Bitcoin Forum

Bitcoin => Development & Technical Discussion => Topic started by: z on August 13, 2010, 03:29:00 PM



Title: fraud > transaction fee
Post by: z on August 13, 2010, 03:29:00 PM
I don't see any way, how transaction fee can be more profitable then fraud.

It's always rational choice to accept transaction with fee, no matter how small it is, since user will not likely re-run the transaction with higher fee.

For example, suppose thet most blocks are generated with transaction fee 0.1%. As a generator it's always best for me to charge smaller fee, so I get 0.1% from last transactions and some more from older transactions with smaller fee.

This will make (computing power -> transaction fee) far less profitable, then (computing power -> fraud).


Title: Re: fraud > transaction fee
Post by: knightmb on August 13, 2010, 04:45:17 PM
That's a social problem, not a technical problem.  :-\


Title: Re: fraud > transaction fee
Post by: aceat64 on August 13, 2010, 04:51:39 PM
Running a business that provides services for bitcoins is also more profitable then generating blocks for transaction fees. I fail to see why this is an issue.


Title: Re: fraud > transaction fee
Post by: z on August 13, 2010, 05:27:41 PM
Isn't it like when you have more computing power then generators, then you can easily manipulate network? I state that current transaction fee system doesn't provide enough incentive to generate, so when inflation will slow down evil people would be able to easily take control over system.


Title: Re: fraud > transaction fee
Post by: knightmb on August 13, 2010, 05:35:45 PM
Isn't it like when you have more computing power then generators, then you can easily manipulate network? I state that current transaction fee system doesn't provide enough incentive to generate, so when inflation will slow down evil people would be able to easily take control over system.
You skip a lot of steps.

It would be me saying "When people are tired of government, I should be able to take over the world".


Title: Re: fraud > transaction fee
Post by: z on August 13, 2010, 05:50:24 PM
You skip a lot of steps.
if bitcoin market is gonna be worth $50 million and this will be possible, then they will invest in those steps and steal a few millions destroying the network. It might be even legal in some countries.


Title: Re: fraud > transaction fee
Post by: knightmb on August 13, 2010, 05:58:05 PM
You skip a lot of steps.
if bitcoin market is gonna be worth $50 million and this will be possible, then they will invest in those steps and steal a few millions destroying the network. It might be even legal in some countries.
You are depending on something that no one know exist yet though. That's it's possible to create an army of CPU for less than it will cost to steal those millions. I'm sure botnet operators would fall into that category, but if anyone owns that much CPU power, they would make more money stealing from the existing electronic currency market (think global stock market) than trying to take control of bicoin. Plus you forget that we are still here watching everything, as this is an open source community. If you can get the entire community to remain silent for a global bitcoin conspiracy, then bigger problems are at hand than someone cheating bitcoin.

Security through obscurity certainly isn't possible with bitcoin and that's it's strong point.


Title: Re: fraud > transaction fee
Post by: z on August 13, 2010, 06:17:19 PM
You are depending on something that no one know exist yet though.
On the contrary, you are depending on non-existence of something never proved to be non-existent. That should be avoided when possible.

You were writing about such scenario in technical paper:
Quote
The incentive may help encourage nodes to stay honest. If a greedy attacker is able to
assemble more CPU power than all the honest nodes, he would have to choose between using it
to defraud people by stealing back his payments, or using it to generate new coins. He ought to
find it more profitable to play by the rules, such rules that favour him with more new coins than
everyone else combined, than to undermine the system and the validity of his own wealth.
But this argumentation is flawed with current transaction fee system.


Title: Re: fraud > transaction fee
Post by: kiba on August 13, 2010, 06:26:45 PM
You are depending on something that no one know exist yet though.
On the contrary, you are depending on non-existence of something never proved to be non-existent. That should be avoided when possible.

You were writing about such scenario in technical paper:
Quote
The incentive may help encourage nodes to stay honest. If a greedy attacker is able to
assemble more CPU power than all the honest nodes, he would have to choose between using it
to defraud people by stealing back his payments, or using it to generate new coins. He ought to
find it more profitable to play by the rules, such rules that favour him with more new coins than
everyone else combined, than to undermine the system and the validity of his own wealth.
But this argumentation is flawed with current transaction fee system.

How it is flawed?


Title: Re: fraud > transaction fee
Post by: z on August 13, 2010, 06:36:46 PM
like I stated in first post. Transaction fees will go towards zero, so it will not be any profitable to charge them. It will give you some tiny part of bitcoin economy per year, while cheating will give you some significant part once, relatively quickly.


