Pages:
Author

Topic: [RFC] New TX fee: 0.0005 BTC - page 4. (Read 16341 times)

db
sr. member
Activity: 279
Merit: 261
May 10, 2011, 11:17:20 AM
#37
Seems like a bit of a stretch to claim that a state of mind is a public good.

Why? It translates to the payment of actual fees. If miners self-interestedly accept all profitable transactions that willingness will fall through the floor and with it the actual fees. That's normal competition and how all market prices work.
sr. member
Activity: 294
Merit: 252
May 10, 2011, 10:52:22 AM
#36
And nobody claims they are. Which brings us to the miner common: the willingness of users to pay substantial fees.

Seems like a bit of a stretch to claim that a state of mind is a public good.
db
sr. member
Activity: 279
Merit: 261
May 10, 2011, 10:40:57 AM
#35
The collective security of the blockchain is the commons

That's the user common. There is also a miner common.

The fees are competing for a time limited resource, inclusion in the next available block;

The space in a block will/should be unlimited and the cost of inclusion negligible.

while miners are competing for those fees.  Those are not commons.

And nobody claims they are. Which brings us to the miner common: the willingness of users to pay substantial fees.
legendary
Activity: 1708
Merit: 1010
May 10, 2011, 09:39:38 AM
#34

You think that miners with not for profit motives are going to be able to maintain a robust, attack resistant network. Others disagree.
Others are wrong.   Tongue
Quote
I can see this working if people voluntarily pay tx fees. If there are millions of users then they each wouldn't need to pay much to incentivise a strong network of block generators, and indeed, a culture of this sort may emerge. But is is a tragedy of the commons like situation.
No, this is where this error keeps coming up.  It's a commons, but not a 'tragedy of the commons'.  The collective security of the blockchain is the commons, but the incentives for senders to add a fee are different from the commons resource incentive.  The fees are competing for a time limited resource, inclusion in the next available block; while miners are competing for those fees.  Those are not commons.
Quote
In another post, you highlighted the fact that there is a transaction acceptance policy encoded in the main client which prioritieses transactions with fees, forcing fees out of merchants needing speedy confirmations, but there is no reason to believe that the mining pools of the future will maintain such a policy when there are fee paying transactions going un-harvested. Luke-Jr, to me, exemplifies the natural dynamic of transaction acceptance policy: accept txs with fees, don't accept txs without fees.

I know you understand this dynamic, however, you haven't made a compelling case as to why bitcoin will work anyway. In any case, I hope you're right. Although, I wish intelligent people on this forum would acknowledge this problem instead of stifling debate.

I have made such a case, it's the 'Wal-Mart' post.
hero member
Activity: 527
Merit: 500
May 10, 2011, 01:43:38 AM
#33
This is just the default ruleset.

Miners are free to choose their own rules.  I believe luke-jr's pool requires all TX's to have a fee, if an extremely minimal one.

I disagree with his policy, but it illustrates the free market already at work.

If the Mining Majority wants a different ruleset, then the default ruleset won't be worth the paper it isn't printed on  Smiley

I'm not sure what conclusion you can really draw from luke-jr only accepting transactions with fees.  When there are no longer any generation rewards for a block, miners would all clearly benefit if they all only allow transactions with a certain minimum fee, however for any individual miner, they would have an incentive to accept any transaction so long as there is some fee...no matter how small that fee is.  This will drive transaction fees down to the point where miners start shutting down and the network becomes more vulnerable to a powerful miner.  Despite there being a strong collective reason for miners to charge some minimum fee and despite the clear benefit to all users to pay a fee to ensure a sufficiently robust network of miners, I think there is a real risk that miners will want to scoop up all transactions paying anything and for users to pay the bare minimum fee (a satoshi I guess).  This is the tragedy of commons argument people make.

Why does this keep coming up?

No, not all miners have the same incentives, and not all will be motivated solely by direct profit (i.e. mining fees).  Some will have an indirect motive to mine, making transaction fees a secondary revenue stream.  You can search for my "Wal-Mart" posts related to this subject.

Because it's actually a problem.

You think that miners with not for profit motives are going to be able to maintain a robust, attack resistant network. Others disagree.

