What that payment is for, in the long run, is securing the network.  

The miners contribute directly to security in a PoW network, but the network still isn't secured if there aren't a meaningfully large number of full nodes available.  

And so far, both PoW and PoS schemes have failed to broadly reward people for keeping a full node up.  Somebody who does that, deserves a full share of that payment regardless of their hashing power or capital invested.  

So, if you want a protocol to survive in the long run, I think it's vital to find a way to provide an incentive for people to keep nodes up - reliably, not just if they win some hugely biased lottery (that they have no chance at without million-dollar investments) and get a chance to form a block.  

I think a coinbase transaction every block with fifty outputs - one fairly-big one of maybe half the award for the miner, and forty-nine others distributed semi-randomly to full nodes that happen to be up at that very minute - would be a good idea regardless of whether you determine block formation via PoW or PoS.

Hmmmm....  Now that I think of it, I have an idea .....  

Ya, we have 10ga and 8ga wiring, but our house was rewired just two years ago and I'm paranoid about such things....  We paid extra. 

Unfortunate, but true.  This is why civilizations eventually require laws and law enforcement.  The old "fuck the law" attitude that some of the aforementioned radical libertarians and anarchists adopt enables the ones who are just out for your money to flourish.  But if we want to build anything worth having in the long run, it has to be part of civilization, and that means it will have to have laws and law enforcement to control the population of scammers. 

I think the best we can hope for is that we get the opportunity to participate and try to make sure that the laws aren't stupid or oppressive.  Controlling the scammers is literally all we want the law to do. 

Sure's heck can't rule it out.  People get so stuck on the "dumb hillbilly" stereotypes that they forget there are also smart hillbillies.  And you cannot run a farm these days without a full-time Internet connection so....

Teaching your kids is easy.  Give them US$150 worth of bitcoin and point them at the NewEgg site.  Show them how to do it once, that's all they'll need.  :-)

No.  They are keysplits.  The thing in the deposit box and the thing in my wife's files need to be combined in a very particular way in order to yield a key that would open my bitcoin wallet.  It is a procedure she knows how to do.  

I have a safe deposit box.  It contains (part of) my key.  If I die, the bank will give my wife a key to it.  While I am not dead, they won't.  My wife has the other part of the key in her files.

I believe that Hal would have distrusted Monero's extended scripting language.  It opens a lot of attack surface, which a designer has to go over with a fine-toothed comb to make sure it doesn't present an avenue for DoS. 

I can't speak much to Satoshi's opinion since I only interacted with him briefly and via email, but I think he would probably have agreed, or at least taken Hal's opinion as valid. 

This is not to say that there's necessarily anything wrong with it.  But it could take a lot of time and effort to study it in detail and figure out all the combinations that could be used in attacks. 

For example, things most people don't consider like, do your arithmetic operators allocate?  If they do, can someone send dozens of transactions that do a lot of arithmetic in an attempt to exhaust memory? 

And so on.  You really have to get in depth on each and every opcode to know what attack surface you present.  Hal disabled about half the opcodes for the scripting language for Bitcoin, and had good reasons for each one why.  Although I think most of them could have been fixed with work on the implementation, he was right. 

And that, in a nutshell, is why deciding chain priority by "coin days destroyed" is a basically broken idea.

It gives the attacker the opportunity to generate more priority in the attack chain, simply by spending the double-spent coins at a later point in the attack chain.

As I said right after the one sentence you quoted, the PoS algorithms we've seen so far are inherently flawed.  That's one of the flaws.  It can be fixed, but not in any of the ways people have done so far.

PoS as an idea isn't inherently flawed.

But the PoS algorithms we've seen so far are IMO very flawed. 

I believe that Transactions-as-proof-of-Stake can work in the long term.  But I only see it as being successful if the people securing the chain (that is, the people making transactions and the people running nodes) are the ones who actually get paid for securing the chain.  And I think that while TaPoS is good for securing the chain in the long run, it's probably best deployed in combination with something else to secure the most recent few blocks.

And I really object to the "lock your coins up for a month to get a chance to form a block" algorithms, because P&D operators can take advantage to "squeeze the market" and then leave, while most others have their coins locked up and can't sell, leaving their stakers flat. Maybe I'm just inherently suspicious, but I expect that when somebody does something that gives another person an opportunity to profit at their expense, they will most likely lose.

Finally, any kind of proof-of-stake requires distribution among a lot of different actors to help stabilize the market.  If you create few whales, rather than many goldfish, then everything depends on individual circumstances rather than generalized and more predictable stable laws of economics.  And so far we haven't seen a PoS launch that really did  that very effectively. 

I'm interested in this, but I don't see how BOINC has the properties needed to secure a cryptocurrency.  I think a lot of people saw possible problems with it and stayed away.  How did Gridcoin manage the problems involved? 

First, a lot of BOINC problems are linear (fastest computer wins ALL the time, rather than in proportion to speed).

Second, it doesn't usually happen, as I understand it, that everybody is working on the SAME problem - some get easier and some get harder. 

Third, it makes their servers a bottleneck and a vulnerability - people have to contact these servers to get a new problem every round, the block chain dies if the servers ever go down, and somebody could arrange to get "easy" problems if the operators collude or can be bribed. 

I'm not saying there's no way for it to work; I just don't know how to engineer around these issues.

Be warned that Bitcoin as a whole is still a pretty woolly area legally.  The law wasn't ready for it, there's a lot of uncertainty, and there are, predictably, scammers and fraudsters trying to take advantage of that uncertainty.

