Boycott 0.8.2

[solex]

Yes. Colored coins is the major casualty, however, this would seem to be a perfect market for Peter Todd's off-chain solution of fidelity bonded banks and might be a real-world proof of concept for the whole idea.

[kjj]

Colored coins are a pretty lousy idea anyway.

I mean they sound like a great idea at first, but they just don't seem to work when you try to actually do it.  That sort of stuff belongs in a merged mining alt-chain.  Namecoin is the proof-of-concept, but doesn't support fractional ownership, so it isn't exactly right.  A chain for general fractional ownership has been described elsewhere.  Just needs to be implemented.

[Elanthius]

You cant send a tx where the output is less than the new limit.

The new limit are based on developer estimates of cost.

But they have unintended consequences, for example, if you had made a business in colored coins, this change may hit you hard. 

OK, but my point was simply that those transactions can be spent at nigh on 0 cost regardless of how small they are. So the change is not in any way designed to save people from receiving tiny transactions because of the costs that person will incur spending them.

[DeathAndTaxes]

You cant send a tx where the output is less than the new limit.

The new limit are based on developer estimates of cost.

But they have unintended consequences, for example, if you had made a business in colored coins, this change may hit you hard. 

OK, but my point was simply that those transactions can be spent at nigh on 0 cost regardless of how small they are. So the change is not in any way designed to save people from receiving tiny transactions because of the costs that person will incur spending them.

Well no.   If you send a spammy low priority tx with no fee there is a very good chance it will NEVER confirm.  There is also a good chance it will never even be seen by a miner in your lifetime because every node following QT rules (which is more than just the QT clients) refuses to relay low priority tx which don't include a fee.  As a result you will be forced to include a fee greater than the value of the output you are trying to spend in order to spend it which is utterly stupid.   The 0.8.2 rules simply prevent CREATING new outputs which can't be spent.  If nodes are willing to relay tx with smaller fees that threshold also decreases.  For example the current default on low priority txs is a fee that is a minimum of 0.1 mBTC and thus that same node won't relay outputs smaller than 0.0543 mBTC.  If today (or in the future) that node lowered the min mandatory fee on low priority txs to say 0.02 mBTC then it would also relay outputs as small as 0.01086 mBTC.

[Elanthius]

Then riddle me this transaction

8bdb360363e17164a1a897f56e8ce9ab1f5deb1fed6786f7a74411e7ef2403b5

Which pays a 1 satoshi fee to get over the 1000B limit and include multiple "dust" transactions. So tell me again how dust can't be spent? Can you show me one transaction with a 1 satoshi (or more) fee that hasn't confirmed in over a couple of weeks?

[jgarzik]

Then riddle me this transaction

8bdb360363e17164a1a897f56e8ce9ab1f5deb1fed6786f7a74411e7ef2403b5

Which pays a 1 satoshi fee to get over the 1000B limit and include multiple "dust" transactions. So tell me again how dust can't be spent? Can you show me one transaction with a 1 satoshi (or more) fee that hasn't confirmed in over a couple of weeks?

Neither of that transaction's outputs are dust.  They are both >= 0.01 BTC.

Transaction inputs may be any amount.

[Elanthius]

No, the argument is that output dust is blocked to prevent the recipient from being charged with tx fees when they are next used. My point is you can freely use dust so there's no need to protect anyone. Now, if you want to say that we are protecting the blockchain from bloat then that's a legitimate argument. But don't lie and claim that people can't freely spend dust if they want to.

[tacotime]

Litecoin banned microtransactions a while ago by requiring huge (0.01) fees.  It's still going strong.  Stop whining.

[Eri]

No, the argument is that output dust is blocked to prevent the recipient from being charged with tx fees when they are next used. My point is you can freely use dust so there's no need to protect anyone. Now, if you want to say that we are protecting the blockchain from bloat then that's a legitimate argument. But don't lie and claim that people can't freely spend dust if they want to.

Let me try to clarify.

There is no rule preventing anyone from spending money they already have in their wallet, even if it is dust. Fees still apply and have always applied if the transaction doesnt qualify as free.(see rules regarding the fee structure).
This change would effectively prevent you from making a transaction with dust as an output.
Regardless of this change(and the guidelines before it) you can still connect to a miner that is willing to mine your 'non standard'(dust) transactions for you.

