Author

Topic: Incentive for Nodes through Economics (Read 789 times)

sr. member
Activity: 254
Merit: 1258
April 07, 2015, 10:57:25 AM
#15
I see the concerns with people that it could become a POS if they had a interest rate or something like that but would it be possible for miners that when miners mine they partition 1% of their hash power and it is shared to the full nodes or when a block is discovered 1% is given to the full nodes that broadcast it?

Would either of these be possible with a hardfork?

Everything is possible with a hardfork, the question "what is a full node" still stands. How to you distinguish proper full nodes from "half-ass" nodes that are just set up to get a piece of the pie.
They would have to contain a full copy of the blockchain and be open to at least 8 connections and have an up time requirement, I mean we aren't talking a ton of coin to be made here but just an bit of the piece of pie.

Wouldnt that make it profitable to snipe other nodes out of the network? If I can somehow (DDoS, etc.) reduce the number of eligible nodes in the network I would increase my part of the reward. This could result in an arms race of nodes trying to get each other unreachable by the rest of the network. This would possible reduce the number of full nodes available as the become a target.

Does not the same concern exists for miners too ? Like one miner trying to knockout other miners ?
People do DDOS pools, all the time.
sr. member
Activity: 254
Merit: 1258
April 07, 2015, 09:55:00 AM
#13
I see the concerns with people that it could become a POS if they had a interest rate or something like that but would it be possible for miners that when miners mine they partition 1% of their hash power and it is shared to the full nodes or when a block is discovered 1% is given to the full nodes that broadcast it?

Would either of these be possible with a hardfork?

Everything is possible with a hardfork, the question "what is a full node" still stands. How to you distinguish proper full nodes from "half-ass" nodes that are just set up to get a piece of the pie.
They would have to contain a full copy of the blockchain and be open to at least 8 connections and have an up time requirement, I mean we aren't talking a ton of coin to be made here but just an bit of the piece of pie.

Wouldnt that make it profitable to snipe other nodes out of the network? If I can somehow (DDoS, etc.) reduce the number of eligible nodes in the network I would increase my part of the reward. This could result in an arms race of nodes trying to get each other unreachable by the rest of the network. This would possible reduce the number of full nodes available as the become a target.
hmm that could be an issue but couldn't this be an issue with any sort of incentive placed on nodes except a interest based one?
copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
April 07, 2015, 11:07:25 AM
#12
Wouldnt that make it profitable to snipe other nodes out of the network? If I can somehow (DDoS, etc.) reduce the number of eligible nodes in the network I would increase my part of the reward. This could result in an arms race of nodes trying to get each other unreachable by the rest of the network. This would possible reduce the number of full nodes available as the become a target.

Does not the same concern exists for miners too ? Like one miner trying to knockout other miners ?
People do DDOS pools, all the time.

Id assume that DDoS among miners is somewhat normal, yes. But since a mining pool only needs a single connection to broadcast a new block its probably difficult to efficiently take a mining node out of the network for a long period of time. An uptime requirement of e.g. 24 hours would make it easy to take a node out of the network for said 24 hours. It would also be difficult to determine whether a node was actualy online over a certain period of time.

They would have to contain a full copy of the blockchain

How do you prove that? Node A and Node B could point to the same blockchain copy on a shared network resource.

Im not sure whether this is an issue, but it would certainly be possible to write custom node software that pretends to be a full node while its actually sharing a single copy of the blockchain with a high amount of other nodes. It might also be possible for a node to just relay block requests to other nodes in order to appear like a full node. This is certainly possible today, but since no one could benefit I doubt its a widespread problem.

---

Another thing that just came to me is: How would a mining node know which nodes to reward? Since the mining node is the one that creates the block it has to verify whether a certain node is eligible for a share of the reward.
legendary
Activity: 1662
Merit: 1050
April 07, 2015, 10:52:52 AM
#11
I see the concerns with people that it could become a POS if they had a interest rate or something like that but would it be possible for miners that when miners mine they partition 1% of their hash power and it is shared to the full nodes or when a block is discovered 1% is given to the full nodes that broadcast it?

Would either of these be possible with a hardfork?

Everything is possible with a hardfork, the question "what is a full node" still stands. How to you distinguish proper full nodes from "half-ass" nodes that are just set up to get a piece of the pie.
They would have to contain a full copy of the blockchain and be open to at least 8 connections and have an up time requirement, I mean we aren't talking a ton of coin to be made here but just an bit of the piece of pie.

Wouldnt that make it profitable to snipe other nodes out of the network? If I can somehow (DDoS, etc.) reduce the number of eligible nodes in the network I would increase my part of the reward. This could result in an arms race of nodes trying to get each other unreachable by the rest of the network. This would possible reduce the number of full nodes available as the become a target.

Does not the same concern exists for miners too ? Like one miner trying to knockout other miners ?
legendary
Activity: 1764
Merit: 1007
April 07, 2015, 09:53:26 AM
#10
They would have to contain a full copy of the blockchain

How do you prove that? Node A and Node B could point to the same blockchain copy on a shared network resource.
copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
April 07, 2015, 09:49:27 AM
#9
I see the concerns with people that it could become a POS if they had a interest rate or something like that but would it be possible for miners that when miners mine they partition 1% of their hash power and it is shared to the full nodes or when a block is discovered 1% is given to the full nodes that broadcast it?

