Topic: rpietila Altcoin Observer - page 105. (Read 387493 times)

There is no fork.

Since pruning does not change the block hash, anyone can just go ahead and unilaterally prune blocks and either store them locally or serve them up to other nodes that request them. The blocks still form a valid chain, and although changes to node logic are obviously needed, permission from pools is not.

A node that receives pruned blocks can still verify transactions and mine on top of a pruned chain. Blocks mined this way are indistinguishable from blocks mined on top of an unpruned chain, so pools can't discriminate against these blocks. Of course concentrated pools can shut out anyone not part of their cartel, but that is independent of pruning.

The efficiency advantages of the mini-blockchain relative to Bitcoin with pruning are quite small, certainly not order of magnitude. Given this I seriously doubt a mini-blockchain can defeat Bitcoin without other advantages.

Two final points and I am outta here for a while...

1. The pruning in Bitcoin is technically doable (and Satoshi very vaguely mentioned it in the seminal whitepaper), not likely as efficient as the MBC, but I wonder if the pools will even allow it to happen. They may have an incentive to make it more difficult to run a full client. We debated upthread whether users can solely dictate hard forks, or whether mining inertia is also in control of that decision. (so I won't rehash that discussion)

2. There are so many of you out there who don't have enough income and have more free time, I don't understand why many of you wouldn't be spending $200 per day on a 10,000 computer botnet and mining these CPU coins in order to generate more capital to invest in Bitcoin or which ever coin you think is the best investment. The link I shared upthread says the mining GUIs are "point and click" in the botnets, so probably even a dummy could do it. Assuming it is profitable, the only issue is legality and morality. I personally think it is immoral to not expose a natural weakness (both in the crypto-coin and on users' computers), because nature requires we do in order to maximize resilience and prosperity overall. Note I didn't comment on the legality. For example, maybe more users would move away from that spyware Windows 8 and to Linux if more users are seeing their computers slow down:

http://en.wikipedia.org/wiki/Usage_share_of_operating_systems#Desktop_and_laptop_computers

Even though other nefarious uses of a botnet can potentially generate more profit, mining is just so much easier and (at least not factoring the theft of other users' computation, the mining aspect is, versus say stealing credit cards with the botnet) less criminal.

Thus I see these CPU coins with near-term high debasement rates and no botnet resistance as sitting ducks waiting for their price to be in a downward spiral. Some whale could end up with too many coins by sitting on the bid (as the botnet miners dump coins), and end up killing the network effects and thus his investment. Look Ma, I own all the currency. Ma replies, "Aren't you a little bit lonesome?".

Edit: the criminal liability is more than anyone with reasonably good opportunity cost would stomach, but for those who have less alternative opportunities and or who under financial distress (or those with a criminal mindset who don't calculate probabilities well), they would be probably wisest to botnet mine an anonymous coin where they have better chance of concealing their coins.

Which means they can't attain the any where near the widespread distribution and network effects that a mined coin can (especially if that mined coin has a design that defeats botnets). Their distribution is tied up amongst who ever was stupid enough to buy their IPO, which can't expand after the fact as popularity rises, thus popularity can't rise organically exponentially (geometrically) through spiraling network effects.

Yeah I think I agree. Ethereum appears to be mathematically DOA.

Apologies for the multiple posts (my thoughts weren't entirely organized before as I am multitasking and distracted).

I didn't know MBC suggested that but rpietila knows I suggested it in a private message Note he has no inside knowledge of anything I am doing. I just mentioned that to him.

But the problem is how to set the demurrage when you don't know the value of the currency in the future?

Yes if there are straggler unspent transactions in old blocks then there will be some merkle branch hashes that need to be kept. Once all the transactions in an old block are spent that goes away and all you have is the header, which is exactly the same as MBC.

This may seem implausible if there are many transactions per block. Except that to keep the ledger from filling with dust and abandoned outputs, MBC suggests (though Cryptonite does not yet implement) some per-account demurrage (idle balances eventually decay to zero and can be pruned). If you do the same thing in bitcoin then you can clean up the old blocks as well. In the end they seem comparable.

