Economically Unspendable Outputs: A Problem On The Radar

[notig]

out of curiosity... is a reason why they wouldn't fix this because they don't think people will want to go through an extra step in creating an account, then sending funds to their site to play with?

[stdset]

out of curiosity... is a reason why they wouldn't fix this because they don't think people will want to go through an extra step in creating an account, then sending funds to their site to play with?

As far as I understand, the system now is provably honest to certain degree.
It is still possible to do some fraud IMO. Lets imagine we are SD. We want people to see, that on average say a 50% bet wins in half the times. But still, we can do some bets from our own speciall addresses which always win, therefore we are able to decrease winning rate for normal gamblers.
Correct me if I'm wrong.

[00null]

The thing is that when you lose at SD, you lose twice. First you lost the bet, and second you have to pay good money to get rid of the SD spam dust.
But I guess this will never be fixed, because Erik has too much influence on TBF and as a result on Gavins work. And with those ever raising profits he can already pay Gavin with pocket change from SD.

[thebaron]

Satoshi Dice is helping to drive the price of Bitcoin up.

Pros....cons...

[alexkravets]

Once hedge funds can funnel cash through Coinlab starting on March 22nd, THAT will become the price driver ... all this retail coin tossing around won't matter

[doobadoo]

This is definitely a flaw.  The BTC protocol should reject any output less than the block's tx fee, or the median tx fee for the last 60 or so blocks.

[zebedee]

"Transaction spam" is the largest issue threatening the fledgeling currency called Bitcoin. A brief synopsis of the problem:

"SatoshiDICE" notifies gambling addicts of losing bets by sending them a transaction containing 1 satoshi back to their original address. A regular transaction fee is also included in this transaction. The problem is that this produces an unspendable and unprunable output. The output is unspendable because it is so small that it would cost more in fees than it would to send. It cannot be pruned because it has to stay around in fast storage to detect double spends, even though ironically it cannot be spent (developers, let me know if this description is technically accurate).

You see, you don't really know what you're talking about and that's leading you to draw the wrong conclusions.

If I have other coins, say 4 BTC, and want to spend 1.25 BTC, it's trivial to create a transaction with 4 BTC + 1 or more single-satoshi inputs, that pays 1.25 + 2.7500000x BTC change as outputs, and pay no fees.

So not only can they be spent, they can even be spent and pruned for free.  Yes, I've done it myself.

[stdset]

So not only can they be spent, they can even be spent and pruned for free.  Yes, I've done it myself.

Good point. But an SD gambler is likely to have some thousands of single satoshi outputs. How much time it will take to spend them this way? It's like a pool with two pipes. Through one water gets in, through another gets out. And it will always be more water getting in than getting out.

[Jutarul]

So not only can they be spent, they can even be spent and pruned for free.  Yes, I've done it myself.

you have a missing incentive structure. A gambler doesn't care about the technological implications of dust output. His main concern is economic.

To get a fee of 0.01 you need at least 10^6 satoshi inputs. So there's no economic incentive. The only thing which would help is an intelligent client which "sweeps" the satoshi output and throws it in as a fee. However, as pointed out before, in a crowded blockchain, that little bit of required extra space may cause another important transaction to be left out, so there's no free lunch.

Overall transactions which have no intent of being spent (for economic or other reasons), pollute the blockchain. We need an incentive structure to discourage that - or at least make it expensive to do so.

[zebedee]

So not only can they be spent, they can even be spent and pruned for free.  Yes, I've done it myself.

you have a missing incentive structure. A gambler doesn't care about the technological implications of dust output. His main concern is economic.

To get a fee of 0.01 you need at least 10^6 satoshi inputs. So there's no economic incentive. The only thing which would help is an intelligent client which "sweeps" the satoshi output and throws it in as a fee. However, as pointed out before, in a crowded blockchain, that little bit of required extra space may cause another important transaction to be left out, so there's no free lunch.

Overall transactions which have no intent of being spent (for economic or other reasons), pollute the blockchain. We need an incentive structure to discourage that - or at least make it expensive to do so.

He's paying the fee implicitly becauser SD are paying it!  And they're making money.

