elective-stereophonic
elective-stereophonic
NXT POS and Bitshares DPOS  
Please login or register.

Login with username, password and session length
Advanced search  

News:

Latest Stable Nxt Client: Nxt 1.12.2

Pages: 1 [2] 3 4 ... 7  All

Author Topic: NXT POS and Bitshares DPOS  (Read 26522 times)

Agent86

  • Newbie
  • *
  • Karma: +0/-0
  • Offline Offline
  • Posts: 6
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #20 on: July 25, 2014, 11:35:15 am »

We'll have to agree to disagree delulo.  You are not understanding the difference in incentives.
Logged

delulo

  • Jr. Member
  • **
  • Karma: +4/-0
  • Offline Offline
  • Posts: 43
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #21 on: July 25, 2014, 11:37:30 am »

Show me a thread / a post of yours that proves me wrong.
Logged

mczarnek

  • Hero Member
  • *****
  • Karma: +68/-4
  • Offline Offline
  • Posts: 898
    • View Profile
    • Nxt Place - Craigslist for Nxt
Re: NXT POS vs. Bitshares DPOS
« Reply #22 on: July 25, 2014, 06:44:40 pm »

So, let's say that Nxt caps the leasing power per forging pool.  In the simplest implementation, this means you can't lease to an account if your forging power plus theirs would push it over some amount.  I believe the number we came up with was 0.5%

Then the differences I see are:
DPOS essentially allows choosing multiple forgers to lease your forging power to
DPOS only allows the top 101 forgers in terms of leased forging power to forger

First of all, seem like simple enough changes and if the market decides it prefers that form.. easy enough for Nxt to switch over.  Just as much incentive for the big guys to have powerful computers to attractive leasers.  However, I personally prefer allowing the little guy to forge on his own if he would prefer to.

Regarding allowing leasing your forging power to multiple forgers.. I don't see a difference one way or another.  Because you give people more votes and it seems to me your average person will be more free with their votes.  Attacker still gets to forge every once in a while.

And Nxt could hardfork too and allow people to switch up their leasing power.. just doesn't seem like a good idea to even consider it to be a solution.  I'm pretty sure the attitude around here is that if you trust your Nxt to an exchange and the exchange gets hacked.. tough luck.

btw Nxt is not 90% attack resistant in it's current form but with upcoming changes, it should be before the end of the year... I think it's been planned for sometime in November.
Logged
NXT Organization: Tech
Donations greatly appreciated: NXT-DWVJ-G89C-RHNL-6QW6Q

2Kool4Skewl

  • Hero Member
  • *****
  • Karma: +396/-246
  • Offline Offline
  • Posts: 1897
  • Banned!
  • Because I'm a Genius
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #23 on: July 25, 2014, 06:54:50 pm »

I personally prefer allowing the little guy to forge on his own if he would prefer to.

+1  This allows the spirit of true decentralization.

There is no need for individuals to use pools except they get payouts faster.  Right now NXT is worth so little it doesn't really matter.  I can guarantee there will be more than 101 forgers in the NXT system.
Logged
We are the descendants of Bitcoin.  We are the continuation of the cause it started, but that perished with its centralization.
An economic system is a manifestation of an ideology.  What was lost, we shall reclaim.
"The Times 03/Jan/2009 Chancellor on brink of second bailout for banks"

ChuckOne

  • Hero Member
  • *****
  • Karma: +293/-17
  • Offline Offline
  • Posts: 3450
  • ☕ NXT-4BTE-8Y4K-CDS2-6TB82
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #24 on: July 25, 2014, 09:46:45 pm »

I also have a question: Is the following possible? I have a lot of nxt (say 5 percent of the whole stake). I then lease my forging power to one of my own accounts that only has a small balance. This would mitigate the insecurity involved with having to have my bunch of next online all the time. In other words: Can I put my NXT in cold storage and be safe and at the same time have my forging power leased to an account I control and earn tx fees there?

Short answer: yes.

