[Cryptography] What has Bitcoin achieved?

[Lodewijk andré de la porte]

2014-06-20 18:27 GMT+02:00 ianG <iang at iang.org>:

> On 20/06/2014 17:05 pm, Lodewijk andré de la porte wrote:
> > 2014-06-20 16:31 GMT+02:00 ianG <iang at iang.org <mailto:iang at iang.org>>:
> >
> >     Take a block.  Calculate the amount of money spent on the mining of
> that
> >     block, which you can use 25BTC as a proxy for.  Divide that by the
> >     number of transactions in the block.
> >
> >
> > I do not, at all, think that's fair.
>
>
> Fair?  What does that mean?  Just curious, do you mean, calculating a
> cost that way seems inapplicable, or do you mean that there is a notion
> of what would be 'fair' in a cost allocation sense?
>

I didn't use another word because they didn't cover the load.

(...)
> One could see things that way, sure.  I'm just trying to be pragmatic here:
>
> What does it cost to run the entire machine?
> How is that cost distributed?  Who pays, who is compensated?
>

I think you should also consider the entire machine that the banking world
consists off. I'm pretty sure that the machine is thusly complex and so
intertwined with the very method of accounting (money) that it is
absolutely unthinkable to create a fair estimation of the cost of a single
transaction, even whilst taking immense liberties. The whole financial
crisis with people speaking of "evaporating money" are a fantastic example
of the complexity of the issue.

The government, with it's many interferences, also contributes vastly to
the cost of maintaining the banking system. Every policeman (greatly,
especially in it's size and threateningness) contributes to the
maintainance of the law, which contributes to the maintanance of the
banking systems. What about the cost in lives over wars fought for the sake
of money?

It just cannot, absolutely cannot, be measured how big an expenditure we
make so that everything in our society is measurable and expressable in a
financial unit ! That's not to say it's not worth it, I dare say I don't
belive the USSR did well enough to compensate for western spending to their
detriment (interesting topic to research but I didn't). And I also dare say
that nobody has tried any differently in the modern age.

Bitcoin, if succesfully implemented, will dramatically shift but absolutely
not retire all these expenses. In fact it may, through it's extreme
objectivity in accounting, be unable to smooth over the troubles of society
and thusly result in very far increased costs. But we gain things too. We
will have truth, liberty, and some crazy sort of objective justice, all of
which have been fought and died for, causing immeasurable costs.

So, simply put, think it's not fair at all to account Bitcoin transaction
costs like that.


> > And that last one is just as intended and a good feature. In a way you
> > can say that your fee buys you security, by buring your transaction
> > under a certain amount of hashes. Now there is money out of nowhere
> > paying for a whole load (maybe around 60x) more security than you paid
> > for. That's a good thing because Bitcoin is (used to be?) kind of small
> > and thus hackable, and is now (maybe around 60x) less hackable.
>
> Well, when we establish a cost, someone has to pay it.  Then, we can
> then build a marketing myth around why Joe-customer should pay that
> cost, as a participant, a consumer.  Idk that I give a lot of attention
> to such a myth.  "It's the security you deserve..."
>
> I think I'm more interested myself in comparing the costs of various
> machines and seeing where the market can be improved.  We can compare
> the cost of say the credit card machine (4%?) as against the Bitcoin
> machine (25BTC/block) as against the world wide slow SWIFT machine
> (20-60?) ... and so far we're not seeing a lot of difference.
>
> Which means happy days -- there is fantastic opportunity to improve the
> matter.  I reckon I can do it for pennies ;)
>
>
> > The question of who pays for that is peculiar as it seems in many ways
> > as if money comes out of nowhere. The complexity of supply and demand
> > does not nessarily change the value (although supply might correlate
> > with amount in existence) of the coins as they increase in quantity. I
> > would say that "newcomers" that push the price compensate for the
> > pulling of the price caused by selloffs through mining. The truth is
> > miners are purchasing new coins from nobody, not affecting supply/demand
> > directly and securing the network through it. It's a very good choice
> > for Bitcoin. So good that you'd almost like to make a coin stable versus
> > hashing power.
>
>
> Right, so the myth is that the money comes from nowhere, sure.  It's
> free.  Don't worry your pretty little head over it.
>
> The reality is that there is huge latent pent-up demand for open payment
> systems from the general public, and people are prepared to pay for
> that.  Currently the exchange rate seems to be say 5% which is pumping
> up a lot of exchange business, and the demand is overwhelming other
> factors such as the cost of mining.  So mining 'appears' to be free, but
> in fact it is only cheap while the system is growing and demand is
> pushing up the price so as to overwhelm the cost-transfer accounting.
>

I mostly agree with this. It is fully intentional that the network will pay
for the security of the network. Many say "it's still cheaper than banks".

