The reason people get BTC for mining is because mining is what secures the blockchain.  Assigning rewards for computational power that secures the network is a really elegant solution that works and that cannot be faked (if you find a valid hash, that means you did do the work).  On the other hand, as you point out, it'd be easy to fake how many clients you have open or what IP addresses you can connect to the network on.  Heck, there are sysadmins at large corporations with entire spare B blocks to throw around (that's 254*254 IP addresses I believe).  They could make a few routing rules, and bam, now they look like they're 64,516 people.  And that's with one B block.  And just forget IPv6.  Absolutely forget it.

Besides, why would we reward having clients open?  Having clients open doesn't really help the network all that much.  Yeah, they relay transactions, but that's not exactly a huge burden.  People have spent thousands of dollars on mining rigs because of the rewards for mining, and all of that money has directly translated into the security of the network.  Would people have to go out of their way at all just to run a client?  Nope.  And does it secure the network?  Not really.

Well that wouldn't work (for the obvious reason that you point out).  This is actually a really tricky issue, and the option Bitcoin chose was the best out of bad alternatives.  Bitcoin needs to be handed out on a basis that is impossible to fake; you can't fake the hashes, but you can equally make as many new receiving addresses as you'd like.  Plus there is a rational reason why we assign BTC to miners, because their computational power is providing security for the block chain.  People generating lots of random addresses wouldn't be providing anyone anything.  If we restarted the blockchain today, at least, there'd be much more interest in it from the very beginning, so the wealth disparity would be a lot less.  I think that's the best we could do.


There's no reason in principle this couldn't be done.  It would certainly counter the claims that Bitcoin has a built-in deflationary spiral.


So basically you would eliminate the difficulty increases.  Thus you would get insane long term inflation roughly on the scale of the size of the growth of the network (around 5% per day right now) times the growth in hashing power due to better hardware (if we use Moore's Law, say it doubles every two years).  Thus, no one would ever use Bitcoin, because it would be losing value due to inflation at obscene levels; really, you could only do worse by using the Zimbabwean dollar.

Plus, how would you assign bitcoins on a basis proportional to hashing power?  Can you come up with an algorithm that would actually manage to do this in a distributed manner?  Keep in mind you'd need to keep the number of blocks generated relatively steady at one per ten minutes on average (the entire system falls apart if you're generating a block, say, every second; most of your time mining would be wasted on blocks that were already invalid, you just didn't know it yet).  So the only way to do it would be to keep the number of blocks being generated steady, but scale the BTC reward per block appropriately.  I guess you could build into the 2000 block window readjustment algorithm that the block generation reward would be multiplied by the same factor as the difficulty level.  So, for instance, this last increase in difficulty from 244K to 434K would've also increased the block generation reward for each block by 78%, up to 89 BTC per block.  I don't like this though.

It's simpler than that.  The data on the card consists of the private key(s) that have the ability to sign transactions spending certain bitcoins.  So, assuming you realize the card has been stolen before all of those bitcoins are transferred elsewhere by the thief, you simply use your home copy of your wallet.dat to transfer those coins to brand new addresses whose associated private keys are not on the lost card.  You'd never have your only copy of private keys for addresses of bitcoins under your control be on the card; you'd always have that info at home too.

Actually, Bitcoins are every bit as universal as credit cards.  Both require network access.


I wouldn't do it with a credit card, I'd do it with a smart card.  That way you could encrypt the data on the card and use a point-of-sale terminal to type in a password to unlock it.  That way, if the card is lost, you aren't automatically out all of your money.  And the way it would actually work is that your smart card would contain some of your private keys (with access to reasonable amounts of your balance; not all of it).  The card and/or terminal (I haven't worked out all the detailed technical aspects) would use the private key(s) on the card to sign a transaction, which the terminal would then send along to the Bitcoin network.  So the terminal needs to be "smart", i.e. it is a node on the Bitcoin network or is in communication with a trusted node on the Bitcoin network (i.e. you could have 20 terminals in a supermarket, all of which use a single Bitcoin node running in the manager's office).  But the card can be "dumb".

Alkor, I think you missed the point of Bitcoin.  Its biggest advantage is that it is decentralized.  Having expensive proof-of-works is simply a side-effect of having a distributed system.  No one's interested in yet another centralized system; there are already hundreds of those, many of which have much larger backing and more trust than your idea ever will.

Going to try to get the Wikimedia Foundation to accept donations in Bitcoins.

Thanks to black8, who graciously admitted defeat and sent the payment: http://blockexplorer.com/tx/caccaad9800a8ffc2284d7986904b30c5e37af0aa08890bf449853803d2b4665

Now, for my next bet ... 

No.  The difficulty level rises to compensate for increased mining, so that the number of bitcoins entering the economy is stable.  Only if the difficulty level didn't adjust would you see an effect on the price.

Yup, right around 438,000.  A lot of people who bought new mining rigs recently and who forgot to take exponentially increasing difficulty into account are going to be losing lots of money.

Everything you're looking for, including a complete analysis, can be found here:  http://forum.bitcoin.org/?topic=7427.0

I should've taken that bet with you Kirian, then I would've made another 1 BTC 

I just fed the bot (chomp chomp )

Yeah, so the thing about that chart is, the axes should automatically scale to fit the data.  It looks like the chart was just configured manually and then forgotten about.  So we'll call it bad programming 

Well yes, that's what it is.  But it adds some flavor to see the reasoning behind the derivation of the 430,000 figure as well.

I'll let you guys in on my original reasoning.  That 500K figure comes from my reckoning as to what the instantaneous difficulty level would be at block 127007.  Yeah, I plum forgot about figuring the 2000-block sliding window into my number, which obviously makes the calculated difficulty a lot less than the instantaneous difficulty at the time of switch-over.  The crazy thing is the instantaneous difficulty level is actually probably going to be above 600K at the time of block 127007 (it's at 520K now and rising).  That means it'll easily be way above 600K for the increase after this next one.  Holy crap!  Here's the chart to prove it:  http://bitcoin.sipa.be/

But between me forgetting to figure in the sliding window, and Black8's bet being way too optimistic, the middle ground between the two has actually resulted in a highly competitive horse race.  It's still too close to say with much certainty which side of 430,000 we're going to see -- and that makes it a good and entertaining wager!


Kirian and BCPokey, you may commence the ingestion of humble pie    Oh, and Kirian, the block solving rate is at 12.83 and climbing right now.  I expect it to see at least 14 before the next difficulty level, which means your numbers are low, and hopefully the difficulty level at the switchover will not be 423,000, but rather, just a tad bit above 430,000.

All the people still assembling mining rigs today need to see this data.

That's not necessarily true.  If you're mining on hardware you already have then it's still quite profitable.  Is it profitable to put together entire new systems for the sole purpose of mining?  No.

Well this is a no-brainer -- for significantly less cost, the 5850 has roughly equivalent performance.  That is if you only care about mining.  If you're also a gamer, the 6950 is the more attractive card.  Keep in mind that, with the currently exponentially rising difficulty level, you are only going to get a few months of worthwhile mining out of this card before it's not even worth running anymore, so what are you going to do with the card after that?

Gaming would be my answer.

The one change I would make is to have pools not incorporate transactions with 0 fees on them.  This wouldn't hurt the pools' payouts because the pools themselves are solving enough blocks that they could include their own transactions in their solved blocks for free, but they wouldn't take anyone else's for free.  Getting people used to fees now is better for the network in the long run.  Note that I'm not suggesting .01 as a minimum fee; that's probably too large.  But it needs to be something.

The estimated difficulty level is hovering around 400,000 right now with over two days to grow.  I can win this still.  It's going to be close!