers.

It is my understanding - please correct me if I'm wrong - Dash Platform is not designed for storing large blobs of data, but rather, structured data with special properties:

  1. data for notarization, such that it does not just notarize what is present but also what is absent i.e. cryptographic proof that at any point in time, the dataset is complete.
  2. to provide full version history of said data.
An example would be a national identity system where you can't just retroactively add or remove people. The data has sequence and history. This would make the following abuses of power harder to execute:
Police use of dead children's identities "justified"
 
qwizzie said:
View attachment 11419
Source : https://knowledge.crowdnode.io/en/articles/6531845-intro-to-crowdnode-trustless-masternodes

This is not trustless !! You are required to trust Crowdnode during the setup process. The Dash masternodes operating on the Dash network and not being setup and operated by centralized hosting service providers on the other hand, are trustless. Know the difference, understand the difference, understand the risks.

You seem to still have this nasty habit of calling anyone that disagrees with your posts, of making FUD posts. Such a weak way of participathing in a discussion.
Click to expand...
For most people this is a pointless distinction (although it definitely shouldn't be hidden). For example, most people, no matter how smart or cautious, don't worry about the security of their local bank branch (trustless MN setup provider) when they open their account. They're only concerned about the long term viability of the institution itself (the Dash network).
 
GrandMasterDash said:
A two-tier masternode network would have to deny all voting powers to the super nodes, or risk being corruptible like any other PoS network. Even now, companies like crowdnode are gradually edging closer to this eventual outcome and it makes me uncomfortable.
Click to expand...

I am also thinking about voting powers to the super nodes, but i am more worried how that could effect voting participation. If super nodes are allowed to vote but are simply not interested in the governance aspect of Dash and are just there to collect their super node payments, then that could undermine voting participation even more (which is already low). Which means we may need to change our governance model as well.

On the other hand super nodes which do have an interest in the governance aspect of Dash, could get too centralized and too large and could exert too much influence in our governance system.
 
Last edited:
qwizzie said:
View attachment 11419
Source : https://knowledge.crowdnode.io/en/articles/6531845-intro-to-crowdnode-trustless-masternodes

This is not trustless !! You are required to trust Crowdnode during the setup process. The Dash masternodes operating on the Dash network and not being setup and operated by centralized hosting service providers on the other hand, are trustless. Know the difference, understand the difference, understand the risks.

You seem to still have this nasty habit of calling anyone that disagrees with your posts, of making FUD posts. Such a weak way of participathing in a discussion.
Click to expand...

Get fucked qwizzie. You have to deposit the share in the masternode to CN initially while all the signing is done because

1664442805673.png


Once that part is complete it is fully trustless, this is the best we can do now, comparing CN to MCM is just plain fucking FUD and quite typical of you!
 
GrandMasterDash said:
A two-tier masternode network would have to deny all voting powers to the super nodes, or risk being corruptible like any other PoS network. Even now, companies like crowdnode are gradually edging closer to this eventual outcome and it makes me uncomfortable.

The potential high price for storing data is of little consequence. The customer is paying for clever and structured access from a fully functional DAO. Otherwise, there would always be cheaper options at the expense of censorship. Let's just say, for example, we add AI contracts to these decentralized databases. In such circumstance the customer is paying for added value, not to fiercely compete with all other centralized storage providers.

It is my understanding - please correct me if I'm wrong - Dash Platform is not designed for storing large blobs of data, but rather, structured data with special properties:

  1. data for notarization, such that it does not just notarize what is present but also what is absent i.e. cryptographic proof that at any point in time, the dataset is complete.
  2. to provide full version history of said data.
An example would be a national identity system where you can't just retroactively add or remove people. The data has sequence and history. This would make the following abuses of power harder to execute:
Police use of dead children's identities "justified"
Click to expand...

SMNs will be granted 10 votes in the case the collateral is set to 10k, this is as one might expect, I see no reason for not allowing them to vote. Any reason you think for not letting them vote is equally valid for denying the current Masternodes from voting, which puts you in a paradox.

What is your beef with Crowdnode??? They are getting more people into voting, all the votes they cast are on behalf of their constituents and fully transparent, thanks to the CN voting block we were able to pass the DCG money grabs last week which otherwise would have failed to reach the threshold.

You are right that Platform is not designed for storing large files, but also the centralised model that QE is suggesting is not designed to store national ID cards, hold a democratic national election, or any myriad of problems we need to solve today. For those we would need to take advantage of the entire network and likely move the network behind TOR so that nodes could not easily be attacked by opposing nation states.
 
xkcd said:
Get fucked qwizzie.