NXT seems more promising, although less pure in some ways.

The question of consensus is a philosophical one. Is a mayority, full
agreement, agreement of involved parties, etc, consensus? We always appoint
either a person, a board, a committee, a bank, a government, anything or
anyone just to tell us what's right. We stick to exceedingly simple systems
for this, with central points of authority. Bitcoin cleverly distributed
the point of authority over the computingpower involved, fuzzing the line
between rule of the wealthy, rule of the public and democracy. Ultimately I
think NXT has the same ruling class, but encoded more honestly.

NXT is trying to be more than it has to be, however, which might cause
significant problems down the road.

Did I mention Bitcoin is actually just a decentralized ledger, and this
whole financial bullshit is actually a mere single implementation that
seemed like it'd be a good one? Mastercoin knows this, and abuses it for
what it really is. Namecoin showed it too, although differently.

NXT merges finance and consensus even further.

It's not really nice to have rich people determine what's right and what's
wrong. But I think voting per moneyunit controlled is actually immensely
fair in some ways.

It's a deeply tied knot of poltics, philosophy, economics and more. It's
not getting untied just like that.


> Well, if the system can be run at scale without the mining reward, e.g,
> when it switches over to fees-based only, then we can also hope that the
> cost will drop dramatically.
>

> But wait -- if so, why did we need the miners in the first place?  It
> was to stop the double spend, right?  Or the false spend, to include a
> wider degree of ills.  So as long as the miners are doing their job on
> this on a fee-basis then it is fine.
>

The security factor will decrease. The unrewritability of the ledger
depends fully upon the costs of writing it since the ledged thing was
introduced. You can bury information under a pile of cash. If you swap the
pile (regenerate the blockchain) you have to pay for it. So if the % of
money spent on mining decreases, the time until a certain certainty is
achieved increases (if mining costs decrease tenfold (miningcosts/10) then
instead of 3 blocks for X money you need to wait for 30 blocks). It'll
still work, just less well.

It now works ridiculously well. The bigger the network the smaller the
average transaction compared to the network, and how higher the "security
factor".


> (idk the answer to this, I've basically ignored fees and distribution of
> same to this point, what's the story?)
>

Obviously people have to pay the miners, most directly this can be done
through transactions. The added benefit is being able to tax annoying
actions, like "transaction dust" or hard-to-compute transactions.

It's also a thing that you have to create money out of nowhere in a limited
fasion. Blockrewards are one of those lightning-strike-genuis things where
you can distribute money and increase adaption in a single strike. They are
the only way money enters the system, and it enters more towards people
that help the network be more secure. It's very elegant.


> > I think you should add the cost of developing software for Bitcoin, if
> > you take initial coin distribution into account. And to think that the
> > price ends up being similar to a bank transfer!
>
>
> Ah, now you're speaking to my heart!  Developers are taken for a ride on
> this one.  There was a sort of proxy reward for the original developers
> in that those close to the code could mine most.  But that was so rough
> a connection that it is pretty well wiped out within months (I speculate).
>

I think there's still the fact that they're a most essential key in
businesses. Getting rewards for certain code features isn't so bad either.

Truth is, for most purposes code monkeys are laborers. Not unlike painters
in a certain era. There's some masterworkers, but most just learned how to
scribble in a way that entertains the masses... Masterworkers pick their
work, though. And they are not usually rewarded properly.


> Now we have the problem that the dev team is the SPF or CVP for the
> whole darn network.  And they're not funded at all, directly.


Indirectly was intended to be good enough. It's not like they'd starve if
they stopped working.


> This is why ripple and ethereum and others have started foundations before
> hand
> and pumped value into them -- so as to ensure the survival of the dev
> team.


To get rich or to make it possible?


> Meanwhile every startup is eating out the core of the dev team
> and bidding the developers up through the roof.  This does not end well.
>  Am I the only one who sees this???
>

It might end like HTML. Could be worse.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20140620/f6285092/attachment.html>