I can see this working if people voluntarily pay tx fees. If there are millions of users then they each wouldn't need to pay much to incentivise a strong network of block generators, and indeed, a culture of this sort may emerge. But is is a tragedy of the commons like situation.

In another post, you highlighted the fact that there is a transaction acceptance policy encoded in the main client which prioritieses transactions with fees, forcing fees out of merchants needing speedy confirmations, but there is no reason to believe that the mining pools of the future will maintain such a policy when there are fee paying transactions going un-harvested. Luke-Jr, to me, exemplifies the natural dynamic of transaction acceptance policy: accept txs with fees, don't accept txs without fees.

I know you understand this dynamic, however, you haven't made a compelling case as to why bitcoin will work anyway. In any case, I hope you're right. Although, I wish intelligent people on this forum would acknowledge this problem instead of stifling debate.
legendary
Activity: 1708
Merit: 1010
May 10, 2011, 12:20:13 AM
#32
This is just the default ruleset.

Miners are free to choose their own rules.  I believe luke-jr's pool requires all TX's to have a fee, if an extremely minimal one.

I disagree with his policy, but it illustrates the free market already at work.

If the Mining Majority wants a different ruleset, then the default ruleset won't be worth the paper it isn't printed on  Smiley

I'm not sure what conclusion you can really draw from luke-jr only accepting transactions with fees.  When there are no longer any generation rewards for a block, miners would all clearly benefit if they all only allow transactions with a certain minimum fee, however for any individual miner, they would have an incentive to accept any transaction so long as there is some fee...no matter how small that fee is.  This will drive transaction fees down to the point where miners start shutting down and the network becomes more vulnerable to a powerful miner.  Despite there being a strong collective reason for miners to charge some minimum fee and despite the clear benefit to all users to pay a fee to ensure a sufficiently robust network of miners, I think there is a real risk that miners will want to scoop up all transactions paying anything and for users to pay the bare minimum fee (a satoshi I guess).  This is the tragedy of commons argument people make.

Why does this keep coming up?

No, not all miners have the same incentives, and not all will be motivated solely by direct profit (i.e. mining fees).  Some will have an indirect motive to mine, making transaction fees a secondary revenue stream.  You can search for my "Wal-Mart" posts related to this subject.
administrator
Activity: 5222
Merit: 13032
May 09, 2011, 10:01:31 PM
#31
That got me thinking. If we can come up with a list of likely attributes of spam transactions, perhaps the client could identify the likelyhood of a transaction being spam. Then, if that likelyhood (expressed as a number between 0 and 254) is greater than a random number between 0 and 254, the transaction is thrown out as spam.

This is also how I think the spam problem should be solved for relaying. A drop probability should be weighted by fees and priority. If the transaction is important, it will be resent.
hero member
Activity: 868
Merit: 1008
May 09, 2011, 09:15:16 PM
#30
This is just the default ruleset.

Miners are free to choose their own rules.  I believe luke-jr's pool requires all TX's to have a fee, if an extremely minimal one.

I disagree with his policy, but it illustrates the free market already at work.

If the Mining Majority wants a different ruleset, then the default ruleset won't be worth the paper it isn't printed on  Smiley

I'm not sure what conclusion you can really draw from luke-jr only accepting transactions with fees.  When there are no longer any generation rewards for a block, miners would all clearly benefit if they all only allow transactions with a certain minimum fee, however for any individual miner, they would have an incentive to accept any transaction so long as there is some fee...no matter how small that fee is.  This will drive transaction fees down to the point where miners start shutting down and the network becomes more vulnerable to a powerful miner.  Despite there being a strong collective reason for miners to charge some minimum fee and despite the clear benefit to all users to pay a fee to ensure a sufficiently robust network of miners, I think there is a real risk that miners will want to scoop up all transactions paying anything and for users to pay the bare minimum fee (a satoshi I guess).  This is the tragedy of commons argument people make.
full member
Activity: 140
Merit: 101
May 09, 2011, 08:59:24 PM
#29
0.0005 sounds about right to me. At any rate 0.01 is now too high and it should be at most 0.001. (Of course in the long run fees should be a matter of individual policy, but there's nothing wrong with defaults.)
sr. member
Activity: 294
Merit: 252
May 09, 2011, 08:40:34 PM
#28
There should be better ways to identify a spam attempt and deny relaying them. But I see the point, it's much quicker to just reduce the limit.

