Topic: WARNING! Bitcoin will soon block small transaction outputs - page 5. (Read 58538 times)

Remember when a Gavin was 5430; man those were the good old days.

What do we call the next default then?

The dust threshold is a variable, there is no minimum (or maximum) just a default.  The default value in v0.8.2 is 5430 satoshis, that default will likely be changed in future versions.  The min mandatory fee (on low priority txs) for example has been changed 4 times in the history of Bitcoin to accomodate increasing value.  

Also (even in the absence of future version) any miner can change the dust threshold by changing a single line in the config file. Right now this very second a miner could (if they chose) run 0.8.2 with a dust threshold of 1 satoshi if they felt the default of 5430s was too high.  They could also run with a dust threshold of 8196s (~$0.01 at current exchange rate) if they felt the default was too low.

Its a minminum we can live with I say, but what if bitcoin goes up to $1000. Does the bitcoin minimum deposit go up?

It is in version 0.8.2 of the QT client.  0.8.2 is sill l RC and not final yet.
There is no hard minimum however the default value for dust is 5430 Satoshis (~0.6 US cents at current exchange rate).

Is this really going to happen or has it already happened. I'm just curious to see if it did or did not.
And do any of yous know what the new minimum limit is?

I think the technical people should be focused on technical stuff, not PR.

By "amounts so small it costs more to spend them than they are worth" I mean it costs more BTC in fees to re-spend than the mathematical value of the output*.  In other words, there is no economic reason to ever spend that output, unless the fees model changes or you manage to spend it without a fee.  It's like seeing a penny on the ground and knowing it will cost 10 cents worth of food to "pay" for the energy expended reaching down and picking it up (probably exaggeration, of course.)  What economic motivation is there to do that?  In Bitcoin, miners and other full nodes have to carry those pennies that people thought it was fun to throw around with no motivation to pick up, for the foreseeable future.  Miners have never been obligated to accept this abuse, and it's looking like many of them will put an end to it now.

Yes, I suppose sending 1 satoshi to one of Satoshi's addresses may have more subjective value to you than to the cold, hard 1-satoshi literal value.  If so, express that burning desire by paying a premium in fees and maybe someone will mine it for you.

Sorry if this feels brusque.  I find the attachment to 1-satoshi divisibility silly as it's just as arbitrary as 5430 satoshis.  I'm disappointed the network has been censoring my attempts at spending 0.1 satoshis.

*By the way, if I'm understanding the 1/3 fraction in the patch, 5430-satoshi outputs are still quite uneconomical to spend.

The fees do.  Lets say the only way to spend USD was to mail them to a recipient.  A stamp costs $0.46 obviously paying $0.47 ($0.01 + $0.46) to make a $0.01 payment is UNECONOMICAL.  The fact that you could make a $0.01 payment (literally mail a US penny to a creditor) doesn't change the fact that it is UNECONOMICAL.  Right now transactions below ~5000 satoshis are uneconomical.  They won't be spent because the cost to spend them is greater than the value gained in spending them. 

In the future (if/when exchange rate rises significantly) the threshold on economic transactions will decline.  That is all "dust" is.  Dust is the threshold where an output isn't economical to spend.  Prior it 0.8.2 you could create outputs that due to economics likely would never be spent after 0.8.2 you can't.  Remember the dust threshold is now a variable.  The default value makes it 5430 satoshis however miners are free to set it to whatever they want.   

Miners have an incentive to both:
a) not exclude economical (meaning will be respent in the future) transactions because that would be turning away paying customers
AND
b) exclude non-economical transactions as they bloat the UXTO and raise miner's costs perpetually.

Thus miners have no incentive to set the dust threshold incorrectly.  Too high means less revenue and too low means excessive future cost.  0.8.2 lays the foundation for a more comprehensive system to balance the needs of small transactions with the cost paid by everyone

There is cost to bitcoin, there will ALWAYS be a cost to Bitcoin.  Now today that cost is partially "hidden" because the block subsidy is so large but the cost still exists.  Bitcoin is an efficient payment mechanism because it is p2p and there are no natural monopolies so (economic theory tells us)  the cost of transactions will be low.  Low doesn't mean zero.

The user experience is horrendous. You can still do it, if you want to. There will undoubtedly be miners who are willing to include such a non-standard transaction for a fee (or maybe even without). As always, miners are free to accept whatever (valid) transactions they want. This simply changes the default to something more sane, improving the general user experience.

Who decides what a transaction is worth? Value is subjective, and even if you can't think of a reason why someone might be willing to pay a lot to send a transaction that doesn't mean they don't exist.

The assumption that this is directed at SatoshiDice, or any other single entity, is fallacious.

The postulation that this doesn't accomplish anything is simply incorrect.

