Bitcoin book draft: first unlicked lump of words sent to betas.

[reddragdiva]

Went down pub last night with a laptop with a book on it. Much help right there and then. I can see people getting into this.

I have the unlicked lump of words ready to send out to beta readers. If you would like to participate, please email me, dgerard@gmail.com, and I’ll send you a link to the .odt and .docx of the rough drafts.

(This is not the prerelease review version, it's the "how am I doing" version ... we'll see how it goes.)

There's more excerpts up if you want tasters: http://davidgerard.co.uk/blockchain/

Comments

[reddragdiva]

Fabulous, fabulous!

There's also bits where I talk about trading as if I have any idea what I'm on about. Also, I'd be interested what you think of the smart contracts and business blockchain chapters in convincing people these things are fucking terrible ideas.

(I realise as I type these words that (a) we have something very like smart contracts in high-frequency trading - that is, rogue AIs barely under human supervision (2) this is where flash crashes come from.)

[hairyears]

Where did you talk about trading? I missed that.

There's no analogy to be found for Smart Contracts in 'Flash Crash' events in the markets.

Smart contracts are irrelevant to algorithmic trading, even if you're only interested in finding cautionary examples from the cases where automated trading goes horribly wrong. Interacting algos are entertaining - look for cases on Amazon of bots bidding each other up - but you'll need to find your own examples of this happening with 'Smart Contracts': the world of banking steers well clear of this particular technology, for reasons you have pointed out quite well.

Blockchain technologies *will* get used in trading systems, although your point about open chains allowing full visibility of the transaction history is, as you point out, a major drawback.

Nevertheless, there are applications where transparency is advantageous, and many use-cases where a hash-verifiable (but not readable) history in a private chain is *exactly* what we need.

These cases are roll-backs and trade fails, and regulatory investigations of fraud and money laundering.

"No laundry tag, no trade" may well be the killer application. This is, of course, antithetical to the principles and purposes of a cryptocurrency.

It is immensely useful to have all the records to hand - or better still, intrinsic to the trade itself - if it is ever necessary to roll back ownership (or recall a pledged asset) when the settlement process fails - or when a counterparty defaults, and all their option trades are immediately closed out. This is going to make stock-lending and repo trades much, much more reliable and efficient.

If it actually works, and this hasn't yet been demonstrated.

You have pointed out that roll-backs are not supported by BitCoin.


So Blockchains have a use in Banking: but fewer uses than are being advertised, and very different uses to those of a cryptocurrency.

[reddragdiva]

That's more uses than I thought blockchains had.

I rambled a bit about wtf Bitfinex was doing, but if nothing triggered your Gladwell detector, I'll count it as a win :-D