Feasibility of zero transaction fee for 'aged' bitcoin?

gredinger

member

Activity: 152

Merit: 61

Quote from: RHavar on February 24, 2021, 07:12:07 PM

Quote from: gredinger on February 23, 2021, 08:36:13 AM

The concept of "zero transaction fees" isn't valid on its own due to the feasibility of an attack on the network.

I'm not quite sure that's a problem. It's just that under the current set of rules bitcoin is running at "full" capacity. So if miners included zero-fee transactions (which they can) they would be doing it at a direct cost to themselves (as they could've included transactions with fee).

One way to solve this would be change the block-limit from a static 4M weight, to one that varied depending on how many "coin days destroyed" for instance. This might work in theory, if you were careful to design the rules to not encourage backchannels for transaction creators to pay miners out-of-band, but in practice it would never work:

a) You'd never get consensus on such a change
b) It potentially makes coin selection & fee selection for transaction creators a complex bitch
c) It makes transaction selection for miners more complex

b and c seem solved through some algebra and code.

a seems like it could be solved by a really rich person.

----

It seems like bitcoin isn't going to be able to have zero fee transactions again.

RHavar

legendary

Activity: 1463

Merit: 1886

Quote from: gredinger on February 23, 2021, 08:36:13 AM

The concept of "zero transaction fees" isn't valid on its own due to the feasibility of an attack on the network.

I'm not quite sure that's a problem. It's just that under the current set of rules bitcoin is running at "full" capacity. So if miners included zero-fee transactions (which they can) they would be doing it at a direct cost to themselves (as they could've included transactions with fee).

One way to solve this would be change the block-limit from a static 4M weight, to one that varied depending on how many "coin days destroyed" for instance. This might work in theory, if you were careful to design the rules to not encourage backchannels for transaction creators to pay miners out-of-band, but in practice it would never work:

a) You'd never get consensus on such a change
b) It potentially makes coin selection & fee selection for transaction creators a complex bitch
c) It makes transaction selection for miners more complex

gredinger

member

Activity: 152

Merit: 61

Quote from: ranochigo on February 23, 2021, 08:41:36 AM

Zero transaction fees are no longer feasible. It was calculated with the coin age as a factor a long time ago before it was removed. The inputs amount was also taken into consideration [1].

There is really no point to try to allow for free transactions to be relayed. The mempool is frequently being pushed to it's limits and it's fairly pointless to reserve any of it for free transactions; miners would rather choose the paid transactions.

[1] https://en.bitcoin.it/wiki/Miner_fees#Historic_rules_for_free_transactions

Aha! I knew it existed at one point in time. I knew I couldn't have came up with a novel idea, but rather just pulling it from memory.

Mempool occasionally clears up. I wouldn't mind waiting a few weeks for my transaction to finally be included in a block.

Quote from: odolvlobo on February 23, 2021, 01:20:34 PM

It is up to the miner to choose which transactions go into the block. No miner will pick a 0-fee transaction, regardless of its age, if there are paying transactions to include. There is no way to enforce your policy.

Well, it's up to the network to broadcast the transaction, which is where the issue is currently.

It's interesting there are some blocks where miners will not include any transactions (due to the method of mining). It seems that not validating transactions only hurts the network as a whole. For example, if most miners were to mine this way, no transactions (other than block rewards) would be processed effectively reducing the overall value of the network to nil.

odolvlobo

legendary

Activity: 4578

Merit: 3526

It is up to the miner to choose which transactions go into the block. No miner will pick a 0-fee transaction, regardless of its age, if there are paying transactions to include. There is no way to enforce your policy.

bitmover

legendary

Activity: 2352

Merit: 6089

bitcoindata.science

Quote from: gredinger on February 23, 2021, 08:36:13 AM

However, what about the concept of 'zero transaction fees' for wallets w/ coins that have sat idle for 210000 blocks? Obviously, miners would have a lower priority on zero fee transactions under transactions with fees.

You are suggesting that if the coins days destroyed (CDD) is too high, the coin should be threated differently.

This cannot happen because bitcoins are fungible.

Every satoshi is equivalent to every other, it doesn't matter how long that satoshi was waiting in your wallet.

Quote from: gredinger on February 23, 2021, 08:36:13 AM

I'd assume that any transaction made from the wallet would reset the block count.

As I said, there is already a metric for that, and it is called CDD. And yes, CDD goes to zero when you make a transaction.

NotATether

legendary

Activity: 1568

Merit: 6660

bitcoincleanup.com / bitmixlist.org

Quote from: gredinger on February 23, 2021, 08:36:13 AM

I'd assume that any transaction made from the wallet would reset the block count.

How would this help? A node could keep making infinite transactions with zero fees and make message spam, but I don't follow what you mean by resetting the block count. That sounds like you're changing the block height by removing blocks from the top.

I'd suppose that invalidating other unconfirmed free transactions from the same input address would've been a slightly better idea, but still largely useless against preventing spam, which is what I assume you mean when you say "attack".

DaveF

legendary

Activity: 3500

Merit: 6320

Crypto Swap Exchange

It is also kind of a loop.
Even if you did offer free transactions they would still no go iton the block until after there were no 1 sat/b transactions left.
Unless 1BTC really goes above several million dollars 1 sat/b is still a small number.

AFAIK for the most part, even worthless alts still require some sort of fee. Could be wrong on that but when a coin is worth $0.0000001 does a fee really matter?

-Dave

ranochigo

legendary

Activity: 3038

Merit: 4418

Crypto Swap Exchange

Zero transaction fees are no longer feasible. It was calculated with the coin age as a factor a long time ago before it was removed. The inputs amount was also taken into consideration [1].

There is really no point to try to allow for free transactions to be relayed. The mempool is frequently being pushed to it's limits and it's fairly pointless to reserve any of it for free transactions; miners would rather choose the paid transactions.

[1] https://en.bitcoin.it/wiki/Miner_fees#Historic_rules_for_free_transactions

gredinger

member

Activity: 152

Merit: 61

The whitepaper mentions using fees as an incentive to keep miners honest.
The concept of "zero transaction fees" isn't valid on its own due to the feasibility of an attack on the network.

However, what about the concept of 'zero transaction fees' for wallets w/ coins that have sat idle for 210000 blocks? Obviously, miners would have a lower priority on zero fee transactions under transactions with fees.

I'd assume that any transaction made from the wallet would reset the block count.

This way, one wouldn't be able to spam the network with zero transaction fees of 1 satoshi each from the same wallet.

However, someone could make a hundred wallets with 1 satoshi each, which would have at least one transaction fee. From there, they'd have to wait a few years, but they'd be able to attack the network with those wallets later on.

Surely this isn't a novel concept for using the age of the coin or wallet as a mechanism for transactions, and I'd love for additional resources into this very specific and niche topic.

Topic: Feasibility of zero transaction fee for 'aged' bitcoin? (Read 151 times)