ppcoin transaction fees

[Jutarul]

In ppcoin transaction fees are not redistributed to the miners, but instead "destroyed". This destruction is compensated by the influx of new money, either through POS or POW blocks. This is a drastic change of the incentive structure for validation nodes and may become a problem, as explained in the following:

In bitcoin, the transaction fee acts as an incentive for miners to include transactions. The first goal of transaction fees is to accomplish spam prevention. The second, more complex aspect is the incentive for validation nodes: By providing a transaction fee, a validation node can earn money by including the transaction in the next block. If the block gets crowded, the validation node can prioritize transactions based transaction fees. (However, that incentive structure posses its own problems, since nodes may have an incentive to WITHHOLD transactions with large fees from other nodes)

In contrast in ppcoin, there is no direct incentive to include transactions into blocks. Thus the mechanism to determine transaction priority may vary dramatically throughout validation nodes. A person creating a transaction has thus NO DIRECT MEANS to influence the priority of the transaction. This can become a problem if the transaction limit in ppcoin is reached.

This is a QOS (quality of service) problem for transactions in the ppcoin network. Users have to be able to increase the priority of their transactions by providing monetary incentive.

[AndyRossy]

In ppcoin transaction fees are not redistributed to the miners, but instead "destroyed". This destruction is compensated by the influx of new money, either through POS or POW blocks. This is a drastic change of the incentive structure for validation nodes and may become a problem, as explained in the following:

In bitcoin, the transaction fee acts as an incentive for miners to include transactions. The first goal of transaction fees is to accomplish spam prevention. The second, more complex aspect is the incentive for validation nodes: By providing a transaction fee, a validation node can earn money by including the transaction in the next block. If the block gets crowded, the validation node can prioritize transactions based transaction fees. (However, that incentive structure posses its own problems, since nodes may have an incentive to WITHHOLD transactions with large fees from other nodes)

In contrast in ppcoin, there is no direct incentive to include transactions into blocks. Thus the mechanism to determine transaction priority may vary dramatically throughout validation nodes. A person creating a transaction has thus NO DIRECT MEANS to influence the priority of the transaction. This can become a problem if the transaction limit in ppcoin is reached.

This is a QOS (quality of service) problem for transactions in the ppcoin network. Users have to be able to increase the priority of their transactions by providing monetary incentive.

Would it not be possible, to, change so that a min tx of 1c is required (as opposed to fixed at this), and then 0.1/tx is destroyed? As such, allowing clients to still give a tx fee to the validating node, if they so desire, for priority?

[Etlase2]

This also raises the issue of what kind of consensus on tx fees could possibly be reached if they are irrelevant to miners. Is the tx fee hardcoded? If so, how can that possibly work when you have no idea what a ppcoin is worth? Perhaps another $100k in development costs are required to find the solution.

[Sunny King]

This also raises the issue of what kind of consensus on tx fees could possibly be reached if they are irrelevant to miners. Is the tx fee hardcoded? If so, how can that possibly work when you have no idea what a ppcoin is worth? Perhaps another $100k in development costs are required to find the solution.

I used to take you seriously, and thought about reviewing your complex designs somewhere down the road and provide help as what I can. So my mistake. I'd say good luck getting anything done with your beloved decrits now with this type of attitude.

[smoothie]

This also raises the issue of what kind of consensus on tx fees could possibly be reached if they are irrelevant to miners. Is the tx fee hardcoded? If so, how can that possibly work when you have no idea what a ppcoin is worth? Perhaps another $100k in development costs are required to find the solution.

I used to take you seriously, and thought about reviewing your complex designs somewhere down the road and provide help as what I can. So my mistake. I'd say good luck getting anything done with your beloved decrits now with this type of attitude.

LOL I agree with Etlase your claim of putting in $100k-$200k time into development of PPC is laughable. $100k is a really good software engineer yearly salary. You haven't even been working on it that long from what I take. And if you have certainly not likely that you put in that much time.

Stop whining about people calling out your bullshit as what you would call an "attitude".

 

[Sunny King]

In ppcoin transaction fees are not redistributed to the miners, but instead "destroyed". This destruction is compensated by the influx of new money, either through POS or POW blocks. This is a drastic change of the incentive structure for validation nodes and may become a problem, as explained in the following:

In bitcoin, the transaction fee acts as an incentive for miners to include transactions. The first goal of transaction fees is to accomplish spam prevention. The second, more complex aspect is the incentive for validation nodes: By providing a transaction fee, a validation node can earn money by including the transaction in the next block. If the block gets crowded, the validation node can prioritize transactions based transaction fees. (However, that incentive structure posses its own problems, since nodes may have an incentive to WITHHOLD transactions with large fees from other nodes)

In contrast in ppcoin, there is no direct incentive to include transactions into blocks. Thus the mechanism to determine transaction priority may vary dramatically throughout validation nodes. A person creating a transaction has thus NO DIRECT MEANS to influence the priority of the transaction. This can become a problem if the transaction limit in ppcoin is reached.

