It's about time to turn off PoW mining

[cbeast]

But even 51% and 720 blocks window is orders of magnitude more difficult to exploit than exploiting PoW.

...

[1714990341]

1714990341

[1714990341]

1714990341

[1714990341]

1714990341

[1714990341]

1714990341

[delulo]

........

I see that we have to disagree on a few very basic economic assumptions.

How do the existing delegates prevent an attack of this nature with their hangout? You wouldn't need to hire 52 people , just a few controlling a thousand personalities and profiles.

You need 52 different people (or more if not all of these are voted in) if they should speak and not be identified as the same person.

Energy costs and asic costs actually. Why do you discount this as a security measure?

Energy and hardware costs are only a necessary security cost in POW. You get the same result with DPOS (pools / delegates that can be voted out; miners leave pools and stakeholders vote delegates out, in case mining pools / delegates perform an attack) without those extreme costs. And the tragedy is that once all costs are carried by users (no block reward, only tx fees) those enegery/hardware costs equal the amount of money it costs to attack the network. By introducing trust and voting in DPOS you brake the equation.

You are ignoring increases in efficiency due to Moores law and better ASIC technologies. Hash rate may decrease in growth in the future but is unlikely to ever drop.

Efficiency increases in mining equipment does not matter at all because miners compete with each other. The growth in hash rate that is due to more efficient ASICS are proportional to a falling price per hash rate.

You keep mentioning the hard fork solution to rollback an attack on the DPoS blockchain as a good thing. I don't want rollbacks to be easy, they shouldn't happen at all. The ease this can happen with DPoS is a sever weakness, not strength. Why again did the Nxt community pay off the hackers who stole all the coins from BTER rather than perform a rollback?

I have no clue what NXT did.
Equally simple to see as the points above: If it happens you WANT it to be easy. During the whole discussion you did not provide a sound argument why hard forking was needed more in DPOS than in POW. I just pointed out that in case of an attack it would be easier to undo the harm than in POW. This is easy to see.

How about the example I have mentioned several times? Government regulators demand delegates prevent certain transactions occurring due to terrorists, illegal actions like drugs, or "think of the children" .... blacklisting. What happens to delegates who don't follow the regulations?

First there is no reason why delegates should be more illegal than miners. If being a delegates is as legal as being a miner than governments would have no interest in attacking delegates. It is just a normal business you do. A competitor of Visa (say mastercard) would also have an interest in attacking the Visa headquarters and it doesn't happen. Now delegates are distributed around the world which makes it harder than just attacking Visa.  
Second as we have seen that delegates are equal in what harm they can do to the network to mining pools, there are only three mining pools right now you would have to compromise but about 50 delegates...Both (mining pools and delegates) can be voted out if they do harm or don't perform (not including tx might not be a reason for miners to leave pools by the way).

Yes, I am in agreement.... I want to encourage p2p mining. This can be accomplished many ways within bitcoin without DPoS or PoS.

P2P mining is not the solution to the ASIC centralization problem. I bet that within 5 years there will be less than 3 companies which have more than 51% of all the ASIC mining power.

The attack is by no mean economically profitable. Consider that you have to buy and sell a huge amount of stake in a relatively short amount of time and that once you have 52 delegates in you could make a lot of money from tx fees by being honest (the same counts for POW miners, only difference is what scarce resource has to be proven to the network: energy consuming hashing power or stakeholder approval).

Overall DPOS provides a whole lot more security than POW per cost. There is no perfect system but the alternative (POW) is just vastly inefficient, tends to centralize over time and can not compete when crypto currencies should ever find mass adoption in terms of the ability to scale (tx per second) and provide cost efficient transaction services to customers.

I am a bit tired of this discussion. I thought we could agree on a few basic assumptions and considerations. It is all not that difficult imo. Seems to me that people stick to POW because they know it but continue to ignore the short comings.  

