Economically Unspendable Outputs: A Problem On The Radar

[Sukrim]

Also if you are worried about performance problems, I would bet that by then the ALU in your watch can do native 1024-bit int arithmetic operations.

How much are you willing to bet and whom are you allowing to escrow this? I'd go with up to 10 BTC against this statement, if it is formulated properly. Your move!

[misterbigg]

People talk about "old" coins, and "aging" coins, to avoid a fee, but what is the actual meaning of "old"? How long do they have to sit, is it a coupe blocks, a couple days, a couple weeks?

It's based on the number of blocks. I'm not sure about the actual number but I think it's on the order of weeks. The reason that age matters is to prevent relay spam where someone just transfers one satoshi back and forth between two addresses without a fee.

[misterbigg]

This is a re-post of my comment from the other thread.

I am not sure why no one has mentioned it here.  But starting today, Satoshi dice already changed its method of sending out 1 satoshi for confirming losing bet.  Instead, they are now sending out 0.5% of bet amount back for losing bets, regardless of bet size.  Given that minimum bet is 0.01 BTC, now all Satoshi Dice losing bet tx starts at 5000 satoshi and up.  I presume that SD changed its policy due to the introduction of SD filter patch.

Therefore, Satoshi Dice now absorbs all tx fee, and return 0.5% of bet amount, even for minimum bet.  Previously, it will deduct a tx fee first and send you 1 satoshi.  So you need to bet at least 0.2 BTC or so, to get 0.5 mbtc back.  But now even 0.01 BTC minimum bet gives you at least 5000 satoshi back.  I presume the change is made today because of all the heat and the patch here.

Hallelujah!!! Can you point me to a tx in the Blockchain and help me understand the outputs and how to verify this? For example:

https://blockchain.info/tx/6e0092ec8abca3ec3df5edbdf4c759024554d8e43c2f1956a3248fe1e6e664f6

Is this a new losing bet confirmation?

FYI 5,000 Satoshi is still considered dust by developers.

[SalvorHardin]

This is a re-post of my comment from the other thread.

I am not sure why no one has mentioned it here.  But starting today, Satoshi dice already changed its method of sending out 1 satoshi for confirming losing bet.  Instead, they are now sending out 0.5% of bet amount back for losing bets, regardless of bet size.  Given that minimum bet is 0.01 BTC, now all Satoshi Dice losing bet tx starts at 5000 satoshi and up.  I presume that SD changed its policy due to the introduction of SD filter patch.

Therefore, Satoshi Dice now absorbs all tx fee, and return 0.5% of bet amount, even for minimum bet.  Previously, it will deduct a tx fee first and send you 1 satoshi.  So you need to bet at least 0.2 BTC or so, to get 0.5 mbtc back.  But now even 0.01 BTC minimum bet gives you at least 5000 satoshi back.  I presume the change is made today because of all the heat and the patch here.

Hallelujah!!! Can you point me to a tx in the Blockchain and help me understand the outputs and how to verify this? For example:

https://blockchain.info/tx/6e0092ec8abca3ec3df5edbdf4c759024554d8e43c2f1956a3248fe1e6e664f6

Is this a new losing bet confirmation?

FYI 5,000 Satoshi is still considered dust by developers.

Yes, that's a good example of new losing bet confirmation.  There are plenty more at SD's website under Results or Unconfirmed.  You can observe the betting action of the owner of your tx sample https://blockchain.info/address/12AKJqjcwX85gaV4veHQFm4ZbsWD1hNPPAf.  Notice how SD is now confirming losing bet by sending back a flat 0.5% of bet amount, without any fees deducted upfront. 

As for 0.01 BTC size bet, they are typical for the really low odd bet like "less than 1".  But most of SD action is taking place in the 48% to 50%, with some in the 24% and 36%.  So most of betting size are larger than 0.01 BTC.  Though obviously people who run bots or martingales would still use 0.01 BTC as standard starting bet, but typically the tx fees would simply kill you over time except for those betting on the really low odd bet, like "less than 1".

[misterbigg]

Okay so there has been considerable improvement?

[SalvorHardin]

Okay so there has been considerable improvement?

If you want my honest opinion, unfortunately no.  In fact, I expect the betting with Satoshi Dice may pick up in the short term since SD has effectively reduce the edge for the house, i.e. net payoffs to the gamblers has increased.  So instead of 1.9% edge for the house, those who bet the minimum amount now only have to deal with a 1.4% house edge, if SD does not lower the payoff of a winning bet in future.  