This is a QOS (quality of service) problem for transactions in the ppcoin network. Users have to be able to increase the priority of their transactions by providing monetary incentive.

Would it not be possible, to, change so that a min tx of 1c is required (as opposed to fixed at this), and then 0.1/tx is destroyed? As such, allowing clients to still give a tx fee to the validating node, if they so desire, for priority?

This is inaccurate, you can still raise your paytxfee to get a better chance of getting into a block. The prioritization of transactions have not been changed from Bitcoin. 1c is the minimum required fee. If the argument is about miners have no incentive to include transactions, I think this concern is false as including transactions does not consume much resources and miners have no incentive to change the default client behavior to exclude transactions.

If transaction fees were paid to minters it would become a serious problem down the road as we described in our design paper.

[Etlase2]

Ohnoes the invisible carrot on the invisible stick, again. Learn to take a joke, brah, or don't fuel the fire.

[smoothie]

Ohnoes the invisible carrot on the invisible stick, again. Learn to take a joke, brah, or don't fuel the fire.

+1 LOL this ^

[Sunny King]

Ohnoes the invisible carrot on the invisible stick, again. Learn to take a joke, brah, or don't fuel the fire.

I am only putting my honest opinion of the work. If you cannot take it or think the type of work in ppcoin does not worth the amount I estimated, I would say that's probably a good indication that you are not anywhere close to a true professional level.

I would also rate the programming job of Satoshi worth $1m (not including his innovations just the quality of programming) and annual programming job of Bitcoin development team around $500k - $1m as well.

If you have no idea how much is the cost of software programming for critical infrastructure work well you are probably just another armchair designers in my honest opinion.

[smoothie]

Ohnoes the invisible carrot on the invisible stick, again. Learn to take a joke, brah, or don't fuel the fire.

I am only putting my honest opinion of the work. If you cannot take it or think the type of work in ppcoin does not worth the amount I estimated, I would say that's probably a good indication that you are not anywhere close to a true professional level.

I would also rate the programming job of Satoshi worth $1m (not including his innovations just the quality of programming) and annual programming job of Bitcoin development team around $500k - $1m as well.

If you have no idea how much is the cost of software programming for critical infrastructure work well you are probably just another armchair designers in my honest opinion.

I do. I am a software engineer. Your estimates of how much "time" you put into it as $100k-$200k is bullshit.

[Etlase2]

I am only putting my honest opinion of the work. If you cannot take it or think the type of work in ppcoin does not worth the amount I estimated, I would say that's probably a good indication that you are not anywhere close to a true professional level.

As I said, learn to take a joke, brah. Besides, after only being interested in your project and asking honest questions, the reply I got from you was a pompous one in a different thread. You haven't set the bar very high.

PS - There was a miner a short while back that had a significant amount of hashing power but included no transactions. There is a very real downside to including transactions and that is network propagation speed. There is an orphaned block in bitcoin every couple dozen blocks or so, so this is a very real disincentive to including transactions. However, I think you might have mentioned something somewhere about coin-days-destroyed being a factor, but I can't be bothered to remember if I am right or not, and you haven't really responded to the concerns of the thread over taking my comment personally.

[markm]

Absolutely, if you don't get paid to include transactions you are wasting joules by doing so, unless enough other people are not, or you are a large enough proportion of the processing, so that enough lag in getting a transaction done is noticed to devalue the coin as not providing sufficient quality of service.

If you can save joules while still getting paid the same, you probably should save the joules.

-MarkM-

[Sunny King]

I am only putting my honest opinion of the work. If you cannot take it or think the type of work in ppcoin does not worth the amount I estimated, I would say that's probably a good indication that you are not anywhere close to a true professional level.

As I said, learn to take a joke, brah. Besides, after only being interested in your project and asking honest questions, the reply I got from you was a pompous one in a different thread. You haven't set the bar very high.

PS - There was a miner a short while back that had a significant amount of hashing power but included no transactions. There is a very real downside to including transactions and that is network propagation speed. There is an orphaned block in bitcoin every couple dozen blocks or so, so this is a very real disincentive to including transactions. However, I think you might have mentioned something somewhere about coin-days-destroyed being a factor, but I can't be bothered to remember if I am right or not, and you haven't really responded to the concerns of the thread over taking my comment personally.

I already responded to the questions above.  As to your argument here I think it probably does not apply to proof-of-stake blocks as you don't really lose that much if your proof-of-stake block gets reorganized.

Even with proof-of-work blocks I think the concern is still minimal. I don't think a significant portion of nodes would alter the implicit network protocol contract with that type of minimal gains. Yes some extremely selfish guys may, but not really a big problem in the whole picture.

