Topic: Blocking the creation of outputs that don't make economic sense to spend - page 3. (Read 3525 times)

eyh for bitcoin equivalent of 0.10 is 10 satashi I can't believe why people dont see how danger for bitcoin is banig any type of transaction, I have impresion then peopol who push it hard have really really band intention.

are you trying to tell us that boost is a bad dep?

jokes aside, i think there being a floor value to bitcoin tx isn't a bad idea. with national currencies there is always a practical limit on how small a tx can get, e.g. in the US, tx below USD 0.10 are very uncommon. there is a fair deal of variation between various currencies, but a floor that is related to a basket of local currencies would make sense.

lower bound tx size = 0.10 x (episodic basket/average of currencies valued in BTC)

Forcing SD to pay for such externalities it causes is a great idea.

Thank you for not using BOOST_FOR_EACH. One more step towards eliminating a dependency.

Yep, that's exactly the reason why I said I'd prefer your solution if the exotic uses can be achieved without dust sized txouts

How much does it cost, if you assume reasonable trends for storage/electricity cost?

Well all the exotic uses come down to either the transaction output being unspendable, in which case it should be flagged as such specifically, or the output is spendable but represents something other than a Bitcoin. In the latter case you can always have your algorithm use non-special txouts and txin's so that the actual value of the special txout in Bitcoins can float as required.

Also, look at it this way: if your special txout represents something worth less than a transaction fee, moving it incurs a net loss anyway. If it's worth more than that, having to keep a transaction fee worth of coins sitting quickly becomes a minor cost.

Oh, we very much agree on this.

Actually, if all the exotic uses of dust sized coins can be done equally well with non-dust sized ones, then I prefer your solution.

Otherwise I proposed one that still permits small valued coins here: https://bitcointalksearch.org/topic/discouraging-dust-without-hurting-exotic-uses-of-small-valued-coins-151177

The point is, a transaction that will never be spent costs the network more than one that will be spent because the former requires expensive, high-speed storage (currently RAM, and maybe SSDs in the future) while the latter doesn't even have to be stored at all by the majority of validating nodes with pruning. To charge the same price for both, which is what the current fee system does, is crazy.

By how much they're willing to pay to create the small valued txout.

How can you tell the difference between spam and a legit use for an anonymous transaction?

Can't we block the spam without nuking legitimate uses for small valued coins?

I just browsed through some old discussions and found this:

Re: Flood attack 0.00000001 BC
https://bitcointalksearch.org/topic/m.7524

A lot of people have been creating various patches blocking SatoshiDice, or as some of you like to call it, Dead Puppies. However the real long term harm of SatoshiDice is that they create small-valued transaction outputs that cost more to spend than they are worth, thus bloating the un-prunable UTXO set which must be stored in relatively fast memory on every validating node. Personally I don't agree with blocking SatoshiDice, I see block space as a limited resource that should be sold to whomever is willing to pay for it, but UTXO space is not being priced appropriately by the fee system.

To that end I've created a patch that blocks any transaction output whose value is less than the per-KB fees that transaction pays. (with an upper limit to allow for deliberate sacrifices of value to miners, and a lower limit of 0.0005BTC) If SatoshiDice simply makes their "you lost" transactions have a value greater than the fee, they will be unaffected. (I suggest simply taking it out of the bet)

For miners and pool operators: this helps you in the long run because when pruning is implemented you won't have to waste money on expensive computer hardware just to hold a UTXO set full of spam. (right now >50% is "you loss" outputs from satoshidice) At the same time, the patch doesn't block SatoshiDice, so you can still collect the fees from their transactions.

For users: this also helps you in the long run, by ensuring that you can continue to run your Bitcoin client on inexpensive hardware for the long term, and continue to fully audit the entire blockchain and help keep Bitcoin safe.