Comment by jacquesm | Hacker News Reader

Comment by jacquesm | Hacker News Reader

Parent Full thread

Bitcoin transaction volume is up tremendously but that is in large part due to 'satoshidice.com'. The reason why they're so large is because you can't really do all that much with bitcoins (yet, this is improving though), so people will gamble their bitcoins away.

As an experiment I put a bitcoin donation link on reocities.com customer pages to see if anybody would donate but the counter is solidly stuck at '0'. I figured if the general public starts to see bitcoin as viable then I'll see some donations there, there have been offers enough of people that want to donate money, bandwidth or server capacity to the project in the past.

The bitcoin ecosystem is tremendously interesting, both from a game-theory perspective as well as an economical one (and technical too). Within a short time we've seen CPU, GPU and FPGA miners to go to insane hashrates with a generation of ASIC based hashers about to turn the previous generation of FPGA/GPU units into expensive doorstops.

The guys behind bitcoin (who I think are related to Michael Peirce and Donal O'Mahony somehow [1]) have done their homework very well and the fact that bitcoin has survived relatively unscathed over the first couple of years has definitely made me shift from 'fad' to 'possible'.

This likely won't be our last crypto currency but it definitely has life and merits serious scrutiny. The bitcoin economy is now already very substantial and it will likely grow enormously in the coming year.

[1] http://www.w3.org/Conferences/WWW4/Papers/228/

Personally I think the inventor of Bitcoin is Nick Szabo. In a blog post predating Bitcoin, he described all its aspects, and even called his invention "bit gold"(!): http://unenumerated.blogspot.com/2005/12/bit-gold.html He also seems to have tried to postdate his blog post (URL says 2005, but date in the post says 2008) to perhaps hide the fact he was the first one to come up with these concepts :D

He was even a contractor for Digicash back in 1994-95.

Nothing he describes there is terribly remarkable - all those ideas were widespread. His description lacks the key innovation that made Bitcoin better than its prececessors, which is that it doesn't rely on trusted centralized timestamping servers.

Szabo's bit gold ideas, or Bitcoin's ideas, when taken individually, are not remarkable.

It is the combination of all of them, that truly gives the whole distributed currency its elegance. And I would argue that Szabo is the first one to put almost all of Bitcoin's ideas together, in writing. Yes bit gold's proof-of-work takes a trusted timestamp as input, but simply changing it to take a head of the "chain of title" as input is really the only (minor) difference between bit gold and Bitcoin.

Here are his own comments on bitcoin: http://unenumerated.blogspot.com/2011/05/bitcoin-what-took-y...

I vaguely remember he also denied being Satoshi Nakamoto somewhere - but I cannot find it now.

Interesting! One thing I'm pretty sure of is that it isn't a single person.

I think it is quite probable that Bitcoin was written by a single person. It is not that complicated: block chain of merkle trees + signing transactions with public/private crypto + proof-of-work with hashing function + peer-to-peer networking. The first public release (0.1.0) was only 20k lines of code. Someone like Szabo, Wei Dai, or Hal Finney, who have each spent more than a decade thinking about and attempting to design decentralized currencies certainly have the expertise to design Bitcoin all by themselves.

My admittedly vague feeling is there are so many balances in the system that it was likely reviewed by a larger group, having lengthy expert-level discussions in a non-public forum, before release.

I've heard the theory that it's a government think-tank project that in asome sense 'escaped', in that it was thrown out to see what happened and has thrived beyond expectations.

ASICs aren't on the market yet, but I'll be excited when I get one on my doorstop.

They've been delayed for months, and it seems this month will be the month they ship.

> and it seems this month will be the month they ship.

And the cows fly...

I heard that at Nokia it takes 3 years to design an ASIC chip. So these bitcoin companies should be able to it in a year.

There has been a lot of back-and-forth over this, I've done my own research and I think that it is possible to make some degree of ASIC in about a year.

Specifically this would be something called a 'structured' ASIC. If you get it right the first time (always a big question) then under a year is doable. A structured ASIC isn't nearly as efficient as a real special purpose design, it will consume more power and will fall speedwise somewhere between an FPGA and a true custom ASIC, probably closer to the FPGA end.

There are currently three companies that claim to be working on ASICs, only one of those three I think has the chops to pull it off.

But anybody that paid their money up-front to these companies without a shipping product is really slightly out of their mind.

All of the companies have insisted that they're not selling structured ASICs, though.

Designing an ASIC that only does SHA256 really fast should be a lot faster than an ASIC that does more than one thing, right?

Isn't the idea of ASIC's in general that they are only supposed to do one thing?

https://en.wikipedia.org/wiki/ASIC "As feature sizes have shrunk and design tools improved over the years, the maximum complexity (and hence functionality) possible in an ASIC has grown from 5,000 gates to over 100 million. Modern ASICs often include entire microprocessors, memory blocks including ROM, RAM, EEPROM, Flash and other large building blocks. Such an ASIC is often termed a SoC (system-on-chip)."

Reply on news.ycombinator.com