Except that Nxt doesn't have 'cold storage' in the sense that bitcoin has, of course. 'Cold storage' in Nxt terms would mean "having removed all possible online passphrase information", as we use a brainwallet.
Ok. Thanks. I crossed out that disadvantage above and on the bitshares forum....
Do people actually use this option?

Indeed, they do.
Logged

delulo

  • Jr. Member
  • **
  • Karma: +4/-0
  • Offline Offline
  • Posts: 43
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #25 on: July 26, 2014, 01:52:03 am »

Thanks for all the responses.

btw Nxt is not 90% attack resistant in it's current form but with upcoming changes, it should be before the end of the year... I think it's been planned for sometime in November.
Can anyone explain how that (the 90% attack resistance) would work?

Quote
I personally prefer allowing the little guy to forge on his own if he would prefer to.

+1  This allows the spirit of true decentralization.

There is no need for individuals to use pools except they get payouts faster.  Right now NXT is worth so little it doesn't really matter.  I can guarantee there will be more than 101 forgers in the NXT system.
How would you define "true decentralization"?

I think every consensus network (Bitcoin, NXT , Bitshares) has to centralize in the end under the conditions of market competition for low fees / customers.

Bitcoin/POW centralizes because miners migrate to the pool with the lowest fees. Atm this is limited only by a commitment of of Ghash to not take any new miners anymore.

For NXT: If crypto currencies have gotten beyond the speculation phase and enter a competition phase where different systems compete for users with low and lower fees all consensus networks are forced to offer similarly low fees (especially if micro payments and/or machine to machine payments matter). Let's assume in addition that in order to forge a forger needs professional equipment and a reliable internet connection so that the consensus network can guarantee high up time and many tx / second. These costs any forger / delegate has to pay are equal for every system. The consequence is that only those with large stakes can mine on their own, the others will have to join a pool that is equally big, because the costs for equipment are the same for everyone. This will result in centralization (to a certain degree) with NXT.  Here is a more detailed description https://bitsharestalk.org/index.php?topic=5564.0;all of this economies of scale phenomenon.

A System where delegation is possible would be more decentralized. But in practice because NXT allows pools it is essentially the same in terms of the potential to stay decentralized as DPOS. Based on the above paragraph the pools in NXT would, under harsh competition for the lowerst fees between different networks, have to be as big as the biggest stakeholders or a bit bigger because they have to charge some fees.

I also updated my assessment in the OP (Quote 1). Looking forward to your feedback. 
« Last Edit: July 26, 2014, 02:15:00 am by delulo »
Logged

2Kool4Skewl

  • Hero Member
  • *****
  • Karma: +396/-246
  • Offline Offline
  • Posts: 1897
  • Banned!
  • Because I'm a Genius
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #26 on: July 26, 2014, 04:29:15 am »

Thanks for all the responses.

btw Nxt is not 90% attack resistant in it's current form but with upcoming changes, it should be before the end of the year... I think it's been planned for sometime in November.
Can anyone explain how that (the 90% attack resistance) would work?

Quote
I personally prefer allowing the little guy to forge on his own if he would prefer to.

+1  This allows the spirit of true decentralization.

There is no need for individuals to use pools except they get payouts faster.  Right now NXT is worth so little it doesn't really matter.  I can guarantee there will be more than 101 forgers in the NXT system.
How would you define "true decentralization"?

I think every consensus network (Bitcoin, NXT , Bitshares) has to centralize in the end under the conditions of market competition for low fees / customers.

Bitcoin/POW centralizes because miners migrate to the pool with the lowest fees. Atm this is limited only by a commitment of of Ghash to not take any new miners anymore.