Further, your mega-gambler has to bet right?  Each bet can take a prior losing satoshi, so even a gambler with a lot of bad luck needn't be building up the satoshis.  Think a little bit, this is a non-problem.

Yes the standard client software might need updating.  I mainly use electrum which makes it easy to specify what outputs to use, but any client updated to spend dust is all you need, and adding such logic is very easy.

[misterbigg]

If I have other coins, say 4 BTC, and want to spend 1.25 BTC, it's trivial to create a transaction with 4 BTC + 1 or more single-satoshi inputs, that pays 1.25 + 2.7500000x BTC change as outputs, and pay no fees.

So not only can they be spent, they can even be spent and pruned for free.  Yes, I've done it myself.

First of all your trick only works once, if the 4 BTC come from a transaction output that is sufficiently aged. So it's not "free", you're consuming the limited resource (old coins in this case). You could not, for example, use the same 4 BTC again to combine another single satoshi input for "free".

Second, assuming that you did not have aged coins you would have to pay a fee. The 36 bytes for the additional input (the 1 satoshi) will drive the fee up by more than one satoshi. This is what we mean by "economically unviable to spend."

[DoomDumas]

This would be taking action against a problem who is only a problem due to how to network currently operates. If you remove SDice from operation, another company will take its place. We should be happy that SDice has made this flaw apparent and take action not against SDice, but at the source rather than the "cause".

+1

I'm not against SD activities at all.  Did'nt play SD, because mathematicaly, it's a lost over time.  I dont care if peoples/bots are using SD.  If not SD, few other could take their place.  It's just pointing out an issue that must/will be resolved.

Always search for solution in regards to the source of the problem.  SD using the network is'nt the source of the problem.  The Bitcoin have to be upgraded to cope with the scaling up that is incomming.  SD helped Bitcoin by showing this scaling issue soon.  It's best like it, as it gives us time to find a solution, in a civil way, to an issue that is nice to have : Bitcoin must adapt to higher volume.  Is'nt very nice issue to cope with already ?

I'm confident that we will arrive to a very good solution, for the best of the Bitcoin futur !

Meanwhile, I've increased max block size of the full node I run since late 2010 !

Happy life to all

[misterbigg]

This would be taking action against a problem who is only a problem due to how to network currently operates. If you remove SDice from operation, another company will take its place.

I'm okay with that, at least we get some peace during the time in between and spare us the block chain bloat even if temporarily. And if the developers fix it before another company takes its place, all the better! But the idea that "it's okay if I do it because if I didn't, then someone else would" is morally bankrupt.

I'm all for libertarian ideals but they can only work when actors behave ethically. Sure, someone who finds a severe security hole in theory has the "right" to reveal the information right away, the "freedom of speech" to tell everyone about the flaw, but the morally correct action is to follow vulnerability reporting best practices. Inform the developers, and then after sufficient time has passed to fix the defect, reveal it to the public.

I hope this example makes it clear that even if we have the opportunity to do something, it doesn't make it right.

[Explodicle]

What exactly is wrong with this "bloat?"
Blocks won't go over the size cap, so it's fair to clients.
SD is paying enough in fees for miners to accept, so it's fair to miners.
Speculators get slightly richer from lost coins and increased popularity.
The only people who are worse off are other senders because they now have to pay a market price instead of tips.

Isn't this what we signed up for? Transaction fees go up, subsidy goes down?

[zebedee]

First of all your trick only works once, if the 4 BTC come from a transaction output that is sufficiently aged. So it's not "free", you're consuming the limited resource (old coins in this case). You could not, for example, use the same 4 BTC again to combine another single satoshi input for "free".

Yes, that's the currently agreed ruleset.  So what's the problem again?

You've gone from saying they can't be spent, to they can't be spent for 6 hours?  Come on.  And not every bet is a losing bet.  And you can spend 3 or 4 satoshis at a time.

Second, assuming that you did not have aged coins you would have to pay a fee. The 36 bytes for the additional input (the 1 satoshi) will drive the fee up by more than one satoshi. This is what we mean by "economically unviable to spend."