You will notice that the betting with SD has actually slow down/stablize despite recent influx of new Bitcoin users.  So by increasing the payoff of losing bets, it has slightly reduce house's edge, and this may renew gambler's interest, as gamblers always love a low house edge.

Of course, not having to deal with thousands of unspendable 1 satoshi input is a big improvement for the gamblers and miners.

[misterbigg]

Okay so there has been considerable improvement?

If you want my honest opinion, unfortunately no.  In fact, I expect the betting with Satoshi Dice may pick up in the short term since SD has effectively reduce the edge for the house, i.e. net payoffs to the gamblers has increased.  So instead of 1.9% edge for the house, those who bet the minimum amount now only have to deal with a 1.4% house edge, if SD does not lower the payoff of a winning bet in future.  

You will notice that the betting with SD has actually slow down/stablize despite recent influx of new Bitcoin users.  So by increasing the payoff of losing bets, it has slightly reduce house's edge, and this may renew gambler's interest, as gamblers always love a low house edge.

Of course, not having to deal with thousands of unspendable 1 satoshi input is a big improvement for the gamblers and miners.

Alright well, let's recognize that there are two problems here:

1) Increased S.DICE transaction volume without a corresponding increase in Bitcoin adoption (likely due to bots)

2) Economically unviable transaction outputs which cost more to send than the amount they transmit

As far as #1 goes there is little that we can do about it, and there is a strong argument to be made that we shouldn't do anything about it. Eventually Bitcoin will have to handle this volume anyway. The problem is that it is driving up fees sooner than expected, which could dampen Bitcoin growth. In the short term this is the bigger problem.

But #2 is the larger long term problem, because the outputs are unprunable and they will incur an ongoing storage cost for all full nodes.

It sounds like #2 has been addressed is this the case?

[TooCasual]

Let's say I shut down SatoshiDice tomorrow

How about just not sending dust to notify losing bets?

+1.  In a net shell.  Why flood the network with shit? (Dust is just not a strong enough word) . It's a waste.  Keep the fees make more BTC on the losing bets. Win-Win. As a miner/user that's my vote.

TC

[tvbcof]

I think this "problem" is only temproary will fix itself when the price for transactions go up.

This could happen when we start hitting the hard limit of 1 MB per block.

I think it still could be a problem if it is worth it to someone with deep pockets to attempt to destroy or manipulate Bitcoin at the expense of buying all or most of the transactions.  (And man would such a bidding war between parties for whom money was little of  no object would certainly be a windfall for miners.)

But, as I said, as long as the long-term damage from caked on dust is predictable and limited to a known and acceptable level I don't see this form of attack as a show-stopper issue.  Floating or vastly increasing the block size would just make an attack of this nature cheaper and that much more likely to succeed.  At least in causing long term damage to the Bitcoin solution.

edit: add missing word.

[Elwar]

Yes, let us focus on Satoshi Dice and make them change their ways so that we can ignore this "flaw".

Meanwhile, to drive the price down so someone can buy a bunch of BTC, they can flood the blockchain with tiny transactions and make it "bothersome" for miners.


I am not a miner and have not been following them, but has there been some huge uproar from the miners saying that they are super bothered by all of the tiny transactions? There are ASIC machines coming out that can do 60 MH/s, are they going to be worthless because of so much "dust"?

I have not seen a huge exodus of miners fed up with the amount of traffic.

[Luke-Jr]

Yes, let us focus on Satoshi Dice and make them change their ways so that we can ignore this "flaw".

The flaw is miners neglecting their job to filter out spam. Do you have a solution to force miners to care about Bitcoin in the long-term?
I'm all for better solutions, but so far I've not heard any.

[justusranvier]

Do you have a solution to force miners to care about Bitcoin in the long-term?
I'm all for better solutions, but so far I've not heard any.

You could make it economical to clean up the mess, which is a better long term solution because it doesn't rely a constant arms race to filter out bad behaviour.

https://bitcointalk.org/index.php?topic=150726.0

At least one client developer is willing to automatically clean up those unprunable dust transactions if the right incentives are in place.

[tvbcof]

Yes, let us focus on Satoshi Dice and make them change their ways so that we can ignore this "flaw".