For NXT: If crypto currencies have gotten beyond the speculation phase and enter a competition phase where different systems compete for users with low and lower fees all consensus networks are forced to offer similarly low fees (especially if micro payments and/or machine to machine payments matter). Let's assume in addition that in order to forge a forger needs professional equipment and a reliable internet connection so that the consensus network can guarantee high up time and many tx / second. These costs any forger / delegate has to pay are equal for every system. The consequence is that only those with large stakes can mine on their own, the others will have to join a pool that is equally big, because the costs for equipment are the same for everyone. This will result in centralization (to a certain degree) with NXT.  Here is a more detailed description https://bitsharestalk.org/index.php?topic=5564.0;all of this economies of scale phenomenon.

A System where delegation is possible would be more decentralized. But in practice because NXT allows pools it is essentially the same in terms of the potential to stay decentralized as DPOS. Based on the above paragraph the pools in NXT would, under harsh competition for the lowerst fees between different networks, have to be as big as the biggest stakeholders or a bit bigger because they have to charge some fees.

I also updated my assessment in the OP (Quote 1). Looking forward to your feedback.

Let's take the example of a user running a business on NXT.  This business is going to need a "high speed" (NXT TF requires 1.5 Mbps, I'm not sure if that qualifies as "high speed") internet connection anyway to operate online.  NXT PoS requires very minimal processing power to sign transactions.  You can do this on extremely cheap hardware that the business would already have or need to acquire for normal operations anyway (web server).  You don't need "professional" hardware.  The only other cost is power which is negligible considering their hardware is going to be running 24/7 anyway (web server).  There is no reason for these businesses, who are incurring these costs from normal business operations anyway, to consolidate their power in a pool unless they want to reap their forging fees in a steady payout instead of one lump sum.  There is no other pressure to force individuals into pools in NXT.  Unlike bitcoin where the longer you solo mine the less likely you are to find a block because of the continual increase in hashpower on the network.  In NXT, their is no additional hashpower increase over time.

When the NXT network grows to have more than 101 different businesses operating on it, each forging individually because they are incurring the forging costs anyway from basic operations, then NXT will be more decentralized than Bitshares.  This isn't including all the other individuals who decided to forge on their own for other reasons, ideological or otherwise.  The more business that run on the NXT network the more decentralized it becomes.  This is not true with Bitshares.  Bitshares will only ever have 101 forgers regardless of the number of individuals or business running on its network.  It is "decentralize" impaired or restricted.
« Last Edit: July 26, 2014, 04:31:31 am by 2Kool4Skewl »
Logged
We are the descendants of Bitcoin.  We are the continuation of the cause it started, but that perished with its centralization.
An economic system is a manifestation of an ideology.  What was lost, we shall reclaim.
"The Times 03/Jan/2009 Chancellor on brink of second bailout for banks"

ChuckOne

  • Hero Member
  • *****
  • Karma: +293/-17
  • Offline Offline
  • Posts: 3450
  • ☕ NXT-4BTE-8Y4K-CDS2-6TB82
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #27 on: July 26, 2014, 09:49:42 am »

Aren't forging pools a threat to NXT/centralization?  For instance, I say "lease your forging power to me and I will send back 90% of the fees relative to the stake you leased so you don't need to bother with it."  Even if you limit it by address, I can set up multiple addresses to accept leased forging.  I can also set up pools that don't appear to be under the same ownership but actually are.

The even more sinister attack is I set up a pool that actually returns even more than 100% of fees.  I basically pay people to give me power over the network.  You have to rely on people being altruistic to avert this attack.  People have to reject the short term financial gain for the greater good of the NXT network... That's asking A LOT!

That is an interesting process you described.

There is something similar possible in PoW: Proof of Idle (you pay others to reduce their mining power) - http://www.peercointalk.org/index.php?topic=3086

If it could work, I do not know. Maybe it does. However, the timeframe is extremely narrow and expensive.
Logged

Come-from-Beyond

  • Hero Member
  • *****
  • Karma: +794/-671
  • Offline Offline
  • Posts: 4013
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #28 on: July 26, 2014, 09:51:06 am »

Can anyone explain how that (the 90% attack resistance) would work?