What the fee structure roughly is: (Age of money) *magic* (amount of money) *magic* (size of transactions) *magic!*.  A transaction can be free if its under that 1000 bytes and its number from that magic math is low enough to allow it to be sent for free.(feel free to look up the exact details)


On the new client your transaction didnt have dust as an output, so its a legit transaction.
Going by your transaction there are 2 possibilities.
-It passed the 'magic' so it didnt need a fee.
-Some miner was willing to mine your low value transaction even if it was considered dust.

To try to reiterate the point of all of this. If someone or several people or hundreds of people send you dust from many transactions, with more then 2-3 pieces of dust as an input you have gone over the size limit for a transaction to be free(unless you can find a miner to mine it). As a result you MUST include a fee if you want it in a block in reasonable time, or if you want it to be relayed by the network. The fee for these types of transactions can as a result be larger then the dust you are gathering from inside your wallet. Which brings back the question 'would you spend that dust if the transaction fee cost you more then the dust is worth?'

I hope this clarifies things a bit.

[tuheeden]

Boy there is a lot of interesting points of view in this thread.

An economic majority is one of the unique Great/Terrible features that bitcoin relies upon.

I saw many posts indicating that bitcoin was not designed/intended for micro-payments, which I do not believe is completely accurate. Since bitcoin has a strong deflationary bias, the average transaction size will DECREASE with time, making payments today that we may consider micro-payments, be standard payments in the future.

Addressing the arbitrary .01 cost of a transaction (or .0005) is an absolute must. Clearly the best solution is to let the miners set the fee that they will accept and process transactions so that competition and deflation can automatically adjust the cost.

As for dealing with wallet dust, probably the easiest way to deal with this is simply to export the keys from any/all wallets and then import them back into one wallet to consolidate your dust. It may still result in a transaction fee but at least you can keep all your pennies in one jar.

[jaywaka2713]

Boy there is a lot of interesting points of view in this thread.

An economic majority is one of the unique Great/Terrible features that bitcoin relies upon.

I saw many posts indicating that bitcoin was not designed/intended for micro-payments, which I do not believe is completely accurate. Since bitcoin has a strong deflationary bias, the average transaction size will DECREASE with time, making payments today that we may consider micro-payments, be standard payments in the future.

Addressing the arbitrary .01 cost of a transaction (or .0005) is an absolute must. Clearly the best solution is to let the miners set the fee that they will accept and process transactions so that competition and deflation can automatically adjust the cost.

As for dealing with wallet dust, probably the easiest way to deal with this is simply to export the keys from any/all wallets and then import them back into one wallet to consolidate your dust. It may still result in a transaction fee but at least you can keep all your pennies in one jar.

I like what I highlighted in red. But the dust we are talking about is mainly from all these transactions from faucets and satoshidice. Other than that though, what do people use microtransactions for?

[DeathAndTaxes]

I saw many posts indicating that bitcoin was not designed/intended for micro-payments, which I do not believe is completely accurate. Since bitcoin has a strong deflationary bias, the average transaction size will DECREASE with time, making payments today that we may consider micro-payments, be standard payments in the future.

When we talk about micro transactions we are talking about the purchasing power not necessarily the nominal units.  Today the dust limit is ~5000 Satoshis worth about half a US penny.  If the exchange rate rose to $5,000 the limit may fall to say 100 Satoshis still worth about half a US penny.

Case in point the min mandatory fee for low priority transactions (an anti DOS mechanism) has been reduced four times from 0.01 BTC to 0.0001 BTC over the years in light on

Regardless of the exchange rate Bitcoin is highly unlikely to be efficient for micro transactions (say transactions with a purchasing power equivalent to a few US pennies or less, circa 2012).

As for dealing with wallet dust, probably the easiest way to deal with this is simply to export the keys from any/all wallets and then import them back into one wallet to consolidate your dust. It may still result in a transaction fee but at least you can keep all your pennies in one jar.

That doesn't do anything.

