Topic: DECENTRALIZED crypto currency (including Bitcoin) is a delusion (any solutions?) - page 59. (Read 91144 times)

Can't argue about this because we failed to come to a common denominator in definitions, particularly of Consistency.

Fuserleer,

This is a design afair I hadn't contemplated until I wrote this. Perhaps this is what you mean by allow Partitions and then letting any user spend on any Partition? But this is essentially a partitioning of access to coins, because there is no cross-spending between forks below...

In the case of write-only database, the inconsistency can always be merged. I was stating that in general due to double-spends, paritions can't be merged. Indeed in a read/write database where clients of the database base their data writes on the state of the database, then inconsistency infects irreversibly same as for double-spends.

My point is that cases of relaxed CAP implications that apply to one type of network do not apply to our case which is strictly inviolable.

For example that article says:


Which is inapplicable to the inviolable point the double-spends in multiple partitions can't be merged. This ends up infecting every possible design you can think of to temporarily enable partitions, unless you accept reversibility of transactions (tolerance for inconsistency) or give up Access during that temporal partitioning. CAP is strict on our situation.

Really?  You want to stick with that train of thought?

Aside from the fact you seem to be conflating the issue of double-spending to a single problem, when it is in fact two issues, a double-spend is akin to many issues that a database has to deal with and is VERY relevant.

Assume we have 3 machines with a copy of a database, and this database contains the state of some object.   The state of some object is A in all 3.  Then, one of these machines sets the state of the object to B,  and another sets the state to Z.  The remaining machine A now has to decide which of the 3 states are correct, A, B or Z.  The database is in a partitioned state, which is exactly the same as what an internal double-spend results in, a conflict of state.

In crypto, A is the existing state, B is a spend to one place and Z is a spend to another.  The machines with state A then have to decide whether to ignore the new conflicting 2 spends entirely and stick with A, or pick between one of the spends B or Z.

In the case of a database, the choice of selecting A, B or Z effects everything irreversibly downstream also, just like it does here, they are the same thing.

Blockstream's Side Chains indeed would cause CAP to apply interchain. This is why smooth and others (including eventually myself) thought Side Chains are going to be intractable.

Exchanges don't link Consistency interchain because the risk of double-spend falls on the counter parties each independently w.r.t. each chain.

If you think of alternative currencies as providers of information on the state of fiat currencies,
then exchanges provide a relationship (information exchange) between partitions. You are
correct in that there is no "formal" transmission mechanism, the question is whether the
CAP theorem renders such a link impossible.

Edit: The question perhaps should be stated as how important could such a mechanism be?

The 1st sentence is correct. The 2nd sentence is incorrect because CAP is inapplicable to a relationship between two block chains. There is no double-spending (Consistency to maintain) between chains.

Actually I explained upthread that crypto currency can theoretically tolerate rare partitioning, but in the case where these are no longer Partitions but treated as orthogonal block chain forks. The Consistency is recovered by allowing everyone to double spend on both forks. However willy-nilly forking would be incredibly chaotic to manage. I suppose it is possible to design clients that could interopt with an unbounded number of forks. There are many issues, such as spending on multiple chains and securing multiple PoW chains.

From my reading of the CAP Theorem, Bitcoin has consistency and availability but cannot be partitioned.
Hence cryptocurrencies, as a group, have inherent partitions, availability, but no consistency as a group.

That article about databases doesn't have to contend with Consistency of double-spends which infect everything irreversibly downstream. This added wrinkle for cypto currency is what makes CAP more strictly inviolable on us, i.e. we can't in general recover Consistency from multiple partitions.

Ive read that many times, and the first time I read it was AFTER I'd formed the basis of theory from which channels operate (believe it or not).

But you are right that is seems that many haven't.  Its a very important article of observation.

I used this and noticed that http://www.infoq.com/articles/cap-twelve-years-later-how-the-rules-have-changed appears on the 2nd page which is read by noone...

My biggest mistakes have happened when the obvious question went unasked.

Thanks for the reply, though.

No offense, but that question is akin to walking into the Large Hadron Collider complex to ask "Where's the bathroom?"   