It's been getting better lately as the law catches up a little and figures out how to do their jobs here. But the scammers are still around.  Making a bet that they're planning how to get you to "invest" your money in their next house in the Bahamas would be a pretty good investment.  

That said?  An investment in mining could be a good idea if and only if you live in an area with heavily subsidized electrical utilities and/or can work out a sweetheart deal that gets you at least a 50% discount on electricity.  Some people are installing bitcoin miners as electric heaters in chilly areas, which gets them free electricity out of what homeowners or business owners are willing to pay for utility costs for heating. Otherwise, mining reliably loses money for two reasons.  First, because you're competing with people who do their mining in exactly those conditions.  And second, because at least a few of them are deliberately mining at a loss, either as an expensive way to speculate on a future price rise or as a premium they're willing to pay for the "virgin coins" they mine, which have no transaction history whatsoever and, if spent carefully, cannot be traced at all.  

Ummm, yeah, that probably means exactly what it sounds like it means if you're paying attention.  Like I said, you should be very careful about whom you do business with and what business you do, because the law is catching up and the sort of people who are likely to come to their attention are by no means gone.  

An investment in Bitcoin startups would be best for the community as a whole, but for the investor it's a risky play.  Could be astoundingly lucrative depending on what happens, but.... I guess it depends on how much money your coalition has to invest in the *whole* portfolio that determines how much you can put on something risky.  

Probably the most reliably profitable, least risky, and lowest-margin bitcoin business right now would be a currency exchange.  Right now there are dozens of players trying to be currency exchanges - and federal money transmitter laws are removing many of them from the business, so if you can navigate that legal maze the competition is getting weeded out for you as we speak.  But on that score you'd have to act pretty fast, because there are already some big players moving.  It's not just the Winklevii either; there are some sharper, faster-moving businessmen in some of those startups.  

Second on the list is people who supply payment services;  Joe merchant wants to accept Bitcoin at his truck stop bar & grill, but doesn't want the bother and complexity of managing it himself.  Payment services guy steps in and sets him up with the appropriate secured equipment, shows him how to use it, processes payments for him, etc.  Bitcoin here is a competitive advantage over Visa etc, because the per-transaction costs are lower.  That is, the bitcoin-to-bitcoin transaction costs are lower.  If you're paying an exchange to convert to and from other forms of money, you'll want to aggregate bitcoins to make fewer and smaller tx at the exchanges.  Or, you know, do vertical integration and be the exchange as well as the payment services provider.

Hey, what a good idea.  I did that. 

Right.  I know you can check the information at a given moment.  But that doesn't call itself to your attention, or allow you to track what it does over time. Well, I guess you could make a script that queries bitcoin-cli every minute or something.

What I was talking about is monitoring it in a way that allows you to see when it changes - a graph like network traffic that shows it over time, or at least some dynamically updated number in the Information panel of the qt node software like the current number of blocks in the information panel.

Good.  That's as it should be. I'm betting the problem was my browser open on the blockchain.info page that was showing me how many tx there were....   


That's actually kind of awesome.  How do you set that kind of test up?  You feed a script of RPC commands to add tx? 

Mostly because a byte is the smallest unit you can easily handle and do anything with -- and you can handle bits only in the context of being parts of larger units for the most part.  So I think of encodings in terms of bytes, making it 256.

Something that I've wished for in Bitcoin core itself the last few days is an option to monitor the size ( in both transactions and kilobytes) of the memory pool of unconfirmed transactions.  I've been seeing it get up to 3000 or so and staying there for hours during the last few days, while some miners continue to turn in tiny 60k blocks that have a few dozen transactions at most.  

It's a perverse-incentives problem.  Some miners are in such stark fear of an orphaned block that they're not processing transactions, and I think that having the ability to at least monitor that situation (outside of blockchain.org) would be a really good addition to the core.  It may turn out that we need to do more. 

I'd like to see a patch to the most-work rule that breaks equal-work ties by how much a proposed block reduces the transaction pool rather than by the order in which it was received.  But that would be risky in a few ways; it would lead to more rather than fewer orphaned blocks, and provide a new way for an attacker to cause a short-term fork.  And it would be a potential compute-time DoS, because you'd have to process all the blocks that are now "orphans" a lot more intensively to find out whether to replace your current chain tip.   So that's an experiment that needs tried out in an alt, long before it becomes a serious proposal for Bitcoin itself. 

Umm, it seems like pretty sound reasoning to me.  much, much better reasoning than the ten words of yours I just quoted. 

I accept as an axiom that the purpose of the block chain protocol is to come to a shared consensus about what actually happened and what therefore can happen next. 

A reorg means that consensus is not shared -- therefore meaning, for the time that it persists, the purpose of the block chain protocol is not being fulfilled. 

Because a lack of shared consensus is a condition that enables people to double spend, or allows people who have been paid to have those payments undone and be deprived of their money, I characterize this failure as a "security failure." 

This is crystal clear.  Are you trolling, or just stupid?

Yeah, as I keep saying, I made a scripting error. 

I set up my tickler to alert me for any mt.gox movements, but I gave it the wrong URL.  I figured the 'recent transactions' page meant RECENT, you know?  And I had my phone turned off at the time, but I didn't think about what that meant....   

So the next morning I woke up bleary-eyed, turned on my phone, and immediately got the text message and went "holy crap", and I came here and posted an OP that's just plain wrong.