I'm drawing comics to explain it.
Logged

valarmg

  • Hero Member
  • *****
  • Karma: +178/-57
  • Offline Offline
  • Posts: 1766
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #29 on: July 26, 2014, 10:11:54 am »


Let's take the example of a user running a business on NXT.  This business is going to need a "high speed" (NXT TF requires 1.5 Mbps, I'm not sure if that qualifies as "high speed") internet connection anyway to operate online.  NXT PoS requires very minimal processing power to sign transactions.  You can do this on extremely cheap hardware that the business would already have or need to acquire for normal operations anyway (web server).  You don't need "professional" hardware.  The only other cost is power which is negligible considering their hardware is going to be running 24/7 anyway (web server).  There is no reason for these businesses, who are incurring these costs from normal business operations anyway, to consolidate their power in a pool unless they want to reap their forging fees in a steady payout instead of one lump sum.  There is no other pressure to force individuals into pools in NXT.  Unlike bitcoin where the longer you solo mine the less likely you are to find a block because of the continual increase in hashpower on the network.  In NXT, their is no additional hashpower increase over time.

When the NXT network grows to have more than 101 different businesses operating on it, each forging individually because they are incurring the forging costs anyway from basic operations, then NXT will be more decentralized than Bitshares.  This isn't including all the other individuals who decided to forge on their own for other reasons, ideological or otherwise.  The more business that run on the NXT network the more decentralized it becomes.  This is not true with Bitshares.  Bitshares will only ever have 101 forgers regardless of the number of individuals or business running on its network.  It is "decentralize" impaired or restricted.

Exactly. Every business that uses NXT will already have a node running 24/7. The forging fees are negligible for most people, so they won't care about maximizing that by leasing (and it doesn't even maximize, just makes payouts more consistent).

For example, jl777 is designing a system have will involve a network of decentralized servers for his InstantDex/Privacy services. Each one will be a Nxt node. That's just a single Nxt business and already hundreds of nodes.

There's no reason there won't be hundreds of thousands of Nxt nodes once Nxt takes off.

What seems to have been lost in this discussion (talking about Nxt becoming centralized in large leasing pools) is that forging fees are not going to be the main incentive to run Nxt nodes. The main incentive will be the advantages that using Nxt provides. Most Nxt nodes won't be run because of fees. It'll be because traders are trading on the MGW, or a businesses are accepting Nxt payments, etc. The forging fees are a nice bonus but for most nodes, they won't be the reason the node is run, and thus they won't care about leasing to the biggest pool to get an extra few pence a day or whatever.

Logged
NXT-CSED-4PK5-AR4V-6UB5V

liondani

  • Jr. Member
  • **
  • Karma: +1/-1
  • Offline Offline
  • Posts: 15
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #30 on: July 26, 2014, 10:24:46 am »



  If small shareholders don't vote it's not a real big deal because their stake doesn't matter much.  Large shareholders will take a couple seconds to do this.  Real companies work this same way, big shareholders care enough to vote.

makes sense!
It is easier for big share holders to vote for security reasons than to get technically involved by them self.  It is not a big deal now because the biggest shareholders are from the crypto community  but that is going to change when mass adoption will take place.  The demographics will change. So it will be much easier for large shareholders to let third trusted party's to secure the network then by they own, since I guess the most of them will not have the IT knowledge and the willingness to setup a super server even if that is a vps. They will prefer to vote a combination of delegates that are suggested for example from future trusted Rating Agencies ...! (?)
Logged

liondani

  • Jr. Member
  • **
  • Karma: +1/-1
  • Offline Offline
  • Posts: 15
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #31 on: July 26, 2014, 12:05:06 pm »

When the NXT network grows to have more than 101 different businesses operating on it, each forging individually because they are incurring the forging costs anyway from basic operations, then NXT will be more decentralized than Bitshares.  This isn't including all the other individuals who decided to forge on their own for other reasons, ideological or otherwise.  The more business that run on the NXT network the more decentralized it becomes.  This is not true with Bitshares.  Bitshares will only ever have 101 forgers regardless of the number of individuals or business running on its network.  It is "decentralize" impaired or restricted.