That got me thinking. If we can come up with a list of likely attributes of spam transactions, perhaps the client could identify the likelyhood of a transaction being spam. Then, if that likelyhood (expressed as a number between 0 and 254) is greater than a random number between 0 and 254, the transaction is thrown out as spam.

but what if it isn't spam?

Since it's not deterministic, it's likely that unless it's almost certainly spam, some nodes would relay the transaction. They could take into account transaction fees, the size of the transaction, the number of recently seen transactions, etc.
full member
Activity: 126
Merit: 101
May 09, 2011, 07:57:21 PM
#27
but what if it isn't spam?

If in doubt, client could display a capcha  Grin
eof
full member
Activity: 156
Merit: 100
May 09, 2011, 07:49:04 PM
#26
There should be better ways to identify a spam attempt and deny relaying them. But I see the point, it's much quicker to just reduce the limit.

That got me thinking. If we can come up with a list of likely attributes of spam transactions, perhaps the client could identify the likelyhood of a transaction being spam. Then, if that likelyhood (expressed as a number between 0 and 254) is greater than a random number between 0 and 254, the transaction is thrown out as spam.

but what if it isn't spam?
sr. member
Activity: 294
Merit: 252
May 09, 2011, 06:34:13 PM
#25
There should be better ways to identify a spam attempt and deny relaying them. But I see the point, it's much quicker to just reduce the limit.

That got me thinking. If we can come up with a list of likely attributes of spam transactions, perhaps the client could identify the likelyhood of a transaction being spam. Then, if that likelyhood (expressed as a number between 0 and 254) is greater than a random number between 0 and 254, the transaction is thrown out as spam.
legendary
Activity: 1106
Merit: 1004
May 09, 2011, 03:39:28 PM
#24
There should be better ways to identify a spam attempt and deny relaying them. But I see the point, it's much quicker to just reduce the limit.
sr. member
Activity: 334
Merit: 250
May 09, 2011, 03:25:11 PM
#23
Please, just abolish this minimum fee. (If you remove this standard fee policy altogether it would be great too Wink)
Oh god no, in-discriminant transaction relaying means DDoSing becomes way too easy.

Agreed. .0005 sounds about right, maybe a little low. But if it appreciates more then should be just right.
hero member
Activity: 755
Merit: 515
May 09, 2011, 03:17:54 PM
#22
Please, just abolish this minimum fee. (If you remove this standard fee policy altogether it would be great too Wink)
Oh god no, in-discriminant transaction relaying means DDoSing becomes way too easy.
legendary
Activity: 1106
Merit: 1004
May 09, 2011, 03:14:40 PM
#21
Please, just abolish this minimum fee. (If you remove this standard fee policy altogether it would be great too Wink)
hero member
Activity: 755
Merit: 515
May 09, 2011, 02:54:04 PM
#20
Transactions with a fee below the minimum fee set by the relay node are treated in the same way as free transactions so they will be relayed if they meet the free transaction criteria.
Yes, but one of the largest motivations behind this was the option to relay transactions which are over the size limit and do not pay a fee and would normally not be relayed.  Remember that this really won't change what ends up in a block much as the very low priority transactions already don't get confirmed until they age.
donator
Activity: 826
Merit: 1060
May 09, 2011, 02:53:30 PM
#19
If the fee is realistic, the transaction value shouldn't matter (because it doesn't affect the processing costs).

Why not 0.001 BTC regardless of transaction value?

This fee has the advantage that it is one millibitcoin, and so can introduce users to this term. "Five hundred microbitcoins" is a little less simple.
legendary
Activity: 1708
Merit: 1010
May 09, 2011, 02:05:49 PM
#18

This is just the default ruleset.

Miners are free to choose their own rules.  I believe luke-jr's pool requires all TX's to have a fee, if an extremely minimal one.

And, last I heard, ArtForz's mining cluster is set to accept all free transactions regardless of the standard blocksize limit for free transactions, so it all evens out in the end.

My opinion is that the rule should be changed to drop the minimum fee in the default client and set back the minimum free transaction as well.  The massive growth we have seen recently has moved these limits from really tiny amounts to non-trivial limits, and the growth isn't likely over.
Pages:
Jump to: