Topic: [SKY] Skycoin Launch Announcement - page 111. (Read 381579 times)

Incorrect.

fed·er·a·tion
ˌfedəˈrāSH(ə)n/
noun
a group of states with a central government but independence in internal affairs.

the action of forming states or organizations into a single group with centralized control.

You appear to have ignored that your partner admitted the consensus model in the white paper does not include modeling the economic incentives to game it.

I do not believe it is bullet proof. I do not believe that consensus can ever be leaderless as I explained upthread, unless the actors (e.g. ants) are incapable of survival without the group. Humans have big brains and thus have a large personal opportunity cost. Thus they maximize how they game the collective for personal benefit at the expense of long-term failure of the collective.


Easy to say when you don't have a real system with the same market cap as Bitcoin, being attacked every day to try to find weaknesses.

I would say at this point Bitcoin's is clearly vastly superior to vaporware.

Until we can analyze thoroughly the economic incentives in your consensus model, we can't declare it is theoretically superior to PoW.

What you want is not "decentralization", it is "federation". You trust a number of servers, but each individual chooses the set of servers they trust. Email is federated. It is not a centralized system and there are multiple domains and anyone can run their own email server if they want to and do not want to use a third party domain.

Yes. It is less than 51% attack. This is key.

There is a trade off, as you mentioned before. We allow a smaller number of nodes to dominate the network, to avoid the double spending attack. So the colluding nodes can determine future blocks, but they cannot double spend or rewrite history!

So we allow a DDoS attack, but prevent double spending! That is the trade off. You cannot have both in any algorithm. Any algorithm that prevents DDoS makes the double spending easier and anything that prevents double spending makes DDoS easier. You cannot have both.

If the nodes are very corrupt, if they start holding network hostage or colluding, then people can identify those nodes and begin pruning them from the network. That is important. So Skycoin tolerates DDoS, but disallows double spending. DDoS is merely annoying, with someone refusing to put your transactions into a block, but double spending means reverting transactions that have been executed and spending the outputs to new addresses.

Agreed. The Communist Party is dying and the free market is rising in Asia.


Perhaps some many decades from now, but not before China rises from the bottom-up and becomes the new financial center of the free world, while the USSA (and perhaps Europe) sinks into a police state here and now.

Try to see the world as cycles and not a flatline.

Apologies but frankly I think your economic models are masturbation (not referring to the consensus algorithm model). Sorry. You are geek who is having fun but IMO you are entirely missing the point. (I come from the small-startup commercial software world where we used to speak extremely bluntly in the office. My boss when we created what is now Corel Painter told me, "If you are so smart, why are you working for us and not still running your own software company as you did before". Note I proceeded to leave shortly after that and created by all by myself CoolPage which grossed a half $million in year 2000).

Degrees-of-freedom and innovation is where all prosperity originates. Inflation and deflation have nearly nothing to do with it, because innovation is:

http://unheresy.com/Information%20Is%20Alive.html#Knowledge_Anneals


Update: you are modeling the factors that were more paramount in the Industrial Age where fixed capital investment was a limiting factor on innovation. We are entering the Knowledge Age where all one needs is a computer and their mind.

False. Not decentralized. More accurate statement is "collectivized consensus that claims to be decentralized but is arguably controllable by a few pools". You mean first attempt at fooling people about what is decentralized.

Typo.

Research Note:

This is a paper on Ryan Fugger's Ripple payment protocol. This was the original Ripple and was a decentralized, peer-to-peer currency which pushed the money creation process from the central bank down to individual users. Ripple Labs bought the Ripple name and is unrelated.

http://is.muni.cz/th/139865/fi_m/dp_139865.pdf

Generation 0:
- eGold
- Hashcash (PoW)
- webmoney, paypal
- RipplePay

Generation 1:
- Bitcoin. First working decentralized digital asset.
- Alts Forked from Bitcoin. Copy/pasted, no original code base. Dogecoin, Litecoin. Name and logo change on Bitcoin but separate user communities.

Generation 2:
- NXT (Proof of Stake, Java code base. New methods for securing blockchain. Speculative 3rd gen projects)
- Ethereum (Scripting, contract execution)
- Ripple (digital asset designed to run on a network of bank computers. Currency pair trading)
- Skycoin (post-PoW, post-mining, focused on making "Bitcoin 1.0" solving remaining problems in usability and security, infrastructure/tool chain. Service revenue, Speculative 3rd gen projects)

So don't you see the obvious solution?

Daniel Larimer was on the correct conceptual track with his BitUSD for BitShares. But he messed up some of the details. In particular I don't think a decentralized BitUSD unit can track the exact exchange value (my logic was expressed long ago as AnonyMint, I haven't monitored to see how BitUSD worked out?), and instead I would design the options to track a mutually agreed exchange rate server (each decentralized option bet could agree on a different server so this isn't centralization if we assume that we when pay we can convert to their desired unit in real-time, i.e. this would depend on liquid markets and well automated decentralized exchanges). I didn't like the BitShares centralized design for market clearing. I would need to go over the details of that again.

We need multiple competing centralized currencies (hidden behind hidden servers on a high latency[1] Tor improvement, because hidden servers are broken in Tor and will never be sure enough due to the low-latency design, ditto I2P) and each of them needs to have a unit which is stabilized by a decentralized option market to maintain a peg to the target unit-of-account. So that solves all the problems with instant transactions, because all the consensus lie crap gets tossed out the window.

Then you have the unit-of-account economies-of-scale without the centralization of collectivized consensus. Then you hide the leaders so they can't be coerced. The market can vote with their money on which leaders (currencies) they wish to hold value in but the key point is they can hold it in a pegged unit to their desired unit-of-account, i.e. you could hold CurrencyA or CurrencyB or CurrencyC while pegging your value to the USD, Euro, Bitcoin, or any asset you choose. The advantage is that if you think CurrencyX is more well managed than CurrencyY, then you vote with your value. The exchange value of each currency can float while your value will remain pegged to your choice of unit-of-account.

Even some of those currencies could be decentralized consensus crypto-currencies, but they would have to compete against the centralized choices. And these multiple choices for currencies is a decentralized paradigm, that is congruent with how nature decentralizes. Collectivized consensus as a winner take-all unit-of-account is not decentralization.

This is the (less complex, more simplified) short cut to where we need to be. Any one ready to work with me?

It doesn't matter if crypto zealots can't religiously agree. The masses don't get a fsck about your ideological mumbo jumbo. They don't know the difference between a centralized hidden server and a decentralized consensus lie. They only care that when they click "spend" it happens easily and quickly and that their value isn't volatile and any other advantages that can be offered, e.g. autonomy, anonymity, etc..

Again I like paradigm shifts. I hate digging canals with spoons.

Update: also there needs to be an option to pay, because Tor is a honeypot because it is free. So there is the demand for micropayments that drives your crypto-currency demand.

And so then you help Bitcoin also, providing better anonymity for sending transactions to the network.

And where is most of the demand for Crypto? Silk Road! Gambling! Hiding wealth! Etc.. Nefarious activities and speculation.

This is the idea I have presented before and everyone ignored. I guess I will just have to implement it by myself!

[1] High latency is relative to the traffic in the system. The more traffic the less the latency and be relatively less susceptible to timing analysis. Also Tor is broken for other reasons including the fact that it interfaces to non-hidden servers via exit nodes (which is a can of worms in many ways and leads to correlation). Etc, etc, etc. We'd have to dig into the details of it. Bottom line is Tor needs to be replaced.

We will need to see some economic game theory models such as the economic incentive to game influence to double-spend the system or game what ever incentive you are offering to those who participate in the consensus algorithm.

I am very skeptical that you've discovered the Holy Grail to turn nature upside down and suddenly collectivism can self-organize without leaders. I think influence can always be purchased, and nodes can follow who they are paid most to follow.