So you say that bitshares would get less decentralized than NXT in future because of the 101 delegates restriction...
So what would you say if in future the number of delegates on the bitshares network would dynamically change depended on business that run each time?
More business? More delegates. Less business? Less delegates (minimum 101).
If that would happen how would you rate BitSharesX compared vs NXT? Which of them  would you prefer then and why?
Logged

delulo

  • Jr. Member
  • **
  • Karma: +4/-0
  • Offline Offline
  • Posts: 43
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #32 on: July 26, 2014, 12:19:20 pm »

Thanks for all the responses.

btw Nxt is not 90% attack resistant in it's current form but with upcoming changes, it should be before the end of the year... I think it's been planned for sometime in November.
Can anyone explain how that (the 90% attack resistance) would work?

Quote
I personally prefer allowing the little guy to forge on his own if he would prefer to.

+1  This allows the spirit of true decentralization.

There is no need for individuals to use pools except they get payouts faster.  Right now NXT is worth so little it doesn't really matter.  I can guarantee there will be more than 101 forgers in the NXT system.
How would you define "true decentralization"?

I think every consensus network (Bitcoin, NXT , Bitshares) has to centralize in the end under the conditions of market competition for low fees / customers.

Bitcoin/POW centralizes because miners migrate to the pool with the lowest fees. Atm this is limited only by a commitment of of Ghash to not take any new miners anymore.

For NXT: If crypto currencies have gotten beyond the speculation phase and enter a competition phase where different systems compete for users with low and lower fees all consensus networks are forced to offer similarly low fees (especially if micro payments and/or machine to machine payments matter). Let's assume in addition that in order to forge a forger needs professional equipment and a reliable internet connection so that the consensus network can guarantee high up time and many tx / second. These costs any forger / delegate has to pay are equal for every system. The consequence is that only those with large stakes can mine on their own, the others will have to join a pool that is equally big, because the costs for equipment are the same for everyone. This will result in centralization (to a certain degree) with NXT.  Here is a more detailed description https://bitsharestalk.org/index.php?topic=5564.0;all of this economies of scale phenomenon.

A System where delegation is possible would be more decentralized. But in practice because NXT allows pools it is essentially the same in terms of the potential to stay decentralized as DPOS. Based on the above paragraph the pools in NXT would, under harsh competition for the lowerst fees between different networks, have to be as big as the biggest stakeholders or a bit bigger because they have to charge some fees.

I also updated my assessment in the OP (Quote 1). Looking forward to your feedback.

Let's take the example of a user running a business on NXT.  This business is going to need a "high speed" (NXT TF requires 1.5 Mbps, I'm not sure if that qualifies as "high speed") internet connection anyway to operate online.  NXT PoS requires very minimal processing power to sign transactions.  You can do this on extremely cheap hardware that the business would already have or need to acquire for normal operations anyway (web server).  You don't need "professional" hardware.  The only other cost is power which is negligible considering their hardware is going to be running 24/7 anyway (web server).  There is no reason for these businesses, who are incurring these costs from normal business operations anyway, to consolidate their power in a pool unless they want to reap their forging fees in a steady payout instead of one lump sum.  There is no other pressure to force individuals into pools in NXT.  Unlike bitcoin where the longer you solo mine the less likely you are to find a block because of the continual increase in hashpower on the network.  In NXT, their is no additional hashpower increase over time.
At the moment you are right. Like I said my assumption is visa scale network sizes and hard market competition. Then the forgers in the network have to be able to guarantee high up time, high bandwidth and it also takes a lot of high end (=expensive) CPU, memory.

When the NXT network grows to have more than 101 different businesses operating on it, each forging individually because they are incurring the forging costs anyway from basic operations, then NXT will be more decentralized than Bitshares.  This isn't including all the other individuals who decided to forge on their own for other reasons, ideological or otherwise.  The more business that run on the NXT network the more decentralized it becomes.  This is not true with Bitshares.  Bitshares will only ever have 101 forgers regardless of the number of individuals or business running on its network.  It is "decentralize" impaired or restricted.
But you assume that the businesses (i guess you mean e.g. a trusted issuers of NXT assets, a lottery company based on nxt, a DNS register; what else did you think of as an example?) also have the biggest NXT stakes. Why should that be?
Logged

delulo

  • Jr. Member
  • **
  • Karma: +4/-0
  • Offline Offline
  • Posts: 43
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #33 on: July 26, 2014, 12:27:17 pm »

Can anyone explain how that (the 90% attack resistance) would work?

I'm drawing comics to explain it.

Looking forward to your art work!

Quote
What seems to have been lost in this discussion (talking about Nxt becoming centralized in large leasing pools) is that forging fees are not going to be the main incentive to run Nxt nodes. The main incentive will be the advantages that using Nxt provides. Most Nxt nodes won't be run because of fees. It'll be because traders are trading on the MGW, or a businesses are accepting Nxt payments, etc. The forging fees are a nice bonus but for most nodes, they won't be the reason the node is run, and thus they won't care about leasing to the biggest pool to get an extra few pence a day or whatever.
As for validating your own tx that is right. But we are talking about forgers, those that find blocks. The number of nodes doesn't matter for a 51% attack if they do not find blocks often enough = have a big stake.
I always talked about centralization of forging/block producing power which is the one that matters to control the network (revise tx, just mess with it due to political reasons). 
« Last Edit: July 26, 2014, 12:32:52 pm by delulo »
Logged

valarmg

  • Hero Member
  • *****
  • Karma: +178/-57
  • Offline Offline
  • Posts: 1766
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #34 on: July 26, 2014, 01:26:19 pm »

As for validating your own tx that is right. But we are talking about forgers, those that find blocks. The number of nodes doesn't matter for a 51% attack if they do not find blocks often enough = have a big stake.
I always talked about centralization of forging/block producing power which is the one that matters to control the network (revise tx, just mess with it due to political reasons). 
But all nodes are potentially a node that will forge a block (providing they haven't leased their power), no? It all depends on the amount of stake and amount of luck. What is the distinction you are drawing between forgers and nodes?


At the moment you are right. Like I said my assumption is visa scale network sizes and hard market competition. Then the forgers in the network have to be able to guarantee high up time, high bandwidth and it also takes a lot of high end (=expensive) CPU, memory.
Why is the high end CPU required? Also, I don't think always-on high bandwidth is necessarily required. If a node is down, then the network will move to the next one. That node will lose it's forging power, but it shouldn't be a big deal for the network.
Logged
NXT-CSED-4PK5-AR4V-6UB5V

delulo

  • Jr. Member
  • **
  • Karma: +4/-0
  • Offline Offline
  • Posts: 43
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #35 on: July 26, 2014, 06:25:12 pm »

As for validating your own tx that is right. But we are talking about forgers, those that find blocks. The number of nodes doesn't matter for a 51% attack if they do not find blocks often enough = have a big stake.
I always talked about centralization of forging/block producing power which is the one that matters to control the network (revise tx, just mess with it due to political reasons). 
But all nodes are potentially a node that will forge a block (providing they haven't leased their power), no? It all depends on the amount of stake and amount of luck. What is the distinction you are drawing between forgers and nodes?
You said it right, "It all depends on the amount of stake and amount of luck". All nodes receive transactions and can check whether they are unspent outputs but they don't have to forge (my definition of forging/mining: Selecting which transactions are included in the next block and selecting which previous block to reference. Agreed?). Everyone can be a node but a node can only be a significant forger if he has a significant stake.