Ethereum basically sold the first 4 years of outright with the genesis sale.  On top of that added .18 of genesis block that was sold for themselves + bounties. The 0.18 may be funding the privileged position.

Perhaps there are also issues around efficiency of how the history is stored with multiple Merkle trees (one per block) verses one Account Tree. As far as I remember not major (not thinking deeply today on this), probably a constant factor.

It is more than plausible, it is explained in the original paper. On the surface less efficient if addresses are reused (which is often not recommended), but not dramatically so. Both end up with block headers plus UTXO.

Cryptonite has some other nice features though. The 10 year reward halving schedule is unusual, perhaps valuable, and will never be adopted by Bitcoin. At least it pushes very-low-reward issues out several decades.

I said they implemented the mini-block chain which is a technology I like, but I never said the Cryptonite (not Cryptonote) had compelling a feature set that made it a wise investment.

One feature wonder coins are not necessary sufficient to sustain against the network effects of Bitcoin. It is up to you all to analyze that further.

Pruning of the blockchain is apparently technically plausible for Bitcoin and some apparently effort has been applied investigating that. Will probably be less efficient than the mini blockchain.

Intra-block period true, but interblock the distribution isn't even Poisson (??) because the difficulty is a not an independent variable.

And one day he starves because he couldn't get any fish for a full month. Example: Bitcoin once had a 6h break between two blocks, I had a transaction in wait for 9 hours because of that. That's like the man not eating 36 days instead of eating every day.

 

So how long until FPGA?

That didn't take long.  GPU miner released before wolf0 and I could even start annoying people by talking about how much we'd sped up the CPU version.

https://bitcointalksearch.org/topic/m.8212829

give a man a fish and he eats for a day.
give a man a poisson distribution and he eats at a known average rate at times distributed independently of one another.

Botnets aren't as profitable to use for mining as if they are used for other nefarious activities, but that doesn't mean it isn't profitable and isn't popular to mine with them.

But there is way to design your GPU-resistant (initially no ASICs) PoW to make botnet mining much less profitable. And I am not referring to making the memory footprint larger than the average botnet system.

Ethereum hasn't even worked out the key design issues. And I think they already made one (probably) fatal design choice. In theory they could back out of that choice—I'll be watching.

Edit: I just noticed they deleted my comment is still stuck in moderation at the following blog, pointing out their math error wherein they assumed the distribution of PoW solutions is a uniform distribution rather than the Poisson distribution it is. It wasn't a fatal error to the point of that blog, but that they deleted my comment speaks volumes about their future.

Moreover they have potential competitors which already realized to some extent few of the declared features which also are going well in advance of the current regulatory environment and nobody can really predict its future utility for society. The ideas may sound revolutionary, but they are not mature yet.

And in addition to all of this, etherium seems to have no real obligations to ether holders even if their grand startup will ever succeed. The snake oil of the highest grade.

This has been seen before in many sectors. Take energy, for example. Medicine. Nuclear physics. Climate "science". What I mean is that the bright minds are harnessed to work together, towards some goal that is not achievable, not made achievable, is not important, or even true. This becomes the "only supported way" to work in that field. So real science, with the base in reality, is almost ground to a standstill. World has functioning water car - but the inventor is murdered instead of acclaimed. Great majority of medicines are useless and dangerous. Yet the alternative cures are suppressed. Climate hogwash still cannot be officially criticized, even though long ago it was proven to be a lie. Made up lie, for the purpose of legistlating more control and human rights violations. The list could go on and on.

What people need is a coin that is easy to use. You cannot, just cannot lose your keys no matter what. Still, they are very hard for thieves to steal. And it must be anonymous, because that is what Bitcoin is lacking. It does not need to have any features. Gold has attributes, but no features.

Ethereum is sucking talent to sit on fat cushions, to be fed by printing-money lords. People that could make a difference if they were not allured by monthly salary.

As with Facebook, if the dumb capital is left unchallenged for too long, it can find ways to use influence and size to take control of a market.

Looks to me that Friendster was paid off to kill itself. Filipinos dominated friendster 2006 to 2008ish, then Friendster started purposely making their system so slow that filipinos migrated to facebook.

P.S. it won't go unchallenged.