[inBitweTrust]

You need 52 different people (or more if not all of these are voted in) if they should speak and not be identified as the same person.

Not if the delegates are anonymous and as I previously indicated if they aren't anonymous than that presents a whole list of problems.

Energy and hardware costs are only a necessary security cost in POW. You get the same result with DPOS (pools / delegates that can be voted out; miners leave pools and stakeholders vote delegates out, in case mining pools / delegates perform an attack) without those extreme costs. And the tragedy is that once all costs are carried by users (no block reward, only tx fees) those enegery/hardware costs equal the amount of money it costs to attack the network. By introducing trust and voting in DPOS you brake the equation.

I detailed an attack that could be carried out with DPoS that was much cheaper than buying up ASICs and paying for electricity. Please refute those details.

Efficiency increases in mining equipment does not matter at all because miners compete with each other. The growth in hash rate that is due to more efficient ASICS are proportional to a falling price per hash rate.

I was responding to this comment:

The result is that the hash rate would probably drop so that is gets cheap...

Perhaps you meant to suggest the hash rate growth would taper off, while the hash rate kept increasing? Bitcoin doesn't exist in a vacuum and needs to protect itself from other attacks like botnets and not just specialized attacking bitcoin asics. In this context it is important that the hash rate keeps increasing and will likely do so.

If it happens you WANT it to be easy. During the whole discussion you did not provide a sound argument why it was needed more in DPOS than in POW. I just pointed out that in case of an attack it would be easier to undo the harm than in POW. This is easy to see.

What? I gave a detailed attack scenario of how one could attack a DPoS coin with both anomynous and non-anomynous delegates.
 

First there is no reason why delegates should be more illegal than miners. If being a delegates is as legal as being a miner than governments would have no interest in attacking delegates. It is just a normal business you do. A competitor of Visa (say mastercard) would also have an interest in attacking the Visa headquarters and it doesn't happen. Now delegates are distributed around the world which makes it harder than just attacking Visa.  

Yes, actually we do want delegates and miners to perform illegal actions equally under the cover of strong cryptography. If you don't understand this you don't understand one of the central purposes of crypto-currencies and the history of the cipherpunks who created this technology. BTSX can become the next government coin an comply with all regulations and see where that leads you; bitcoin will remain sovereign.

What is legal often conflicts with what is ethical or just.

can be voted out if they do harm or don't perform

Sure, they can be voted out, just like in representative politics. How is voting for the "lesser of the two evils"(or group of sociopaths) working out for people during the last couple hundred years?

P2P mining is not the solution to the ASIC centralization problem. I bet that within 5 years there will be less than 3 companies which have more than 51% of all the ASIC mining power.

Possibly, but that is something we should struggle against and a battle worth fighting.

I thought we could agree on a few basic assumptions and considerations. It is all not that difficult imo.

I have agreed that DPoS has several security benefits with the current Bitcoin status quo. You have yet to give a detailed rebuttal to how cheaply an attack could happen within DPoS. Or are you content with DPoS becoming the next highly regulated government coin by having non-anonymous delegates follow the laws of a territory?

[delulo]

...Their voice does not make them not anonymous entirely.  

...I have shown how hard it would be to carry out such an attack at all, how it is impossible to do this profitably and that there is an emergency plan that can be applied faster than in POW.

As for the p2p pools: Overall you can not fight / struggle against economies of scale which leads to ASIC centralization....

An attack would be way more expensive with DPOS (or impossible) than in POW when both systems are matured. POW maturation means no block reward / no inflation = extremely cheap to buy 51% of the hashing power and ASIC centralization. DPOS maturation means delegates being well recognized exchanges and other companies trusted by stakeholders.

Thanks for your contributions to this discussion inBitweTrust. I will retire from this thread I think. There are a few things we will not come to a conclusion on.

[cbeast]

...Their voice does not make them not anonymous entirely.  

What does this even mean and what does it have to do with security?

[inBitweTrust]

DPOS maturation means delegates being well recognized exchanges and other companies trusted by stakeholders.