[tuheeden]

The dust I am referring to is generated from mining (< .0001) or from transactions where division is required and the results are many digits out (maybe pay per click web adds). While Satoshidice is certainly the biggest dust generator at the time, there is no way to avoid transactions getting smaller with time due to deflation so a fixed cost will absolutely fail in the long run. This is already become an issue because bitcoin went from $10 USD per coin to over $200 per coin, so suddenly the minimum fee went from .10 USD to $2.00 USD - This is deflation and it is already negatively affecting what we would consider small transactions (a cup of coffee now has a $2 transaction fee..)

So while I don't think that verion .82 is the answer, its an short term solution to a bigger problem that has to be fixed, so I support the effort if for no other reason to keep bitcoin viable while it is growing in adoption. If you kill the small transaction market, you lose a big market of potential users.

[DeathAndTaxes]

The dust I am referring to is generated from mining (< .0001) or from transactions where division is required and the results are many digits out (maybe pay per click web adds). While Satoshidice is certainly the biggest dust generator at the time, there is no way to avoid transactions getting smaller with time due to deflation so a fixed cost will absolutely fail in the long run. This is already become an issue because bitcoin went from $10 USD per coin to over $200 per coin, so suddenly the minimum fee went from .10 USD to $2.00 USD - This is deflation and it is already negatively affecting what we would consider small transactions (a cup of coffee now has a $2 transaction fee..)

So while I don't think that verion .82 is the answer, its an short term solution to a bigger problem that has to be fixed, so I support the effort if for no other reason to keep bitcoin viable while it is growing in adoption. If you kill the small transaction market, you lose a big market of potential users.

The min fee has been lowered many times of the history of Bitcoin.  It was never $2.00 (USD equivelent) at any point in Bitcoin's history.

If you kill the small transaction market, you lose a big market of potential users.

The min fee on low priority txs is ~1 US cent and the dust threshold is ~0.5 US cents.   What small transaction market are you talking about?

[tuheeden]

The point of consolidating dust is that it is the same premise as the penny jar. If you use multiple wallets, various transaction identifiers, etc, etc - You always end up with small amounts of bitcoins left over. If you consolidate them all into one wallet (AKA - the penny wallet), then one day you get a free cup of coffee.

It doesn't address any big issue, but there was a previous post about how to deal with small bitcoins left over in various wallets so I was sharing what I have done with mine.

As for dealing with wallet dust, probably the easiest way to deal with this is simply to export the keys from any/all wallets and then import them back into one wallet to consolidate your dust. It may still result in a transaction fee but at least you can keep all your pennies in one jar.

That doesn't do anything.
[/quote]

[tuheeden]

The min fee on low priority txs is ~1 US cent and the dust threshold is ~0.5 US cents.   What small transaction market are you talking about?

Agreed. The low priority txs is still acceptable, but if bitcoins value was say $1000 USD then a fixed threshold would represent an ever escalating minimum fee.

Summary: The minimum fee needs to be adjustable and probably the best way to accomplish this is via an open market solution - let the miners who are processing the transactions set the fee.

[DeathAndTaxes]

The min fee on low priority txs is ~1 US cent and the dust threshold is ~0.5 US cents.   What small transaction market are you talking about?

Agreed. The low priority txs is still acceptable, but if bitcoins value was say $1000 USD then a fixed threshold would represent an ever escalating minimum fee.

Summary: The minimum fee needs to be adjustable and probably the best way to accomplish this is via an open market solution - let the miners who are processing the transactions set the fee.

Miners do set the fee and as a result the min dust threshold (54.3% of min fee).  The min mandatory fee is simply a default and is used to protect the network from denial of service attacks.  The default has been lowered four times in the history of Bitcoin to keep the min fee (on low priority txs there is no min fee on high priority txs) at a low level of around 1-3 US cents.

[jaywaka2713]

I honestly don't have an issue with the fee as I have never had the need to send a transaction worth less than 10 cents.

[justusranvier]

Summary: The minimum fee needs to be adjustable and probably the best way to accomplish this is via an open market solution - let the miners who are processing the transactions set the fee.

This is precisely what Gavin described in his "State of the Coin" speech at San Jose, and in the pull request on GitHub for 0.8.2.

It's not necessary to convince anyone that this is what should be done because people are already working on implementing it. This entire thread is about to be rendered moot.

[polrpaul]

Heh, no longer have to hate on 0.8.2...

News: Bitcoin-Qt 0.8.3 has been released. Download http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.8.3/