Quote from: ArticMine on January 15, 2016, 04:02:14 AM

Quote from: tifozi on January 15, 2016, 03:41:01 AM

Quote from: ArticMine on January 15, 2016, 03:07:05 AM

Quote from: aerbax on January 15, 2016, 01:29:52 AM

Mike Hearn has an interesting new article on some of the failures of Bitcoin...

https://medium.com/@octskyward/the-resolution-of-the-bitcoin-experiment-dabb30201f7

He makes some very valid points and he is of course right about the negative impacts of the 1 MB blocksize. He is also correct in saying the "bloat" is not the issue, in fact the small blocksize in Bitcoin may have led to the massive concentration of the Bitcoin hash rate in China effectively allowing the Government of China to control Bitcoin. On the other hand I do not agree at all with the criticism of people such as Gregory Maxwell, who I must say has also made many very valid points on the matter of scaling Bitcoin. The simple reality is that the Bytecoin solution of adaptive blocksize limits without a tail emission is also a prescription for disaster.

The problem with Bitcoin is that nobody has found and it may well be impossible to find a way to develop a fee market, in the, absence of a block subsidy, that does not over time converge to one of two undesirable results: Fixed blocksize and infinite fees or infinite blocksize and zero fees. Mike Hearn has made a very persuasive argument as to why a fixed blocksize and infinite fees is such an undesirable outcome; however I am sure that Gregory Maxwell can make an equally persuasive argument as to why an infinite blocksize and zero fees is an equally undesirable outcome.

Maybe the real reason why there has not been a solution to Bitcoin blocksize debate is that a solution may in fact not be possible, if one keeps the 21,000,000 maximum number of XBT limit in place, rather than because of the personalities involved.

So in all of this where does Monero stand? Well Monero is the highest capitalization coin that has solved this problem in a pure proof of work coin. My philosophy on this is that when one takes care of the long term the short term will take care of itself. On the other hand focus on the short term and expect grief over the long term. Monero has taken care of the long term, unfortunately Bitcoin and for that matter most other crypto currencies have not.

Excellent insight. The problem of hash concentration could still be something that needs to be smartly tackled. Otherwise XMR has got some pretty solid design for a crypto currency. I am a big believer in PoW but hash concentration is a potential problem (down the road)

On the markets, seeing some minor positive action even at this hour for a change.

The problem of Bitcoin hash concentration in China was paradoxically a direct result of keeping the blocksize small. This happened because ASIC production is most economical in China. With an adaptive blocksize limit in Bitcoin this likely would not have happened since The Great Firewall of China would have placed miners in China at a very significant disadvantage due to latency. This would have forced Chinese ASIC manufacturers to sell their devices for export. The lesson I see here is that a well designed crypto currency will route around censorship and drive mining to those jurisdictions that do not engage in Internet censorship. Monero is way less vulnerable here since it is likely that both China will be the most cost effective place to manufacture electronics and will also continue to censor the Internet.

Well the myth that "bloat" leads to centralization has been debunked.

Edit: Monero has also other ways to mitigate this risk such as an ASIC resistant algorithm, and initiatives such as smart mining; nevertheless I do agree that this risk will need to be addressed and minimized.

We already discussed these issues upthread[1], and I already explained to ArticMine upthread that tail emission (i.e. Monero's erroneously claimed solution) to insure that block rewards do not come entirely from fees, does not solve the Tragedy of the Commons issue w.r.t. to choosing an optimum block size or transaction fee. Apparently he didn't fully understand what I wrote upthread, or I didn't explain my point clearly enough, because he repeats the incorrect information again above after we had already discussed the issue upthread.

The point is that for any given level of block rewards (whether they come from debasement and/or transaction fees is irrelevant!), if the block size is hard-coded in the protocol then there is some level of transaction rate in which mining becomes unprofitable. Also since a miner is paying all transaction fees to himself when he wins a block, he can spam the network with unlimited block sizes if the block size is not hard-coded in the protocol. This is insoluble dilemma that has no solution for as long as PoW mining is profitable. Monero's claimed solution does nothing to resolve the dilemma and thus ArticMine is making an erroneous claim that Monero solved the dilemma.

The only solution is the one in my design, where I make mining unprofitable and thus mining is forced by the protocol and miners can't pay transaction fees to themselves and they also must include a PoW share with each transaction. Iota makes mining unprofitable too, but it doesn't prevent outsourcing of the PoW to ASICSs, thus centralization of the PoW hashing is the threat in Iota[2] because Iota's mathematical model of consensus can't converge without centralized enforced of that model.