Thanks for your contributions to this discussion inBitweTrust. I will retire from this thread I think. There are a few things we will not come to a conclusion on.

Yes, and thank you for your contributions as well. I believe the crux is your solution can be more "secure" and efficient against small time crooks as it ultimately relies upon trusted and well known delegates to properly secure the protocol. We have different definitions of security however as I want to promote an open source protocol that is secure from both hackers, gangs, corporations, and governments and what you propose only solves part of those considerations.

This doesn't mean that DPoS cannot fullfill an important role in the white market. In reality I forsee a likely scenario where an inflationary DPoS type coin gets adopted by certain governments, but much to the chagrin of bitshare proponents it will be unlikely to be their proposal as you have made the fatal flaw of creating a deflationary asset that governments won't appreciate.

For the blackmarket/greymarket (which is larger than the whitemarket worldwide) bitcoin will remain an important tool with anonymity and strong cryptography. This is not to suggest that BTC will never go mainstream or used in the whitemarket but that it will fullfil the role of a nuetral asset that doesn't descriminate upon the legalities of a transaction just like cash for the last couple thousand years.

[CoinHoarder]

You have yet to give a detailed rebuttal to how cheaply an attack could happen within DPoS. Or are you content with DPoS becoming the next highly regulated government coin by having non-anonymous delegates follow the laws of a territory?

That's probably because you exaggerated how cheap/easy an attack could be done...

You claim you can attack it with a few thousand dollars and a lot of sock puppets. I dare you to try, I doubt you can get more than a few sock puppets voted in, and you would need 51....

The large stakeholders are only voting in trusted members of the community and/or people that are working on a project that will help BitsharesX in some way, which are also unlikely to perform an attack. Revealing/proving your identity is also a deterrent from this, most delegates can be tied to community members or identities. Before you say this makes it less secure, I point you to the fact the large Bitcoin pools are not anonymous either.

Like seriously... your recent arguments as to the weaknesses of DPOS are filled with exaggerations as to how easy they would be to perform and are borderline FUD. I admit they are possible, but they are just as unlikely as a lot of Bitcoin's possible attack vectors.

We have different definitions of security however as I want to promote an open source protocol that is secure from both hackers, gangs, corporations, and governments and what you propose only solves part of those considerations.

Neither is Bitcoin, or any consensus algorithm for that matter, and no consensus algorithm will EVER be as secure as this... it is a pipe dream.

[inBitweTrust]

I dare you to try, I doubt you can get more than a few sock puppets voted in, and you would need 51....

The large stakeholders are only voting in trusted members of the community and/or people that are working on a project that will help BitsharesX in some way, which are also unlikely to perform an attack. Revealing/proving your identity is also a deterrent from this, most delegates can be tied to community members or identities. Before you say this makes it less secure, I point you to the fact the large Bitcoin pools are not anonymous either.

Like seriously... your recent arguments as to the weaknesses of DPOS are filled with exaggerations as to how easy they would be to perform and are borderline FUD. I admit they are possible, but they are just as unlikely as a lot of Bitcoin's possible attack vectors.

You seem to be selectively reading my statements out of context. I have repeatedly stated that BTSX would be 1 step forward and 2 steps backwards in security. I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

I wouldn't attempt to attack BTSX because stealing is unethical but no worries I'm sure such an attempt won't happen because competition between BTSX delegates will force a majority of the candidates to have public profiles thus making such attack much more difficult. I think BTSX will remain a very niche product for sometime without an attack but is unlikely to take off because :

1) It won't appeal to governments or banks as the largest stakeholders aren't part of their constituency, and isn't inflationary which won't appeal to them. Governments and Banks will use a ripple like solution or launch their own PoS.
2) The fact that a majority of the future delegates will be high profile and public figures in the future will prevent the currency for facilitating Black market transactions as  regulators can easily place pressure upon delegates
3) Bitshares is already so far behind Bitcoin in marketshare and development that it will be unlikely to ever catch up
4) Bitcoin can solve many of these so called shortcomings with side-chains and tree-chains which are being developed right now.