No, you can easily get a total of combined inputs and outputs of 6 or more without raising the cost from free.  Have you looked at what's happening in the actual blockchain, or the logic for free transactions?  And if you're paying the 0.0005 "minimum" anyway you can get away with a lot more inputs and outputs.

[solex]

What exactly is wrong with this "bloat?"
Blocks won't go over the size cap, so it's fair to clients.
SD is paying enough in fees for miners to accept, so it's fair to miners.
Speculators get slightly richer from lost coins and increased popularity.
The only people who are worse off are other senders because they now have to pay a market price instead of tips.

Isn't this what we signed up for? Transaction fees go up, subsidy goes down?

Everyone who has accepted the spirit of Bitcoin wants a new, peer regulated currency independent of central-banks and civil-liberties-crushing government control. The next thing is they want is for Bitcoin to take some market share from the existing fiat dominated world.

Applications like SD are generating new business within the Bitcoin domain, business which could never have existed before. This is fine to a degree, but this new, internal, business is not replacing general fiat usage or taking market share from paypal or mastercard, yet it is close to 70% of Bitcoin's transaction flow. This means that only 30% of Bitcoin's processing is replacing existing real-world fiat&payments. Even this might be fine, but Bitcoin has a finite capacity at the moment, the max block size, and a further unknown block size limit which is set by the network block propagation latency and verification time (remember, only 10 mins between blocks).

There is a potential risk, which looks to me a big risk, that SD-like applications will grow to consume nearly 100% of Bitcoin's capacity. At that point the "improve the world" dream will be on life-support. There is also the problem of unspendable outputs bloating the blockchain.

We are asking that SD optimize itself to at least buy some time.

[Explodicle]

Everyone who has accepted the spirit of Bitcoin wants a new, peer regulated currency independent of central-banks and civil-liberties-crushing government control. The next thing is they want is for Bitcoin to take some market share from the existing fiat dominated world.

There is a potential risk, which looks to me a big risk, that SD-like applications will grow to consume nearly 100% of Bitcoin's capacity. At that point the "improve the world" dream will be on life-support.

When did the spirit of Bitcoin tell you which applications are most important? Some of my dreams involve weird new inventions that weren't possible before, why should they take a backseat to direct conflict over occupied territory? If we both want the same shared resource for different ends, bidding on it seems like the fairest way to resolve our disagreement. Who knows, maybe provably fair gambling is the "killer app" for Bitcoin.

There is also the problem of unspendable outputs bloating the blockchain.

That "bloat" when maxed out will increase chain size linearly, while the cost of storage goes down exponentially. It's really an anomaly of young bitcoin that we had blocks below the cap this long - for most of bitcoin's future, the blocks should be nearly full.

[salfter]

Free transactions are only relayed or included in a block if the coins are of sufficient age, to prevent attackers from flooding the network with messages.

So if someone wants their 1 Satoshi they can just send it as a free transaction. Once it is of a sufficient age they will be sent.

Problem solved.

You most likely wouldn't send 1 satoshi by itself. One of the free-bitcoin sources has been sending 1-satoshi payments. I combine them with mining income and other income with my CheapSweep script (google this forum for the thread). It can take a few hours for the transaction to get accepted, but I've not had one not go through yet.

[solex]

Who knows, maybe provably fair gambling is the "killer app" for Bitcoin.

And I thought Bitcoin was a fiat-killer, the most powerful idea for the internet since the WWW.
I do hope some other people share this view...

[tvbcof]

That "bloat" when maxed out will increase chain size linearly, while the cost of storage goes down exponentially. It's really an anomaly of young bitcoin that we had blocks below the cap this long - for most of bitcoin's future, the blocks should be nearly full.

My thought too.  As long as things are predictable I'm not terribly worried.  At least in terms of bloat, and as long as it is possible to be an infrastructure operator without spending huge amounts of money and thus limiting Bitcoin to be reliant on a handful of well capitalized entities.

Seems to me that as long as the community does not choose to fiddle with the block size, the fees will start to take care of SatoshiDice-class problems and it should remain practical for the system to be operational on the backs of a wide class of operators. 

Start monkeying with the block size and all bets are off.  Possibly this is why things are coded as they are?