Title: Re: fraud > transaction fee
Post by: Insti on August 13, 2010, 07:03:22 PM
cheating will give you some significant part once, relatively quickly.

Please explain how.


Title: Re: fraud > transaction fee
Post by: Red on August 13, 2010, 07:25:29 PM
All processes for bitcoin cheating involve either:
1) Inserting a non-validated transaction into the block list. Which is easily spotted by anyone with the list, generating or not.
2) Removing previously validated changes from the list, which involves substuting the entire tail (most commonly scrutinized section) of the list.

All other theats are common concerns to all secure systems. (man in the middle, wallet stealing, cross site scripting)

One and two are really easy to add additional defenses too. In fact there are more existing defenses than are currently discussed. (hardcoded checksums, prohibitions on large branch switches.) as attacks are spotted it's easy to bump up defenses.


As such, this argument boils down to, "if there were a way to cheat it would be more profitable than not cheating." Duh! That is why we don't rely on incentives to prevent cheating!


Title: Re: fraud > transaction fee
Post by: FreeMoney on August 14, 2010, 03:47:57 AM
I don't see any way, how transaction fee can be more profitable then fraud.

It's always rational choice to accept transaction with fee, no matter how small it is, since user will not likely re-run the transaction with higher fee.

For example, suppose thet most blocks are generated with transaction fee 0.1%. As a generator it's always best for me to charge smaller fee, so I get 0.1% from last transactions and some more from older transactions with smaller fee.

This will make (computing power -> transaction fee) far less profitable, then (computing power -> fraud).

That's neat that you can compare them. Will you share with us how profitable each will be?


Title: Re: fraud > transaction fee
Post by: z on August 14, 2010, 01:43:44 PM
In short: I state that current transaction fee system provides almost none incentive for generating. That's my main argument. This means that power invested in generating will be very low. If it isn't obviously dangerous, then this thread is not important.


Title: Re: fraud > transaction fee
Post by: Red on August 14, 2010, 03:54:09 PM
It is written somewhere that after bitcoin generation stops, transaction fees will provide the main incentive for people to keep generating blocks and verifying transactions.

I've always thought that statement was silly and unnecessary. It should be deleted to avoid confusion.

Currently fees are charged to avoid "transaction spam" as identified in the source code. Also for verifying very complicated transactions. I think both policies are simply deterrents for known annoying behaviors. They are not intentional incentives.

The main reason people will continue to validate blocks into the future absent new coin generation, is personal self interest. If you hold lots of value in bitcoins and you don't defend them, no one will. The people with the most coins have the highest incentive to make sure the system remains safe and secure.
 


Title: Re: fraud > transaction fee
Post by: z on August 15, 2010, 08:05:21 PM
The main reason people will continue to validate blocks into the future absent new coin generation, is personal self interest. If you hold lots of value in bitcoins and you don't defend them, no one will. The people with the most coins have the highest incentive to make sure the system remains safe and secure.
1. It contradicts many statements on the site and in technical paper.
2. It sounds very idealistic. It's not rational choice for agent to generate: if there are many generators, then you won't change much; otherwise it's not safe to keep capital in bitcoins, better to sell them. Even when it will work (what I don't belive), it's really not fair, since many people use bitcoins, but only some of them pay for it. I think it will be much better to have well-designed system providing compensation for generators, based either on inflation or fees. This system should ensure some economical equilibrium, which guarantee that quoted fragment from technical paper is true.


Title: Re: fraud > transaction fee
Post by: Red on August 15, 2010, 09:23:16 PM
1. It contradicts many statements on the site and in technical paper.
2. It sounds very idealistic. It's not rational choice for agent to generate: if there are many generators, then you won't change much; otherwise it's not safe to keep capital in bitcoins, better to sell them. Even when it will work (what I don't belive), it's really not fair, since many people use bitcoins, but only some of them pay for it. I think it will be much better to have well-designed system providing compensation for generators, based either on inflation or fees. This system should ensure some economical equilibrium, which guarantee that quoted fragment from technical paper is true.

There exist several people with large amounts of value stored in bitcoins now. They also run business using bitcoins so they have a vested interest in protecting the process and the trustworthiness of the system now and into the future. They also run significant amounts of CPU generation, and watch the blocklist for fraud. If you would like, I can introduce them to you.

So NO. I don't accept your assertion that these people do not/will not exist. Nor do I accept your assertion that these people are not rational. Take that up with them.

If you would like to assert that *you* would be more willing to validate transactions if people will pay you better, then knock yourself out.