The clock is ticking on BTSX as well because it needs to gain considerable momentum before the initial funding runs dry which is fast approaching.
People complain all the time about the lack of development in Bitcoin when its happening all around them ... that development will dry up quickly with BTSX without the 40-50k a week that is being blown.

I don't think DPoS is superior vs a PoW/PoN,PoS, or DPos hybrid but even if it were the facts above are enough to prevent BTSX from taking off. You do understand that the best product in the market doesn't always win right?

[krach]

2) The fact that a majority of the future delegates will be high profile and public figures in the future will prevent the currency for facilitating Black market transactions as  regulators can easily place pressure upon delegates

yes



keep in mind we are talking to a delagate right now

[CoinHoarder]

You seem to be selectively reading my statements out of context. I have repeatedly stated that BTSX would be 1 step forward and 2 steps backwards in security.

I am not meaning to selectively read your posts. I understand you believe it would be 1 step forward and 2 steps back in terms of security, but the attack vectors you have brought up are incredibly unlikely to happen.. similarly as unlikely as the possible Bitcoin attack vectors. You are also ignoring the benefits such as the speed, scalability, and efficiency the DPoS provide and only focusing on these unlikely vulnerabilities.

I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

Although you may not support centralized mining, it is a fact most PoW mining takes place on centralized mining pools whether you like it or not. I don't like to compare possible idealisms but I like to compare realities.

I wouldn't attempt to attack BTSX because stealing is unethical but no worries I'm sure such an attempt won't happen because competition between BTSX delegates will force a majority of the candidates to have public profiles thus making such attack much more difficult.

Yes, you are starting to see why that attack vector is unlikely.

I think BTSX will remain a very niche product for sometime without an attack but is unlikely to take off because :

I agree BTSX will remain a niche product for the time being and is unlikely to take over Bitcoin, but I think it has a good chance of securing its place and a large market share in certain niche markets. BTSX will attract people that care about the environment and the impact of PoW, people that like cryptocurrencies but don't care for the volatility, people that trade different cryptocurrencies but want to limit third party risk such as centralized exchanges, people that don't care for how slow Bitcoin transactions are, people that don't like Bitcoin prints new money to secure its network and the economic implications that brings, and currency, hard metal and other commodity traders and speculators. All of these people combined could potentially be a huge market if widely adopted and it is why people see value in BTSX.

1) It won't appeal to governments or banks as the largest stakeholders aren't part of their constituency, and isn't inflationary which won't appeal to them. Governments and Banks will use a ripple like solution or launch their own PoS.

You are using a lot of arguments that apply to all other cryptocurrencies. Same applies for every single one as to this point.

2) The fact that a majority of the future delegates will be high profile and public figures in the future will prevent the currency for facilitating Black market transactions as  regulators can easily place pressure upon delegates

Again, this applies to Bitcoin too.. People are using it for illicit transactions yet the government/regulators haven't crushed it they've worked with them. Bitcoin pools, exchanges, payment processors.. The same argument applies.

3) Bitshares is already so far behind Bitcoin in marketshare and development that it will be unlikely to ever catch up

I would argue that BTSX is already ahead of Bitcoin in terms of its speed and features, to each their own but I would rather use BTSX over Bitcoin any day.  Maybe I am naive, but it seems perfectly secure to me and I haven't heard any likely plausible vulnerabilities to make me think otherwise. Market share is ever changing and not set in stone, the Bitcoin infrastructure is well ahead but Alt coins will eventually catch up in these regards.. Bitcoin just had a several year head start. At some point there will be not many ways to improve upon the infrastructure and they will catch up. I think Bitcoin will remain king of cryptos for sometime, but it is not invulnerable and it will need to innovate and improve itself to remain on top in the long term.