Afaics, the way decentralization appears to work in nature is multiple competing entities. When ever there is a collectivization in nature, it is always gamed by leadership, except for perhaps ants. But ants are innately constrained by their inability to survive individually, so they are acting in their best interests to organize to prioritize the colony over self. Humans don't act this way, apparently because we don't rely on a collective intelligence in order to carry out our daily life. Ants can't even function individually without the colony. It appears the ant's brain is not sophisticated enough individually to survive as an individual. It is the collective brain of the hive that raises the collective intelligence so the colony can do amazing feats such as build bridges, etc.. I like to think of an oversimplified analogy where each ant is a neuron or synapse in a human brain.

Yep, I only work on the technical aspect. The other guy who often wrote long posts is more proficient at high vision, economical and strategic aspect.
Maybe he will have a in depth discussion on this with you later.

But I still want to emphasize the vote power paradigm difference again:
- In PoW, you voting power is given by yourself.
- In our alg, you voting power is give by others.

It's something like the difference between emperor regime and democrtic regime.

One could pool hop for reasons of wanting to spread their voting power around not for gaming the payout algorithm. There is a variance disincentive to dilute your mining shares across many pools though. But I also am not convinced your algorithm has a disincentive to not participate in aggregatingconcentrating influence.


Your statistic counts nodes irrespective on their relative influence? I doubt it. If not, then the number of nodes could be much less if their relative influence is much higher?

I think you are obfuscating here. The point is that in Bitcoin the influence is much more concentrated in a few pools. You apparently assume the influence will be less concentrated in your algorithm, thus you claim it will take more nodes to collude.


Bitcoin pools are highly suspicious but nodes get paid more, so they mine there. That is the reality of life. Profit talks, ideology walks. As you pointed out, GHash could hide its influence so that the ideology surely walks and miners get on with maximizing their profit.

I don't see how it will be any different for your algorithm. The nodes will maximize their profit and the influence can hidden with Sybil attack (as it is for Ghash hiding that it is really multiple pools or what ever technique they use to create obfuscation).

In fact I(the guy who wrote the paper) am NOT a westerner, I am Asian, LOL.
Thanks to the internet, the world is now a village and we can collabrate together from all over the world.

The free market has more resources than the USA government. But most of the free market is not interested in mining Bitcoin. I had a design that could potentially get everyone to mine (they didn't need to buy an ASIC, they already have an ASIC inside their recent model CPU) and by making it unprofitable to mine at scale.


Afaics, they can still buy and/or confiscate influence. Afaics you haven't thoroughly analyzed the ways to game your system economically.


There is more ASIC computing power inside existing CPUs than the US govt could manufacture. Moreover they couldn't source enough electricity to match the world's existing computing power.

Let's stay on point. This is drifting off point.

Pool-hopping is the practice of mining in a pool only during the good times, and leaving during the bad times; by so doing, a pool-hopper can get more out of the pool than the value he contributes to it, increasing his rewards at the expense of other miners. Pool-hopping gets its name from the act of constantly hopping into and out of the pool (to either other pools or solo mining).

Pool hopping does not make sense for counting voting power. The power of the miner is still the same.

Yep, so if of the same scale of bitcoin network(8000 nodes), it will need 150 top nodes to collude.

aggreating is inevitable, that's the reason why there are top influential nodes.
actually in our test dataset, each of the most several top nodes has about 30% followees.

a node can follow MANY nodes, so the power it gives out is decentralizated.
If a node is paid to only follow exact ONE node, then the followee is highly suspicious.

I'm sure you are aware of "pool hopping". Miners can surely mine (follow) for many pools. See your point below...


This is not a clear statement. Your whitepaper says roughly 2% of influential nodes can game the system.


You are asserting that there is no disincentive to not participate in aggregating influence in your algorithm. Prove it.

I believe an influential node can offer to pay those who will participate in gaming the system. Which is what PoW implicitly does.


I wouldn't answer that because I am not here to make the developers my enemies. I am continuing here to see if we can reach any mutual understanding that might help us. And so the people who follow me can read my logic not just about skycoin, but generally what I think about crypto-currency and where I want to lead it.

