Reggie Middleton v Paul Krugman pt 2, Inflation, Bitcoin and the Alt.coins

Here is the next segment in the presentations given at the North American Bitcoin Conference in Miami Beach.

Comments

[madtechnician]

"The hashing rate has nothing to do with transaction speed."

Okay - but please can you provide a source for that information because as insofar as I can understand the software source code a percentage part of the hashing process is actually allocated to verifying transactions , it is not just creating random hashes to solve a new block checksum - the mining process itself actually verifies transactions.

According to the source code once all coins have been 'mined' there will be no random hashing at all , 'mining' will be just pure transaction verification.

The percentage of 'mining' divides by half every 4 years until it reaches zero.

[Matt]

I'll see if I can find anything, but essentially solving the block is how those transactions in that block are added to the blockchain (public ledger). 

If the sheer volume of transactions increased to the point that it slowed down the average time it took to solve blocks, then the difficulty would automatically go lower so the blocks were easier to solve, so that the average time trended back to the target time - in the case of bitcoin, every 10 minutes.

If the transaction fees grow, or desire for anonymity grows, then people could simply have their wallets with a service, and that service would record the transactions, and then bundle a bunch of the transactions together and send them as one transaction on the bitcoin network, which lowers the fees for everyone and hides the personal information from the general public, while still being available for law enforcement, like the woman on the public panel said.

[madtechnician]

As far as I can understand it "solving a block" is creating a random checksum which matches the checksum produced by the transactions broadcast by the network and received by a miner in the last 10 minutes or so. It is not unfeasable to scale this up so there would be no reason to have seperate blockchains or "off block" transactions. If you have ever worked with RF systems you could easily see that bitcoin is in fact simply a high speed serial data bitstream with regular 10 minute checksums , with each "block" being a 10 minute segment of that bitstream. This could theoretically operate at gigahertz speeds I will let your imagination decide what is possible here - my imagination says this technology could easily operate entire stock markets ,.   So anyway , maybe I am preaching to the choir ?

[Matt]

The problem is that the blockchain contains a record of every single transaction that has ever occured. Also, it tends to grow exponentially. That does not mean a mechanism for trimming off old transactions cannot be used.

If you had HFT algos trading stocks on a blockchain, the problem is that the chain could grow by terabytes each day. As the amount and complexity of hardware and IT services gets more complicated, you end up with a more and more centralized system.

Each transaction takes up about the same amount of data, whether it is a transactions for $0.10 or $10 million, so since there is a cost in transmission and storage of data, there needs to be things in place to disincentivize excessive quanitities of tiny transactions that cause bloat.

We'll see what Reggie's Derivitives-on-a-blockchain ultra coin does to handle the technical challenges.

[madtechnician]

I don't think Reggies Ultra Coin app. will have anything at all to do with high frequency trading.. but maybe I am wrong...

But - There is no reason to suggest that every single transaction within a stock market needs to be recorded over the blockchain , in the case of HFT , the blockchan is only required to hold a checksum for each segment of HFT ..

[Matt]

How would a pure Proof of Stake coin work? Only people who already have money, can print more money, so how would anyone get the units in the first place, in order to produce more?

There are a lot of interesting ideas for making true alternative coins (as opposed to coins identical to Bitcoin or Litecoin but with slightly different variables) but most of them turn out to have significant flaws once you bounce the ideas around a bit.

[tip e. canoe]

most of them turn out to have significant flaws once you bounce the ideas around a bit.

well, one could say the same about bitcoin ;~)   good questions on PoS.  NXT is an attempt at a pure PoS protocol, so it will be interesting it develop (or if not, why not).

personally, methinks a PoW/PoS hybrid that limits processing power to simple home machines is most promising.   but how exactly, i have no idea.   but i don't think anyone else does either...yet.

but that's what bouncing ideas around is for, yes?

[Matt]

There are Proof of Work / Proof of Stake coins that use cryptography that is designed to make it hard / impossible to use on GPU, thus requiring CPU work only. 

Some coins also have it so that the higher the difficulty, the lower the block reward, as a way to disincentivize people from trying to have an arms race to mine the most, in an effort to keep power consumption under control - PeerCoin works that way, I think.

There is probably an alt-coin that combines these features, but with hundreds out there, who can keep track of them all?

[tip e. canoe]

yeah, at a glance, peercoin looks closest to what i'm talking about.     not a bad brandname either, which is definitely an asset in the Age of Simulation.

[smlbizman]

u need to ask fonestar...im sure he has all ur answers...

[disabledvet]

in theory you could finance a war effort on bitcoin actually.
"everything else you'll have to get on the battlefield."

I know I wouldn't want to have been on the receiving end of one of Napoleon's "encampments"...let alone one of his campaigns.

Some of those towns...are still towns today.