PS I just speak my honest opinions so if that time you were not comfortable with it I apologize. The other time when you critique QBC I praised you as well since your critique were quite spot-on.

[AndyRossy]

Etlase2 and smoothie, why dont you guys go and try and trash another thread, or go give LTC threads some trash.

[Jutarul]

In ppcoin transaction fees are not redistributed to the miners, but instead "destroyed". .... A person creating a transaction has thus NO DIRECT MEANS to influence the priority of the transaction. ...This is a QOS (quality of service) problem for transactions in the ppcoin network.

...change so that a min tx of 1c is required ... and then 0.1/tx is destroyed? As such, allowing clients to still give a tx fee to the validating node, if they so desire, for priority?

This is inaccurate, you can still raise your paytxfee to get a better chance of getting into a block. The prioritization of transactions have not been changed from Bitcoin. 1c is the minimum required fee. If the argument is about miners have no incentive to include transactions, I think this concern is false as including transactions does not consume much resources and miners have no incentive to change the default client behavior to exclude transactions.

If transaction fees were paid to minters it would become a serious problem down the road as we described in our design paper.

I recommend to take the discussion about the value of the design work to PMs.

on the topic:
Please help me to understand the prioritization schedule for transactions and how that can be enforced on a protocol level (as you indicated in the design paper). As far as I understand it, right now validation nodes have neither an incentive nor a disincentive to include transactions. If the number of transactions hit the transaction limit, how can a user enforce a higher priority? In bitcoin this is achieved by providing a monetary incentive to the validation nodes.
Can you think of any other means of implementing an priority system for transactions, which does not require monetary incentives for validation nodes?
 

[Jutarul]

I don't think a significant portion of nodes would alter the implicit network protocol contract with that type of minimal gains. Yes some extremely selfish guys may, but not really a big problem in the whole picture.

I urge you to revise that opinion. You should always assume the worst case: 100% of selfish participants who are trying to exploit the shit out of the existing infrastructure.

[Etlase2]

Here are the minimal gains:

1) 0 joules > X joules (as markm said) - probably not a very big one as I think CPUs will outdo network growth in this area, but ECDSA verification is a very slow operation
2) faster block prop > slower block prop
3) less incentive to actually participate in the network via bandwidth costs which I think will always be a problem (this especially applies to PoS which very likely won't be pooled)

I think it's a bit more forward-thinking to address these concerns rather than dismissing them as minimal. It is hard to predict the future.

Plus you didn't answer if the 1c tx cost was hard-coded or can be changed by the protocol. If it's hard-coded, that's an awfully blunt-force way to set the tx fee. What if 1c is $1 in the future? A hard fork would be required to be competitive with other currencies.

[Sunny King]

This is inaccurate, you can still raise your paytxfee to get a better chance of getting into a block. The prioritization of transactions have not been changed from Bitcoin. 1c is the minimum required fee. If the argument is about miners have no incentive to include transactions, I think this concern is false as including transactions does not consume much resources and miners have no incentive to change the default client behavior to exclude transactions.

If transaction fees were paid to minters it would become a serious problem down the road as we described in our design paper.

I recommend to take the discussion about the value of the design work to PMs.

on the topic:
Please help me understanding the prioritization schedule for transactions and how that can be enforced on a protocol level (as you indicated in the design paper). As far as I understand it, right now validation nodes have neither an incentive nor a disincentive to include transactions. If the number of transactions hit the transaction limit, how can a user enforce a higher priority? In bitcoin this is achieved by providing a monetary incentive to the validation nodes.
Can you think of any other means of implementing an priority system for transactions, which does not require monetary incentives for validation nodes?

These are two separate issues. Including transactions is default client behavior and an implicit protocol but not enforced.

When transaction volume gets to the point there is competition getting into a block, ppcoin has the same Bitcoin logic that above half megabyte region of the block only higher and higher fee paying transactions are included. This is again implicit protocol and default client behavior and not enforced.

What I mean implicit protocol is that any benign alternative client software should follow the same behavior.

[Sunny King]

Yes I understand all your concerns. If it becomes a widespread problem down the road (although I highly doubt) then protocol will be revised to correct the problem.

Adjusting transaction fee would require a hard-fork yes. I am not as strict as Gavin on the issues of hard-fork. I believe even if you design a convoluted transaction fee model most likely hard fork would be necessary somewhere else anyway.

[smoothie]

Etlase2 and smoothie, why dont you guys go and try and trash another thread, or go give LTC threads some trash.

Hey if I ask a question and there isn't an answer to it besides bullshit propaganda I will keep asking the question.

Obviously your "leader" sunny KING doesn't know the average salary of a software engineer in making his $100k-$200k  appraisal of his L33T programming skills.

Seriously even Coinshiiter talked so highly about how much he was paid as a software enginner/programmer.

AndyRossy, if you don't like it keep crying okay?

Thanks

Your best friend,

Smoothie