4) Bitcoin can solve many of these so called shortcomings with side-chains and tree-chains which are being developed right now.

It is yet to be seen whether tree chains and side chains can be successfully implemented and overcome the network effects of alts that solve the same problems. They are only ideas and concepts, not realities and they could not work for multiple reasons. The country coins have failed by giving out coins to people for free and then they dump them on the market, this may happen with these solutions as well.. it is anyone's guess.

The clock is ticking on BTSX as well because it needs to gain considerable momentum before the initial funding runs dry which is fast approaching.
People complain all the time about the lack of development in Bitcoin when its happening all around them ... that development will dry up quickly with BTSX without the 40-50k a week that is being blown.

DPoS has ways to fund development and infrastructure projects by voting in delegates. You write this off as an insignificant feature, but it is better than the alternative by only being able to fund development by donations. Delegates can currently make 250$/month, give a developer a few delegate spots or so and that will fund his time to continue working on the project maybe not full time, but it will fund some improvement and infrastructure development... It is already being used for this. Since we are a cryptocurrency that is ran like a company, ideas have been floating around about offering more shares as well to fund further development or a huge marketing push, just like big companies do on the stock market. We think about cryptocurrency differently and if we think it is a good business decision it can happen... anything is a possibility and we don't look at it like Bitcoiners look at Bitcoin whereas the rules are set in stone. If it improves the value of everyone's stake in the long run, it may not be a bad idea and has a chance of happening.

I don't think DPoS is superior vs a PoW/PoN,PoS, or DPos hybrid but even if it were the facts above are enough to prevent BTSX from taking off. You do understand that the best product in the market doesn't always win right?

I think you are blind or biased if you can't see some benefits to DPoS over PoW/PoN and PoS. Each system has benefits and shortcomings, and you are focusing on the shortcomings. You are ignoring the scalability, efficiency, and speed DPoS provides and thus the reason it was conceived and implemented. You need to stop and think there must be some benefits otherwise BTSX would have chosen an older consensus algorithm. You think we are all retards? Or DPoS is just a marketing scheme? Come on man...

[Inotanewbie]

I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

Although you may not support centralized mining, it is a fact most PoW mining takes place on centralized mining pools whether you like it or not. I don't like to compare possible idealisms but I like to compare realities.

Pools are not centralized. The largest pool only controls ~25% of the hashpower, and most other pools control between 2% and 5% of the network hashpower. This is hardly centralized. Also as it has been said many times if a pool acts in a "bad" way then miners can easily either start to solo mine or direct their hashpower to another pool with no loss of mining revenue

[CoinHoarder]

I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

Although you may not support centralized mining, it is a fact most PoW mining takes place on centralized mining pools whether you like it or not. I don't like to compare possible idealisms but I like to compare realities.

Pools are not centralized. The largest pool only controls ~25% of the hashpower, and most other pools control between 2% and 5% of the network hashpower. This is hardly centralized. Also as it has been said many times if a pool acts in a "bad" way then miners can easily either start to solo mine or direct their hashpower to another pool with no loss of mining revenue

I realize this, but this is still centralization that is applicable to the conversation. Two pools have 25% opens up another attack vector due to pools colluding, getting hacked, or getting approached by thugs with guns. In DPoS in comparison each delegate controls less than 1% of the network. If they act in a bad way they can be voted out.

It works much the same way as Bitcoin's centralization, except with a couple benefits. My point is centralization is a problem in just about any consensus algo due to mining pools and lease forging (PoS). People act in a way that makes them the most or the most steady amount of money. DPOS recognizes this and has a form of controlled centralization that can be managed by shareholders, whereas you have no control over where people mine or who does the mining.

[inBitweTrust]

....

No, I don't believe you are retards or hinted at this either. Despite someones intellect they are easily blinded by their investments. Mine in BTC and yours in BTSX. I suspect this thread was created out of a mixture of genuine concern for the Bitcoin community, partly to promote BTSX, and partly out of curiosity to listen to opinions outside of the echo chambers of bitshares forums.