Meanwhile, to drive the price down so someone can buy a bunch of BTC, they can flood the blockchain with tiny transactions and make it "bothersome" for miners.


I am not a miner and have not been following them, but has there been some huge uproar from the miners saying that they are super bothered by all of the tiny transactions? There are ASIC machines coming out that can do 60 MH/s, are they going to be worthless because of so much "dust"?

I have not seen a huge exodus of miners fed up with the amount of traffic.

Those actually mining are not 'peers' in the supposedly peer-2-peer system per-se.  Transfer nodes are, and yes, there have been an exodus of these due to bloat and overhead.  It is quite possible to render these almost completely outside of the reach of all non-commercial participants depending on how things evolve.  How much of a problem that is depends on where one stands in the Bitcoin ecosystem from a philosophical and/or monetary investment standpoint.

[niko]

Yes, let us focus on Satoshi Dice and make them change their ways so that we can ignore this "flaw".

The flaw is miners neglecting their job to filter out spam. Do you have a solution to force miners to care about Bitcoin in the long-term?
I'm all for better solutions, but so far I've not heard any.

I don't see SD dust transactions as "spam" - a better example of spam would be your insertion of jesus messages into blockchain: https://bitcointalk.org/index.php?topic=38007.0

[Elwar]

So I should be able to start CPU mining again and make some money?

[cbeast]

Yes, let us focus on Satoshi Dice and make them change their ways so that we can ignore this "flaw".

The flaw is miners neglecting their job to filter out spam. Do you have a solution to force miners to care about Bitcoin in the long-term?
I'm all for better solutions, but so far I've not heard any.

Let me see if I understand how this will play out. Eventually only very large mining operations will be profitable. They will police each other by agreeing to accept each other's spam-free (or at least within a tolerable threshold) blocks out of necessity by branding (i.e. VISA/MC) their blocks. There will be the temptation to take the transaction fees anyway, but it will be balanced against the speed their blocks will be verified. Merchants that subscribe to mining operations to clear their transactions will prefer the fastest block verifications.

[Prattler]

Satoshi dice, thank you for exposing dust output spam vulnerability in bitcoin (bloating unspent tx output data, which is RAM and is much more expensive than HDD!). You have already proved your point, would be nice if you stopped spamming until devs and mining pool operators can act on it.

misterbigg, thank you for raising awareness of the problem and discussing it in a polite manner. Some evangelists keep saying bitcoin is perfect. Well it's not! Bitcoin is still an experiment! It has some design problems that need to be solved. Dust output spam vulnerability is one of them.

Developers and pool operators, please address the dust output spam vulnerability. I personally think that transaction outputs should not be smaller than the fee/kB paid. Additionaly, it would prevent mistakes like https://bitcointalk.org/index.php?topic=149926 – a very nice side effect!

[ShadowOfHarbringer]

Satoshi dice, thank you for exposing dust output spam vulnerability in bitcoin (bloating unspent tx output data, which is RAM and is much more expensive than HDD!).

Exactly my point. We should THANK them for exposing it - i would rather have SDICE exploit this flaw now rather than somebody else with really bad intentions some time in the future.

You have already proved your point, would be nice if you stopped spamming until devs and mining pool operators can act on it.

No, it wouldn't be nice. If not SatoshiDICE, then somebody will use this flaw in the protocol.
SDICE is putting a pressure on the parties interested (miners, devs) to fix this. So I say don't bother SDICE and let it continue until the protocol bug is resolved.

misterbigg, thank you for raising awareness of the problem

For the thousandth time: There is NO "problem". Just a bug in the software.

----
Also, OP of this topic is simply spreading FUD.

[Prattler]

For the thousandth time: There is NO "problem". Just a bug in the software.

Bitcoin is not just software, it's also the users who use that software (or ruleset/fork). P2SH change was simple from the software standpoint, but quite difficult because of the social aspect of pools having to agree to upgrade. The fix for this might be as slow as P2SH change. Bugs in societies are much more difficult to fix than in software

[blockbet.net]

If not SatoshiDICE, then somebody will use this flaw in the protocol.

It's not likely that this "somebody" would ever reach the kinds of volumes that SD have, so it wouldn't be much of a problem if this "somebody" would do so if their operations were minimal.

Second, by raising awareness of the issue, SD, the "others", and the gamblers can decide and do what's best for Bitcoin, since they would only be hurting themselves by harming the network with their spamming.