Relevant Info:
http://lmgtfy.com/?q=CAP+theorem

"You must use PoW or PoS and thus the problems of centralization and government control are the end result of all of what we are doing here. And helping to force a world government cooperation of regulating the internet, encryption, and Bitcoin mining. Bitcoin is a Trojan Horse that weakens the nation-states and traditional banks, which must fight back by cooperating with a world governance regulation of the internet and Bitcoin mining."

Disagree - once inequality passes a critical limit revolution is inevitable. While some would prefer
a world government, it is not an optimal solution, indeed, there may not be a single optimal solution.

With regard to cryptocurrencies, I doubt that a single cryptocurrency ie bitcoin, can provide an optimal
solution to the paradox of thrift, and for that reason alone other currencies must exist.

Before taking this further, what is the CAP theorem referenced earlier?

You guys should take a look at SpreadCoin- the truly decentralized cryptocurrency (no pools, prevents centralization of hashpower, etc).

https://bitcointalksearch.org/topic/ann-spreadcoin-decentralize-everything-decentralized-blockexplorer-coming-1045373

Mining rewards can still be paid to PoW and it can still be unprofitable if the level of difficulty rises too high because every transaction is including a PoW share. Transactions just keep growing and growing while the debasement rate is a constant percentage. If the value of the currency grows as fast as the transaction rate, well then I guess that is a great problem to have.  

If mining farms are dumping their coins while the price is rising that fast, they are leaving all that money on the table for us.   Meaning the banksters are shrinking their relative net worth.

I assume you are referring to ongoing (a.k.a. tail) emission. For initial distribution, then profitability of mining appears to not be a flaw. I would argue that distribution to HODLers is a market failure plan though. I'd prefer to initially distribute to users than investors. Let the investors buy from the users in secondary markets such as your metaexchange. I would prefer if earliest investors have to work even harder to get coins early by creating businesses which accept the coins from the users. So the early investors add even more value to the ecosystem. As I had explained to you in my vaporcoin thread, I think users would have no idea what an exchange is and how to access it, thus earliest investors would have to work harder to source coins in volume. And that IMO is the way it should be. It shouldn't be so easy to invest as an early adopter, that all you need to do is buy an IPO. There is no capital created from such an activity, unless you can argue that these investors work on improving your coin (but they seem to mostly just clog up the forums with shrilling) or they are funding your coding (but that would normally come from angel investors not an IPO in a non-pump&dump). At least with Kickstarter, your customers who are your early investors actually help you design and refine your product.

Think out of the box.

Okay let's get back on track of trying to determine what is the best design (in terms of best of the imperfect possibilities) and enumerating the tradeoffs and flaws.

I think we all just need to accept that there's a high possibility that there is going to have to be some compromise in the system.

We are all striving for perfection, which is commendable of course, but no one knows if there is a solution to achieve the perfection we all desire.

Just accept the compromise, but take solace in the fact that why it might not solve ALL problems, future generations of this technology will likely be an improvement over Bitcoin and clones by some degree.

Going back to my cancer statement, surely its better to provide a solution that prolongs the life of someone by 10 years, if an outright cure is not available.

The same is true here, Bitcoin is the first attempt to cure a cancer, and while it might not be an outright cure, its inspired and motivated many to also start looking for that cure.  Many will follow that are not outright cures, but does that make them entirely worthless?  Of course not.

Those solutions are not (as far as we know now) limited by physics. Thus they may be worth continued effort to solve. Crypto has certain limitations imposed by entropy (Second Law of Thermodynamics specifically) which are inviolable. Being able to state this taxonomy is an example of a highly abstract conceptualization.

For example, zero knowledge proofs can't be done without a trap door. Thus they can never be information theoretic security. There are many other examples of inviolable truths.

Knowing when the end of the inviolable rope has been reached is a high IQ trait. If my IQ was 160 I would have perhaps sorted it out a year ago. (or maybe if I wasn't so ill doing much of the past 3.5 years)

Also I played for example football (the one with the helmet) and other very intense sports. So I also attained a demeanor in life that bullshit walks and performance talks. I am hard on myself this way too. But when I am in relax mode, I am very amiable, joker, smiling to everyone, etc.. I have two modes. Even my gf says she feels terror from eyebrows when I am in "compete and get shit done" serious mode. Sorry. I prefer to do that alone.

Getting this topic back on track, I have a question:

* If mining rewards cause centralisation, how do you handle distribution fairly in a system with no mining rewards?