At the moment you are right. Like I said my assumption is visa scale network sizes and hard market competition. Then the forgers in the network have to be able to guarantee high up time, high bandwidth and it also takes a lot of high end (=expensive) CPU, memory.
Why is the high end CPU required? Also, I don't think always-on high bandwidth is necessarily required. If a node is down, then the network will move to the next one. That node will lose it's forging power, but it shouldn't be a big deal for the network.
If you would want to process say >100 tx per second and have 10 second block times you would not only need high bandwidth, reliability / up-time (determining that someone missed his block with TF costs time and makes the network less predictable which might not be acceptable under high market competition) as well as a lot of CPU and memory although I can not assess at which point (how many tx per sec etc.) the CPU/memory begins to play a role.
Logged

ChuckOne

  • Hero Member
  • *****
  • Karma: +293/-17
  • Offline Offline
  • Posts: 3450
  • ☕ NXT-4BTE-8Y4K-CDS2-6TB82
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #36 on: July 26, 2014, 08:17:43 pm »

If you would want to process say >100 tx per second and have 10 second block times you would not only need high bandwidth, reliability / up-time (determining that someone missed his block with TF costs time and makes the network less predictable which might not be acceptable under high market competition) as well as a lot of CPU and memory although I can not assess at which point (how many tx per sec etc.) the CPU/memory begins to play a role.

1000 transactions are not an issue for today's processors. Even less for those to come. Within several years, that will be peanuts. I suspect that even smartphones will be able to handle that without effort.
Logged

toast

  • Jr. Member
  • **
  • Karma: +8/-4
  • Offline Offline
  • Posts: 23
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #37 on: July 26, 2014, 10:03:31 pm »

If you would want to process say >100 tx per second and have 10 second block times you would not only need high bandwidth, reliability / up-time (determining that someone missed his block with TF costs time and makes the network less predictable which might not be acceptable under high market competition) as well as a lot of CPU and memory although I can not assess at which point (how many tx per sec etc.) the CPU/memory begins to play a role.

1000 transactions are not an issue for today's processors. Even less for those to come. Within several years, that will be peanuts. I suspect that even smartphones will be able to handle that without effort.

CPU speed is not really the limiting constraint. Memory capacity (need more RAM), memory bandwidth (need more CPUs), and network bandwidth (need access to bigger cables) are the real killers.
Logged

valarmg

  • Hero Member
  • *****
  • Karma: +178/-57
  • Offline Offline
  • Posts: 1766
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #38 on: July 26, 2014, 10:19:15 pm »

You said it right, "It all depends on the amount of stake and amount of luck". All nodes receive transactions and can check whether they are unspent outputs but they don't have to forge (my definition of forging/mining: Selecting which transactions are included in the next block and selecting which previous block to reference. Agreed?). Everyone can be a node but a node can only be a significant forger if he has a significant stake.

A possible way to solve that would be to create a portal site. Those with high capacity nodes point their server at the portal, and those with Nxt to lease point their Nxt at the portal. The portal ensures that the leased Nxt is distributed throughout the network of high capacity nodes, creating a robust decentralized network.
Logged
NXT-CSED-4PK5-AR4V-6UB5V

ChuckOne

  • Hero Member
  • *****
  • Karma: +293/-17
  • Offline Offline
  • Posts: 3450
  • ☕ NXT-4BTE-8Y4K-CDS2-6TB82
    • View Profile
Re: NXT POS vs. Bitshares DPOS
« Reply #39 on: July 26, 2014, 11:38:14 pm »

CPU speed is not really the limiting constraint. Memory capacity (need more RAM), memory bandwidth (need more CPUs), and network bandwidth (need access to bigger cables) are the real killers.

I do not see killers here. Memory is cheap and is getting cheaper the minute I write this. My home desktop scans (verifying + building up internal datastructures) the entire blockchain (200,000 blocks and 300,000 transactions) within 8 seconds. I consider my desktop mediocre (right now) if not low-end within 10 months. Bandwidth is the only thing I worry about a little bit.

However, considering the growth rate, we will not outpace the real-world technology and infrastructure.
Logged
Pages: 1 [2] 3 4 ... 7  All
 

elective-stereophonic
elective-stereophonic
assembly
assembly