I agree with you and have mentioned multiple times their are security benefits of DPoS over Bitcoin and PoW. There are also security weaknesses as well that you fail to acknowledge. 250 a month for dev isn't much but possibly enough to keep BTSX patched with a slow dev crawl once those funds dry up. What is the average delegate salary though?

Part of my argument comes from security considerations comparing PoW and DPoS. Another part comes from the fact that crypto-currencies in general are very fragile at this point in time. While I appreciate a healthy discussion about seeking solutions I do not appreciate an alt trying to cannibalize on bitcoins marketshare with misleading propaganda. Notice the differences between how I am willing to admit advantages within PoS and DPoS and you insist upon their are no advantages of PoW above DPoS? Notice how I entertain the idea of rolling in some of these advantages of DPoS or PoS into bitcoin as potential solutions and you don't want to have anything to do with PoW besides perhaps a transitional speedbump? I believe you really do think that DPoS is the best, end all be all, and I have given rational arguments in which I disagree that you that you claim are far-fetched.

I think it is being disingenuous or naive to suggest Bitcoin could be transitioned from PoW to DPoS. Creating a hybrid consensus mechanism alone is likely not going to happen let alone removing PoW completely.

I have no nostalgia or allegiance to the label "Bitcoin" but I also understand what is at stake here and why it is important we first try and fix Bitcoin's problems before giving up on it so quickly. Bitcoin's failure can possibly put the sovereignty of all cyrpto-currencies at risk.  There are much larger stakes at play than simply having a cheap and efficient currency. Security is very important and simply scaling up the quantity of delegates doesn't address all the possible attack vectors. Yes, I am being picky but this is the process PoW went through and what it is continually going through.

[inBitweTrust]

I'm going to exit this discussion because what is needed more at this time is coding rather than talking. Good luck to everyone.

"Cypherpunks write code."  -- https://w2.eff.org/Privacy/Crypto/Crypto_misc/cypherpunk.manifesto

[zxcv731]

I agree with you. Also not a good time for cloud mining too.

[Brangdon]

The dilemma any ALT faces is that they better hope their coin doesn't beat Bitcoin because doing so will undermine their own currency in relation to Fiat.

Bitcoin has the most market support, the most users, the most developers, the most investors, the most hardware, the most wallets, the most apps, and about 8 times greater than all the other crypto currencies combined in market share.

If Bitcoin fails than it will cast a shadow of doubt within all much , much weaker currencies long term viability as a store of value. People will be unlikely to make investments if any of these digital currencies could become obsolete because of a new feature or a passing meme.

It's an interesting discussion point. The argument for PoS put here is not a passing meme, or even a feature, so much as a fundamental efficiency. I expect to see fees in PoS currencies ending up much lower than PoW ones, and then, if other factors are equal (obviously people disagree about that), then people will shift to the cheaper system if they can. I expect payment processors will make it easy for merchants to accept payment in a variety of crypto-currencies. It's software - they can write it once, roll it out to thousands of merchants, and there-by get a competitive advantage over other payment processors that support fewer coins. There's no loyalty. The best coin will get the bulk of the business, and lesser coins will fade gradually into irrelevance until they are not worth having as an entry in a combobox. Certainly industries do switch technologies when it makes sense to.

The protocol and security design wouldn't save the image either because technically Bitcoin could be converted to PoS, DPoS or a PoS/PoW hybrid

I don't believe that will ever happen, in the sense that the original block-chain and protocol will continue indefinitely. Bitcoin won't evolve; it'll be replaced and the original will fade away. That's because there will never be 100% consensus to upgrade it, and no-one can force the holdouts to shift. It's easier not to try.

Whether the replacement will be some new thing devised by the Bitcoin community itself, or an outside coin (perhaps one that already exists) is another matter.