My design also solves the concern about centralization of the PoW hash due to the greater efficiency of ASICs, because in my design ~~I explained how using latency is the only way to prevent the PoW from being outsourced to ASICs~~[mining is unprofitable]. Monero claims that by using a PoW hash designed to be efficient on CPUs that they have solved the problem, but the fact is there is no way to design a PoW hash that is immune to not being more efficiently implemented on an ASIC (which is the concern correctly expressed by tifozi).

Satoshi's design (including Cryptonote coins such as Monero) can't solve these issues. Period. They are all dead ends. My design potentially solves these issues.

Quote from: TPTB_need_war on January 15, 2016, 09:46:02 PM

Quote from: Come-from-Beyond on January 15, 2016, 09:30:38 PM

Quote from: TPTB_need_war on January 15, 2016, 09:23:48 PM

Quote from: Come-from-Beyond on January 15, 2016, 09:16:42 PM

Quote from: TPTB_need_war on January 15, 2016, 09:13:56 PM

However it is a fact that what he says about China controlling Bitcoin mining and this having potentially very negative impacts.

China controlling Bitcoin mining is not an argument. China is 20% of the population, not surprising that China outnumbers here. One could claim that Bitcoin is completely centralized because Earth controls 100% of its hashing power.

If what you wrote were logical, then decentralization would always be the same as centralization.

Decentralization is where the decision making power is spread out to as many actors as possible. A few miners in China making the decisions for all the users of Bitcoin, is not decentralization. You are conceptually conflating the sham of representative democracy (or in this case taxation without representation because we don't even get the illusion of voting) with decentralization.

I recall those days when GHash.io was controlling 51% and people were saying that it's not decentralization, because miners will leave the pool should it abuse its power. Are you sure it's few miners making decisions, not few pool operators?

As I explained upthread, moving between pools doesn't necessarily enable decentralization, because marginal cost of mining will dictate that miners MUST seek out a pool with significant portion of the system PoW hashrate, otherwise the miner will be unprofitable. For the professional miners not at the margins of profitability, they are already centralization.

Even the Chinese claim their mining is highly concentrated by pools not by each miner, and observe how the Chinese government was able to force those pools to adopt a policy of not supporting a block chain increase.

[1]

Quote from: TPTB_need_war on January 12, 2016, 12:22:13 AM

[...]

This paper analyzes the flaws in Bitcoin's algorithm which:

* effectively make it centralized over time
* limit scalability and transaction rate
* cause an irresolvable tension over the ideal block data size
* induce a Tragedy of the Commons on the economics of funding mining
* deny instant confirmations
* allow less than 50% Byzantine fault tolerance (due to selfish mining)

I propose a new design to resolve all these issues that retains the core principle of a proof-of-work block chain.

Quote from: TPTB_need_war on January 09, 2016, 12:40:29 PM

Monero

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

Quote from: TPTB_need_war on January 09, 2016, 12:34:04 AM

Quote from: monsterer on January 09, 2016, 12:26:17 AM

...POW ensures permissionless behaviour

PoW doesn't insure permissionless nor decentralized if mining is inherently centralizing.

In Satoshi's formulation of it and every other one I have seen, it is. Please for example refer to my prior reply to monsterer. Also the upthread post I made about how adding instant transactions via LN (or any transaction rate scaling) forces further centralization.

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

Quote from: TPTB_need_war on January 09, 2016, 12:34:04 AM

There are many reasons that Satoshi's design is centralizing. One is for example that mining is profitable, thus there is a competition to drive difficulty higher and higher such that EACH transaction costs ~$10 in electricity (which is currently paid by mining the investors as professional miners who have loans from the oligarchy+banksters mine with huge farms near hydropower at $50 per BTC cost and sell all that they mine). Even if no debasement was paid to miners, they have a monopoly on transactions added to blocks, so they can charge what ever transaction fees they want if they control 51% of the hashrate. It begs for a oligarchy on mining to develop (which is the direction it appears to be headed and normally government steps in with regulation to aid the oligarchy).

And increasing the transactions per block also forces centralization, thus furthering a trend towards oligarchy control, so saying cost per transaction will fall as transaction rate increases is only true if the resultant oligarchy decides to go for market share first (perhaps as a longer-term strategy of 666 enslavement and world government trend).

Also since mining is only done for profit, then pools are required to deal with huge variance of winning a block for typical (non oligarchy) miners. Pools centralize mining, even if we argue that miners can switch pools, the government can more easily target the pools even if miners switch since by definition there will never be as many pools as miners (not even close).

There are many flaws in Satoshi's design which make it entirely broken from my view. I can't support Bitcoin (nor Monero). I support things that I feel enthusiastically could work out well for mankind and be successful. Bitcoin will succeed only because it is fitting in well with the existing oligarchy's (e.g. Peter Thiel) plans for world domination (and that includes that an illusion of decentralization will persist long enough for the centralization to finally take hold).