This patch removes the ability to CREATE outputs (using standard transactions) that cannot be economically spent. It also removes a "magic" number from the source code and introduces greater configurability for miners.

The result is that far fewer people (newcomers especially) will end up with a wallet full of outputs that they cannot spend without incurring a fee greater than their value.

In other words, the change addresses a real, existing problem and enhances the experience for newcomers.

You know what the biggest reason this is a terrible idea is? It doesn't even accomplish anything. Consider SatoshiDice, the biggest "offender" when it comes to noneconomic translations. All they've got to do is have people send an extra 0.001 BTC with their bet and send it back either with the winnings or alone if they lose. They could even throw in an extra satoshi for tradition sake.

IMO this sounds like a terrible idea in general, but would require a hard fork, as it would mean miners would be allowed to claim funds they didn't have the private key for.  I don't think any substantial portion of the network would adopt that fork.

Some of the recent posts seem to believe dust will be unspendable.  That's not correct, it will continue to be as spendable as it is today (that is, at high fees if you're not careful.)  Those high fees should perhaps be addressed, but it's outside the scope of this solution.  If you believe this solution should not be adopted if it won't address that, that would actually be a reasonable objection as opposed to "this is censorship!"  Personally I think it's reasonable to first slow the creation of dust before turning your attention to getting rid of the existing dust.


He has responded here, here, here, here, here (good one IMO).  gmaxwell has responded here, here, here, here, here, here, here, here, here, here (good one IMO), here (to name the most recent ones; I believe there's plenty more.)  jgarzik has responded here, here, here, here, here, here, here, here (good one IMO) , here (again these are the most recent; there are more.)  What more can you ask for?  There are only so many ways and so many times they can explain the same things over and over before they get sick of it.  And I doubt anybody has made any point the developers haven't already considered, so there's no reason they need to be deeply engaged in the multitude of threads there are about this.

Bottom lines: a) this change is voluntary and adjustable, b) there have always been ways to discriminate against transactions, and there have long been transactions that have been discriminated against by default (non-standard scripts, zero-value outputs), c) you never had any enforced right to have your transactions processed, so complaining that this will now lead to your tx not being processed is nonsensical (offer a large fee for the best chance), d) this change is trying to prevent the sending of amounts so small it costs more to spend them than they are worth (how is it sane to want to keep this property?!).


How are they lost?  You have always been able to spend dust (if at high fees because of large tx size), and this change doesn't change that.  I do agree that it would be nice to be able to recombine dust without high fees.  I think this would be best handled by the market -- pools could announce they will accept low-fee dust-combining transactions even if they're large, as long as they have the effect of shrinking the UTXO to some degree.  I am not sure all the existing dust outputs can be efficiently gathered though.  It seems like you would have to pay people quite a bit more than the dust is worth to get them to bother adding their outputs/signatures to a big tx that consumes all the dust.

My son, PRab, and I were discussing this and he described a notion; the idea in this posting springs from that source; as dust ages eventually it should be reclaimable by a miner.  For example, 1 Satoshi might be claimable after 1 week of being idle, 10 at 2 weeks, 100 at 4 weeks and 1000 at 8 weeks.  Perhaps that is too aggressive but you get the idea -- just has to be tuned.  Anything greater than dust would have to be idle for a long enough time to be sure enough that it is never coming back into circulation.  Maybe there's a threshold above which idle Bitcoins are never reclaimable.

All good points.  Another reason this should not be permanent, at least without some more thought, is that over the long term, while these "lost" dust amounts are insubstantial, there could ultimately be substantial (and unnecessary) permanent loss of BTC.  There should, perhaps, be some way of recombining these lost amounts into something useful (without costing more than they're worth).

I also thought that fighting the FUDers in constructive manner is the way to go, but then realized that in btc's case FUD fails severely... although you can actually scare the latent new people from joining at the time being, you can't really scare the people that are already in, yet alone to scare the code itself, utp cable or the graphic card...

Code will proactively change to allow as much trashy transactions as the current price makes affordable... Eventually, when we reach 0 block reward, price is projected to be at the right level to support the miners with just a transaction fee...

This thread was created only so it could later be linked to by http://www.redstate.com/2013/05/16/tech-at-night-bitcoins-central-bankers-kim-dotcom-censors-mega/

I understand that Gavin is a busy man, but IMHO, promptly and accurately responding to such disinformation and FUD attacks, should be given high priority, because it may damage bitcoin more than any technical attack could.

What does everyone else think about this?

The bad news is, that you (not you personally) played right into the troll's trap, and helped create this 24 page long thread, which was then linked to in yet another article trying to spread FUD about bitcoin. This thread was started to do just that (create and spread FUD), and it worked flawlessly. My only wonder is, why Gavin himself didn't gave a clear answer, to stop the FUD from spreading?