The best an ALT should and can hope for is being adopted as a sidechain/treechain or maintaining a unique quality with a distant 2 or 3rd place in marketshare.

I just don't see that. If one coin is demonstrably better (eg by having lower transaction fees and no downside), why wouldn't people switch to it? They switched from Alta Vista to Google. They switched from Myspace to Facebook. They switched from leaded to unleaded petrol, and many switched again to diesel. It's one thing to claim a given coin isn't significantly better, and another to claim the best coin won't win.

[a447513372]

I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

Although you may not support centralized mining, it is a fact most PoW mining takes place on centralized mining pools whether you like it or not. I don't like to compare possible idealisms but I like to compare realities.

Pools are not centralized. The largest pool only controls ~25% of the hashpower, and most other pools control between 2% and 5% of the network hashpower. This is hardly centralized. Also as it has been said many times if a pool acts in a "bad" way then miners can easily either start to solo mine or direct their hashpower to another pool with no loss of mining revenue

I realize this, but this is still centralization that is applicable to the conversation. Two pools have 25% opens up another attack vector due to pools colluding, getting hacked, or getting approached by thugs with guns. In DPoS in comparison each delegate controls less than 1% of the network. If they act in a bad way they can be voted out.

It works much the same way as Bitcoin's centralization, except with a couple benefits. My point is centralization is a problem in just about any consensus algo due to mining pools and lease forging (PoS). People act in a way that makes them the most or the most steady amount of money. DPOS recognizes this and has a form of controlled centralization that can be managed by shareholders, whereas you have no control over where people mine or who does the mining.

What would happen if someone were to become multiple delegates? What would stop them from doing this if it does not cost anything to mine? Wouldn't they have an incentive to do so if they wish to protect their "investment" in the PoS coin?

[cbeast]

It's an interesting discussion point. The argument for PoS put here is not a passing meme, or even a feature, so much as a fundamental efficiency.

I haven't seen a PoS that solves the Byzantine General's problem. I see PoS as efficient as fiat and just as trustless as fiat.

[Traffic4u]

I don't support centralized mining pools and have never suggested that is an ideal solution to security yet you still insist on comparing BTSX with centralized mining.

Although you may not support centralized mining, it is a fact most PoW mining takes place on centralized mining pools whether you like it or not. I don't like to compare possible idealisms but I like to compare realities.

Pools are not centralized. The largest pool only controls ~25% of the hashpower, and most other pools control between 2% and 5% of the network hashpower. This is hardly centralized. Also as it has been said many times if a pool acts in a "bad" way then miners can easily either start to solo mine or direct their hashpower to another pool with no loss of mining revenue

I realize this, but this is still centralization that is applicable to the conversation. Two pools have 25% opens up another attack vector due to pools colluding, getting hacked, or getting approached by thugs with guns. In DPoS in comparison each delegate controls less than 1% of the network. If they act in a bad way they can be voted out.

It works much the same way as Bitcoin's centralization, except with a couple benefits. My point is centralization is a problem in just about any consensus algo due to mining pools and lease forging (PoS). People act in a way that makes them the most or the most steady amount of money. DPOS recognizes this and has a form of controlled centralization that can be managed by shareholders, whereas you have no control over where people mine or who does the mining.

What would happen if someone were to become multiple delegates? What would stop them from doing this if it does not cost anything to mine? Wouldn't they have an incentive to do so if they wish to protect their "investment" in the PoS coin?

This is a very good point. The additional cost two have two stakes that each have 1/2 of your holdings of a PoS coin above only having one stake are very small. Since almost all altcoins are very deeply divisible it would be easy to effectively control a PoS network with only a very small holding of a PoS coin.

[achimsmile]

It's an interesting discussion point. The argument for PoS put here is not a passing meme, or even a feature, so much as a fundamental efficiency.

I haven't seen a PoS that solves the Byzantine General's problem.

You present this as if it was a fact. Can you prove it with an empirical example?