Quote from: monsterer on January 09, 2016, 12:26:17 AM

It seems like you have hit an impasse, but without knowing the details of the problem

I had thought so, but I think I explained in this thread how to break the impasse I thought I had hit. But I reiterate I will go over all the details when I am offline to see if I haven't missed something in my thought process while I am writing here online.

First I do agree that Satoshi's design is flawed for the following two reasons:
1) Mining cannot be paid for entirely by fees, so a tail emission is a must.

Mining can be paid for by users who must add PoW to submit their transactions. The other reason for tail emission is because otherwise the coin supply is shrinking towards 0 as users lose coins, which will be much more likely once doing microtransactions and more frivolous widescale adoption.

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

The reason is that a fee market cannot properly develop in the absence of a block subsidy. One has either a fixed blocksize with a mining oligarchy and infinite fees or an infinite blocksize where competition between miners drive fees to zero.

For layman technophobes, your astute point (which many of us technophiles already aware of and I wrote Spiraling Transactions Fees thread in 2013) is that when the block size is unlimited then anyone can include all the transactions that were excluded by a miner that was trying to force higher transaction fees. But if block size is unlimited then transaction spam can be unlimited, so a mining oligarchy must form to constrain block size. The problem is not due to relative block size (so compression wouldn't fix it), but due to the fact that someone centralized needs to decide which transactions are spam or not (otherwise in decentralization at least in Satoshi's design then all spam transactions are allowed).

But you are incorrect to assert that tail emission fixes this. This is a fundamental flaw in Satoshi's design which is not fixed with tail emission. It is fundamental problem of a design that includes blocks. This is one of the flaws that Iota is attempting to correct by eliminating blocks, but then I have conjectured in this thread that Iota has other flaws because it eliminates flaws. This is why crypto gets so frustrating because it seems every design always ends up with centralization, because the CAP theorem can not be avoided!

My idea for fixing all of this, has to do with the way the temporal intrablock partitioning is done.

[...]

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

There is a cost to censorship which has already been demonstrated in Bitcoin. It is manifested in the opposition to an increase in the blocksize by those miners that are based in China. The reason for this is the latency introduced by the Great Firewall of China puts a centralized data centre in China at a significant disadvantage with respect to say a residential connection in Canada. The irony here is that if Bitcoin had allowed for larger blocks earlier on this situation would not have developed since the Chinese ASIC manufacturers would have been forced to sell their devices for export leading to a much more decentralized Bitcoin mining situation. The lesson from Bitcoin is that a small blocksize can actually lead to mining centralization by accommodating censorship rather than the other way around.

Astute, well stated, and agreed. The block size is a double-edge sword no matter which direction it is increased or decreased. Real solutions will change entirely the design.

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

The second factor of PoW that must be taken into account is PoC (Proof of Cold). The key is that in a situation where electricity is used for space heating the marginal cost of mining is zero; however the heat is only valuable if it is decentralized. The key here is that one can easily distribute electricity but not heat forcing decentralization. This in effect gives the small player a very significant cost advantage over the large centralized operation.

Heat from electricity is the most inefficient way to heat (better to use wood, gas, or any carbon). Besides Larry Summers' 21 Inc is dreaming up ways to hook users into being slave oligarchy miners by giving them free heaters and cell phones in exchange for using their device to mine on behalf of 21 Inc (last time I looked 21 Inc was already a significant % of the global Bitcoin hashrate and climbing fast).

Quote from: TPTB_need_war on January 09, 2016, 01:24:53 PM

Note I added point #3 as why PoS is MUCH less secure than PoW.

Basically this boils down to the same conceptual tradeoff as the post where I replied to ArticMine, in that allowing decentralized participation enables unbounded conditions (which is a flaw) but centralizing participation also leads to another set of flaws.

Quote from: TPTB_need_war on January 10, 2016, 02:07:08 AM

Quote from: TPTB_need_war on January 10, 2016, 12:39:20 AM

Decentralization is the normal mode of society. It is when the normal oscillating balance between decentralization and top-down control entirely fails in one extreme direction that society enters a Dark Age for 600 years. Without a balance of decentralization and top-down control, nothing functions and everything collapses.

We don't want entirely decentralization and no centralization as that is just as bad as entirely centralization, e.g. see the reply I made to ArticMine upthread and how 100% decentralized control over what goes in the block chain means a choice between unbounded spam or oligarchy control. Thus the problem is the lack of balance and Bitcoin flip flops either to too much decentralization forcing too much centralization (a Tragedy of the Commons).

Here is that linked post:

Quote from: TPTB_need_war on January 09, 2016, 12:40:29 PM

Quote from: ArticMine on January 09, 2016, 02:14:19 AM

The reason is that a fee market cannot properly develop in the absence of a block subsidy. One has either a fixed blocksize with a mining oligarchy and infinite fees or an infinite blocksize where competition between miners drive fees to zero.