Once that part is complete it is fully trustless, this is the best we can do now, comparing CN to MCM is just plain fucking FUD and quite typical of you!
Click to expand...

'the best we can do for now' may be good enough for you and if you want to call this trustless, eventhough its setup is not trustless (which kinda reminds me of Zcash and their trusted setup) then that is fine. I have the right to disagree with that, which i do.

I wish you could refrain of personal attacks and using f words, but i guess thats too much to hope for. Some people never change. It does make discussions with such a person a bit pointless. So i think i go back to ignoring you. Have a nice day.
 
Last edited:
@xkcd I think you are assuming that a 10x collateral is equal to 10x masternode work, which may or may not be the case. I'm not sure how we can verify how much work nodes perform compared to others hardware. Is this a planned feature of Platform?

Currently, voting power is not based on collateral alone. As you previously pointed out, the collateral is there to protect against Sybil attacks i.e. disincentivize the fast accumulation of evil nodes. Proof of Service per-node must still be achieved for voting, so I'm wondering how we determine the voting power of these supernodes?

I think it is already recognized by the developers that large data blobs such as photos are best kept on IPFS and notarized to Dash Platform. There are some potential shortcomings with this but it's not a deal breaker for MVP.

My beef is not specifically with crowdnode but with all mass accumulated custodial solutions. Unfortunately, it's very difficult to disincentivize this form of centralization but it shouldn't mean we embrace it either. How's it going to look when one day it comes to light that a custodial provider lost their voting keys to a hacker?

When there is a large custodian, proposals can be effectively down-voted by distributing the same number of Yes votes across multiple nodes. Likewise for the distribution of No votes, ultimately distorting the outcome of voting.
 
GrandMasterDash said:
@xkcd I think you are assuming that a 10x collateral is equal to 10x masternode work, which may or may not be the case. I'm not sure how we can verify how much work nodes perform compared to others hardware. Is this a planned feature of Platform?

Currently, voting power is not based on collateral alone. As you previously pointed out, the collateral is there to protect against Sybil attacks i.e. disincentivize the fast accumulation of evil nodes. Proof of Service per-node must still be achieved for voting, so I'm wondering how we determine the voting power of these supernodes?

I think it is already recognized by the developers that large data blobs such as photos are best kept on IPFS and notarized to Dash Platform. There are some potential shortcomings with this but it's not a deal breaker for MVP.

My beef is not specifically with crowdnode but with all mass accumulated custodial solutions. Unfortunately, it's very difficult to disincentivize this form of centralization but it shouldn't mean we embrace it either. How's it going to look when one day it comes to light that a custodial provider lost their voting keys to a hacker?

When there is a large custodian, proposals can be effectively down-voted by distributing the same number of Yes votes across multiple nodes. Likewise for the distribution of No votes, ultimately distorting the outcome of voting.
Click to expand...

I am just relaying what QE told us about the voting power of the Super MNs, if the collateral is 10k, they will have 10 votes, if it is 4k, then 4 votes and so on. regarding the votes of clusters like Crowdnode, they are so transparent, we even capture them here https://mnowatch.org/crowdnode/ if anything went awry there, we would see it and call it out instantly, in the case their votes keys got lose, which might happen since by definition they need to be online, then they can quickly re-key the MNs with new keys, the threat of their votes being used nefariously is minuscule.
 
I don't understand... I thought if I choose to upgrade my node(s) to run Platform I get paid additional Platform usage fees to offset/eliminate/surpass (short/mid/long-term) the higher operation cost of a High Performance node?
We already have a 1k Dash barrier of entry. To introduce another one would be a horrible decision. To think about performance issues before the actual product is even out is typical over-engineering.

Get it out the door already, we've been waiting long enough. Once you collect empirical data from a mainnet product you may think about optimization, not before.

Geez.
 
Macrochip said:
I don't understand... I thought if I choose to upgrade my node(s) to run Platform I get paid additional Platform usage fees to offset/eliminate/surpass (short/mid/long-term) the higher operation cost of a High Performance node?
We already have a 1k Dash barrier of entry. To introduce another one would be a horrible decision. To think about performance issues before the actual product is even out is typical over-engineering.

Get it out the door already, we've been waiting long enough. Once you collect empirical data from a mainnet product you may think about optimization, not before.

Geez.
Click to expand...