I haven't bought any altcoins yet. I also didn't invest in Bitcoin yet either. Partially because I was bankrupted by a chronic health issue and I made some poor investment decisions in 2012 and lost most of my stored wealth (in a sort of zombie and high adrenalin mental state due to suffering a combination of Dengue, Acute Peptic Ulcer that Exploded, and Chronic Possibly Terminal Viral Infection). I didn't know anything about investing before 2006 when I got into gold and silver. By about 2007 I was starting to make headway on understanding and even tried to sell my silver at $21 at the peak but my sell order was refused by the people holding my silver in the USA. So I had to ride it down as I had to physically remove my silver from there. I published an article predicting the move from $25 to $48 and then crash, a year before it occurred. But again in 2011, I wasn't positioned well to make a sell order at $48 (physical silver that was illiquid in the Philippines), so again I had to ride it down and I tried to get cute by playing with options to try to recover what I thought was what I should have had by that time in net worth. It was about 2012 that I started to take Armstrong more seriously, but it wasn't until after I lost most of what I had remaining that I realized Armstrong was the only analyst whose calls had always been correct. I had followed his writings when he was in prison, but it read to me as discombobulated noise. Only when he starting writing his blog daily, with much feedback from readers, was he able to communicate effectively. When rpietilla gave me a heads up in early 2013 on Bitcoin at $10, I couldn't enter the investment even though I thought he was probably correct, because my tolerance for risk has completed evaporated, because my health was in a tailspin and I had lost most of my wealth and had become very conservative trying to hang on to what I could. Since then I sort of stabilized my situation somewhat, but bankruptcy approaches in a few months, so I am more focused on projects that can generate an immediate income at the moment.

One salient point is that complexity is difficult to manage. That is why it is best to simplify and understand well all the factors that can come into play.

Nope. If you are a miner of PoW, then if you join a mining pool, you voting power goes there.
But if you a node in our algorithm, then you can follow MANY nodes,
as a result, for a network with N nodes, if a node is followed by all the other N-1 nodes, it still can't control the consensus.
But....if in the PoW case, the pool will have 100% computing power.

Not considering other factors here, just for voting power, the paradigm is apparently different:
- If you a miner, you voting power is given by your investment, you can split your power into different pools, but the result is your power is weaken by yourself.
- But in our case, you give other nodes voting powers, meanwhile, your voting power is also given by other nodes.



The pionts are:
- *how many* top influential nodes can control the network. If the number is not small, then it is decentralization.
- If they do evil, can we resist them.

Considering an extreme case, if US gvmnt develop a super miner, then it can beat all of the other mining power.
And so it can control the bitcoin network without even collude with other miners, and we can do thing against it even we know it.

But in our algorithm, US gvmnt will fail on both of the two aspects.


Whatever PoW algorithm is, it's always a paradigm of voting power by invest.
So anytime US gvmnt step it, the decentralization is over.

iamback
are you going to purchase some skycoin?

I have been trying to think of a paradigm shift. That would radically shift the problem space in order to provide a breakthrough.

One potential solution (and I think it is the one the world currency must adopt also because a reserve currency held by one nation causes so many distortions) is to make the unit-of-account a basket of currencies.

Thus there would be multiple competing currencies and we would hold our value in multiple currencies simultaneously, and when we pay we can pay in multiple currencies according to the basket weights.

But the centralization problem remains who will decide the basket.

Money is just a really big problem. It is all about confidence and trust.

Decentralized currency is the concept that we can trust the broad consensus, but the reality is technically there is no way to eliminate influence = power = money.

Sorry I don't think there is any solution to this.

Thus I am thinking now more of how to gain market share for a currency and make sure the trusted parties can't be coerced.

I am giving up on eliminating trust and influence.

For how long are we doing to delude ourselves? We (idealistic Westerners) can pontificate and do research experiments until 2032, but the reality is decentralized currency won't exist except as a figment of our imagination and delusion. Meanwhile the Asians are busy building real capital.