For layman technophobes, your astute point (which many of us technophiles already aware of and I wrote Spiraling Transactions Fees thread in 2013) is that when the block size is unlimited then anyone can include all the transactions that were excluded by a miner that was trying to force higher transaction fees. But if block size is unlimited then transaction spam can be unlimited, so a mining oligarchy must form to constrain block size. The problem is not due to relative block size (so compression wouldn't fix it), but due to the fact that someone centralized needs to decide which transactions are spam or not (otherwise in decentralization at least in Satoshi's design then all spam transactions are allowed).

[...]

My idea for fixing all of this, has to do with the way the temporal intrablock partitioning is done. I will probably find some flaw in my design too, but I will spend more time thinking about it.

Correction: my idea for temporal intrablock partitioning doesn't solve the problem above, but rather it makes block announcements a constant size (and enables instant transactions). The problem above remains and remains for Iota too in the sense that sending unlimited transactions is still an unbounded load on all the full nodes in the system. Remember that market based transaction fees are not a solution to the problem. My proposed solution (ditto for Iota) is again that every transaction has to include a PoW share. Since PoW is unprofitable (or if the PoW share provided is not computing a block solution), then this solution can be employed. However note it works much better in my (and also Iota's) design because in a Satoshi design it is possible your transaction won't get included in the current block so then you need to recompute your PoW share and resubmit your transaction. This is one of the details I was referring to. I think my design doesn't have that problem. Again I don't want to detail it entirely yet. I will get some rest now.

Those naive who are contemplating making their own copycoin lack exposure to level of unresolved problems in crypto land, and the intense level of arcane details they would have to become knowledgeable about.

Quote from: TPTB_need_war on January 10, 2016, 10:09:02 AM

Thus my idea for permissionless improvement is to make mining unprofitable by limiting the debasement rate yet requiring each transaction to include a level of PoW difficulty. The goal is to keep the PoW power in the hands of the users so their client software can (perhaps automatically?) redirect to delegate full nodes which are not censoring transactions or otherwise maliciously modifying the protocol (e.g. censoring transactions that don't sign a KYC/666 id number).

As I stated in the previous post that if mining is unprofitable then requiring some PoW with each transaction solves the "unbounded transactions spam versus centralized oligarchy mining control tradeoff" pointed out by ArticMine which in Satoshi's design requires contention over the block size choice which thus reduces Satoshi's design to a centralized politics because there is no block size which is ideal.

There are at least three design challenges introduced:

The PoW only has to be recomputed if the prior block on which the PoW share was computed is orphaned.
The delegates could provide the PoW (compute it more efficiently employing an ASIC) in exchange for a fee.
The difficulty has to be adjusted so that the block period remains constant.

If the PoW share difficulty is constant and chosen such that it requires at most only a millisecond to compute on most client computers, then the 100s of milliseconds round trip latency cost of paying a delegate server to compute the PoW is more costly to the user than just computing the PoW share locally on his client. The user doesn't care about the increased electricity cost for computing the PoW on less efficient hardware because the cost is miniscule compared to the value of the transaction.

The difficulty can be adjusted as it always is in Satoshi's design as this is orthogonal to the amount of difficulty required for each PoW share.

Thus so far my idea has passed the initial process of searching for a flaw. This is promising.

[2]

Quote from: TPTB_need_war on January 15, 2016, 03:59:09 PM

Quote from: monsterer on January 15, 2016, 03:35:13 PM

Quote from: TPTB_need_war on January 15, 2016, 03:16:19 PM

I think you missed my point, which is that if the PoW can be outsourced, then it is always more economical to do it on an ASIC farm near to cheap hydropower. Thus the mining PoW in Iota will trend to professional miners same as for Satoshi's design! They haven't accomplished anything at all on economic centralization (but have accomplished scaling but with other critical flaw I asserted upthread) because the economics of mining remains centralizing!

Note as we transition to mobile, battery life is important to users. Not needing to waste battery life on frequent microtransaction PoW computations would be desirable. Even IoT devices are often necessarily low powered. My design will face the same economic reality, but as I said if the latency cost is much higher than the cost to compute the PoW locally, then my design principle holds because in my decentralized, permissionless design the payer signs the PoW thus to outsource it would require a round-trip network communication (and wireless network communications consume battery life also).

It will only centralise like that if it becomes profitable enough for any such ASIC farm to do that work, compared to the reward they might achieve mining a coin with a block reward with all that expensive hardware.

That is an incorrect conceptualization of the economics. The relative profitability will drive the fees extracted from payers such there is no difference between the two scenarios you contemplated. Thus mining in Iota will centralize the same as for Bitcoin.