Yes, but If you release the DashPlatform and the transaction fees are huge, nobody will write apps for it, and DashPlatform will be a failure.
In that case you should have a plan in order to reduce the transaction fees, and reduce them as soon as possible, because if the excitement and the curiosity of the developers fades and most of them abandon DashPlatform and focus on competitive projects, then DashPlatform will be a desert place.
 
What does it mean to have HUGE tx cost for devs? HUGE = ???$
 
splawik21 said:
What does it mean to have HUGE tx cost for devs? HUGE = ???$
Click to expand...
Nobody knows what huge means. You can discover the correct tx cost, by using the "try and error" method.

You set a random number of tx cost, and if nobody develops applications, you reduce it.
If the tx cost is too cheap and many devs create apps, you may also need to increase the tx cost, in case the mnos start complaining of high administration costs due to huge dashplatform databases.

Thats why this "try and error" method of fast increasing or fast decreasing the tx cost, should be definitely part of the protocol.
Vote the number, as always.
I am asking for this since 2016, but nobody listens to the obvious. The Dash community is blinded and stupidified, as the prophecy says.
 
Last edited:
I think it would help this discussion about having Platform run on all nodes or only on High Performance nodes, if we received some data about the fees difference and the TPS difference between both solutions, directly from the Dash Platform team or directly from Quantum Explorer.

All nodes solution versus High Performance nodes solution

Higher fees / Lower fees
Lower TPS / Higher TPS

Above is just too vague at the moment and only leads to speculation.

Does higher fees = huge fees ? (i suspect not)
Does lower TPS = under-utilization of Dash Platform ? (not sure)

But since the Dash Platform fee system is still being worked on, i suspect that data is still subject to change and not yet available for this discussion.
I guess we could raise this question at their next Dash Platform Product Update, to get more clarity.
 
Last edited:
qwizzie said:
But since the Dash Platform fee system is still being worked on, i suspect that data is still subject to change and not yet available for this discussion.
I guess we could raise this question at their next Dash Platform Product Update, to get more clarity.
Click to expand...

The devs of Dash Platform cannot give you accurate information about the optimum cost of tx.
The market will define the correct tx cost, and nobody knows what the market thinks.
The only thing you can do is to "try and error" many times, in order to discover the optimum number of the market, or a number close to the optimum.
The faster you can do this "try and error" method, the faster you will discover the optimum number.
 
vazaki3 said:
The devs of Dash Platform cannot give you accurate information about the optimum cost of tx.
The market will define the correct tx cost.
Click to expand...

That is your opinion. I would like to hear directly from devs, as i suspect they have direct influence over these fees and TPS.
 
Last edited:
Looks like a presentation about the High Performance Masternode solution has been prepared by Dash Platform team, to be presented at the end of today's Dash Platform Development Update (see : www.youtube.com/watch?v=yOMbduMPzv0 -timestamp 5:17), but it got postponed to a later date (Friday) due to sickness of the researcher/ presenter.

I am not sure if this means consensus has already been reached among the Dash Platform team in favor of the High Performance Masternode solution and if they are perhaps planning to create a decision proposal for masternodes to vote on after that presentation, or if this is just to highlight that specific solution and provide more clarity.

That presentation on Friday should hopefully provide more details with regards to the differences between the Masternode solution and the High Performance Masternode solution.
 
Last edited:
I believe a set of options with pros and cons will be presented and yes a masternode vote will be the final outcome, exciting times indeed! I'll update the thread with details as they come to light.
 
If we say the factor of a supernode is 10x of a masternode, we can calculate the number of new supernodes as 3671 / (10 + 1) = 333.

So the question is, if 333 nodes qualifies as "sufficiently decentralized", then what the hell are we doing with 3338 redundant nodes? (3671 - 333)

I'm making the assumption that the security and reliability of Dash Platform should be the same as payments. If not, why not?

For comparison, there are 4058 Zen Secure Nodes and 37747 Secure Nodes. Similar to Dash's plan, Zen has two node types.

Bitcoin has 14414 nodes which receive zero financial incentivizes.

Now watch the justifications from the people that champion Crowdnnode because your average Joe can't afford a dash masternode.. yet they are defending 10x for a supernode.

And finally, tell me, why must Dash Platform be cheap to use? I thought it was going to be revolutionary, nothing else like it is out there? What is this, cold feet or proof that Dash Platform / Evolution is impossible / impractical to achieve?
 
Status
Not open for further replies.
Back
Top