Would either of these be possible with a hardfork?

Everything is possible with a hardfork, the question "what is a full node" still stands. How to you distinguish proper full nodes from "half-ass" nodes that are just set up to get a piece of the pie.
They would have to contain a full copy of the blockchain and be open to at least 8 connections and have an up time requirement, I mean we aren't talking a ton of coin to be made here but just an bit of the piece of pie.

Wouldnt that make it profitable to snipe other nodes out of the network? If I can somehow (DDoS, etc.) reduce the number of eligible nodes in the network I would increase my part of the reward. This could result in an arms race of nodes trying to get each other unreachable by the rest of the network. This would possible reduce the number of full nodes available as the become a target.
copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
April 07, 2015, 09:37:44 AM
#8
I see the concerns with people that it could become a POS if they had a interest rate or something like that but would it be possible for miners that when miners mine they partition 1% of their hash power and it is shared to the full nodes or when a block is discovered 1% is given to the full nodes that broadcast it?

Would either of these be possible with a hardfork?

Everything is possible with a hardfork, the question "what is a full node" still stands. How to you distinguish proper full nodes from "half-ass" nodes that are just set up to get a piece of the pie.
legendary
Activity: 1764
Merit: 1007
April 07, 2015, 09:15:53 AM
#7
Would like to know whether a full node can be identified distinguishably ?

It's called proof-of-work.
sr. member
Activity: 728
Merit: 256
April 07, 2015, 05:49:49 AM
#6
Define "full node". Sits on one machine? Has its own IP and bandwidth? Has its own blockchain copy? There could be lots of "half-ass" nodes that someone set up to use shared resources but that would try to be compensated as full nodes. How would you prevent that? There would need to be some kind of web-of-trust approach.

WoT is a failed approach that marries centralization. But "half-ass" nodes claiming full node compensation is a concern. Would like to know whether a full node can be identified distinguishably ?
legendary
Activity: 1764
Merit: 1007
April 07, 2015, 05:34:39 AM
#5
Define "full node". Sits on one machine? Has its own IP and bandwidth? Has its own blockchain copy? There could be lots of "half-ass" nodes that someone set up to use shared resources but that would try to be compensated as full nodes. How would you prevent that? There would need to be some kind of web-of-trust approach.
copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
April 07, 2015, 05:03:25 AM
#4
If such an incentive was available, you could have some form of interest on your wallet, not based on the value, but based on the uptime of your node. It is an interesting idea, but it needs some form of funding, possibly done by farmers. But how will they be pursueded to do such a thing. Enforcing it by the protocol could be misused as wel, what would stop me for setting up a node farm to rake in the extra bits?

I dont like the idea of interest on the wallet thats borderline PoS. It would also result in funded full nodes. I like that I can just run a full node on a cheap VPS and dont have to worry about my money, as there is none. The bitnodes incentive[1] is a good indicator that not many people have interest in running full nodes even when there is a chance to get paid for it. The number of nodes did not change significantly since they started the incentive. They still list ~6000 nodes. There are probably many others that are not accessible from the outside. Since those are not much help for the network anyway I dont think it matters much.

Btw Satoshi predicted that running a full node will be less common over time.


[1] https://getaddr.bitnodes.io/nodes/incentive/
sr. member
Activity: 462
Merit: 250
I can draw your avatar!
April 07, 2015, 04:44:57 AM
#3
If such an incentive was available, you could have some form of interest on your wallet, not based on the value, but based on the uptime of your node. It is an interesting idea, but it needs some form of funding, possibly done by farmers. But how will they be pursueded to do such a thing. Enforcing it by the protocol could be misused as wel, what would stop me for setting up a node farm to rake in the extra bits?
legendary
Activity: 3248
Merit: 1070
April 07, 2015, 04:37:55 AM
#2
it would be cool to have something like, as a myself i would run the client 0-24 if there was some incentive for that, it would eliminate the need to go for a pos system in the future(like many have suggested)

but payout should not be too small like faucets or few satoshi as usual, but in the range of 10k-100k, this could also encourage more people to adopt bitcoin and store bitcoin locally instead of using unsafe online wallet
legendary
Activity: 1204
Merit: 1002
RUM AND CARROTS: A PIRATE LIFE FOR ME
April 07, 2015, 03:49:16 AM
#1
I know the idea has been discussed before, making a protocol change to enable supporting full-nodes for their services, as well as the idea of making nodes bundle bit message relays, tor-exit nodes, etc....

Perhaps it's not necessary to make any protocol changes to make this happen- perhaps if we were to come up with a strong smart contract system, pools could voluntarily agree to share part of their earnings with the nodes. While it's not directly in their best interest, it is in their best interest to support a strong network. Indeed perhaps users could put enough pressure onto pools that they would willingly pledge some reward from each block they solve towards these pools.

As a user running two full nodes myself, (plus bitmessage) I don't particularly feel the need to be compensated, but I'm sure others do. It might just be nice to build a software bundle that automatically runs a full node + tor exit node + bittorrent seed of the blockchain + bitmessage (and any other services) and have a way for the software to verify that you were indeed running the software in case pools agreed to help compensate nodes.

Jump to: