Topic: Re: Fees for full nodes? (Read 8464 times)

No chances either.

How about the dash model? It follows the argument that people invested should run a node.

Here's a thread on it in the mining subsection where it is explained:
https://bitcointalksearch.org/topic/empty-blocks-1085800

Can you help me understand how it is possible to produce empty blocks, or what the incentive would even be to do so, when there are still unconfirmed transactions in the mempool?

Look at this picture.
Can you prove that F2Pool, AntPool and Slush are running full-nodes, not SPV? 

387615 https://blockchain.info/block/0000000000000000044e3a71cdb513d85d892313fada6905ccb675778857e86b
387616 https://blockchain.info/block/00000000000000000302c0857d3f71ef4f264ff95c99a058170256a1f330f07d
387617 https://blockchain.info/block/0000000000000000099b095a982f5987017885e033d1c3b39296b8b552b696e9

This is a distinction without a difference.  An SPV mining node can always produce blocks containing valid transactions.  All it needs is a collection of previously valid transactions using keys that it controls. This will effectively hide the fact that it is an effectively empty block.  There will no risk that any of the transactions in this block will prove invalid.  Of course, if the previous block contains invalid transactions there will still be a risk that this otherwise valid block will be orphaned.

I think that the science that says "no known solution" is still young in the timeline scale of humanity. As they say: necessity is a mother of invention. I'm willing to wait for either a better solutions or a better proof of incompleteness.

https://en.wikipedia.org/wiki/Tragedy_of_the_commons
There is no known solution to it. Only centralization.

OK, now I understand that you have an oddball sense of humor.

The "full node" is traditionally defined as a node that:

1) knows the entire blockchain history
2) propagates transactions and blocks after verification

The "pro-pools" legacy meaning included one more function:

3) mines blocks

Bitcoin started in the situation where mined coins paid for all 3 things:
3) electricity spent on PoW
2) network bandwidth
1) block storage

Now that node specialization is progressing I think the first two functions will get unbundled too. I thus don't consider "full node" a very important concept. Perhaps it would be better to split it into 2 or 3 terms:
i) real-time propagating node
ii) UTxO query-service node
iii) archival query-service node
depending on how far we would want to go into unbundling.

The problem with unbundling is that there's nobody paying the fourth cost:

0) research and development

Therefore after mining got unbundled the developers were left to funding themselves by extortion,  scams, rip-offs, vulture/venture capital, preying on the naïve and other miscellaneous funding sources.

There are chances that a SPV Miner produces valid empty blocks. A full node should be able to produce valid blocks including transactions.

If mining in SPV-mode produces valid blocks - it can be treated as holding full-node.
I do not see any problems here.

I don't know if you are joking or forgot about SPV mining. It isn't that simple.

Sorry, Meant to put:   http://web.media.mit.edu/~cebrian/p78-tang.pdf

there is a way to prove that you are running the full node.
there is only one way.
this way is well-known for 6 years.
the answer is: mining.
just create and broadcast new valid block and this would be a proof that you are running full node.

Running a full node is fully hectic nowadays.

Well, I may do it as long as I get some conventional fees for that.

I think running a full node should be implemented in recent miners so that the miners will perform mining as well as running the nodes simultaneously.

I don't know if this is possible though!

so there is no way to prove the full node I run is real?

This thread is getting hard to follow. There seem to be various mixups going on.

Altchain is better terminology than altcoin. It is like subordinate chain or side chain, but ultimately pays in the coin of the mainchain.

"nodes don't mine" is more or less temporary. After few more knees on the distribution curve the incentives will align more toward propagating transactions instead of trying to snipe the coinbase with a small block.

I think that this opinion seems to miss the that there are two p2p protocols within the Bitcoin network:

a) real-time transaction and block propagation network
b) historical block synchronization network

Currently they are awkwardly rolled into a single p2p protocol, with a branch appearing fairly recently "relay network" that concentrates on task (a). The task (b) is currently badly served by the original legacy protocol. Sooner or later the Bitconin developers will understand that they need to import more features from the well developed bulk transfer protocols like Bittorrent and maybe even from paid on-demand streaming. "headers first" isn't a real solution, more of a temporary workaround. Certainly it doesn't subsume the bulk-transfer functionality of a better-engineered protocol like bittorrent.

Pseudonode cheats only on the task (b) above, I think it does decent job for task (a). Other than this I don't really understand the above comment.

The current situation in the Bitcoin brain trust is that almost everyone suffers from some sort of brain constipation about how to change a single line of code
Ultimately this problem is going to solve itself in some way and people will again resume working on other long-term issues.

You're talking to the wrong person. I wasn't pushing for incentivising anything; only pointing out the flaw with the idea.

How about something like this? Miners send a portion of the mining fee to the nodes that sent them the transactions in the block they solved.

There's no symbiosis in that.

It's simply:

"I have a txn I want it in a block"
OK that happens.

"I want it in a block faster"
Then you need to transmit it to miners faster and (usually) give it a higher fee if you want to be prioritised over earlier txns.

Perhaps you meant http://research.microsoft.com/pubs/156072/bitcoin.pdf for the second.

Hmm, both links point to the same place.

This is incorrect. Although, you can't incentivize the process of just running a node, you have to incentivize both the speed and volume of the process of tx propagation.


http://research.microsoft.com/pubs/172840/Bitcoin-Red-Balloons-SIGEcom.pdf

Read that paper to learn more

And here's a reflection on it by MIT Media Lab

http://web.media.mit.edu/~cebrian/p78-tang.pdf

Node incentives are important, and likely play a role in dismissing certain concerns regarding issues inherent in larger block sizes by potentially representing a more Symbiotic relationship between miners and the nodes that propagate the transactions to them.

Reading this thread just influenced me to better understand how to support Bitcoin.  Been mining but that's like playing the lotto.

I have a Node setup now

75.187.165.159:8333 (dynamic)

and i am setting up one in NY (brother is on a military base)
>remoted in to his PC

I am pretty sure that if we can address node incentivization, we need a better incentive for users to run nodes instead of relying solely on altruism". Then we will have an easier time putting concerns about raising the block size at rest. The solution could present a Symbiotic relationship between nodes and miners.

I'm still wondering what 3rd world internet country people live in and say they can't run a full node?

I live in Aus and the internet here sux so bad I think I must be in the 3rd world.
I pay $110/mth for speed: 8Mbit down and 1Mbit up (with a data limit up+down of 1Tbyte)
I'm broke ... I'll admit it
I get by on that with all the net I use and I also run:
2 full nodes at home, one for my wallet and one for the pool wallet.

I run 'a few' full nodes on the net for the pool.

I gotta wonder about comments further up linking to reddit saying they can't run a full node ...

That's a bit hard to do when their are things like pseudonode.

Bitnodes uses a scoring system https://bitnodes.21.co/nodes/leaderboard/#peer-index.  Not sure how close that gets to identifying full nodes.

What about a lottery system instead of guaranteed payment?  Only way I can imagine this working is as an Etherum contract so it is verified how the registration and payouts work.

Some organization would need to fund it like Bitnodes did with incentive program, or you could have weekly/monthly entries.  Costs xxx bitcoin to register your node (IP and payout address).  Your node has to have a certain PIX score before you can qualify.   i think that would be interesting and may motivate some people to setup a node.  Lot of people like gambling

As far as I can tell, this point by -ck is sorta a showstopper for the idea of paying full nodes.  As he says, unless we could guarantee people aren't cheating, surely people would take advantage of that.  There may indeed be some way to prove this, but I think it's definitely a requirement going forward.

p2pool is an alt chain of its own and it has miners. Bitcoin nodes don't mine though.

nuli   good

Why altcoin? It is more like p2ppool, but instead of shares for mining one would get shares for proven transaction propagation.

You can wish all you like but there's absolutely no way to confirm a full node really is a full node and not just faking it. If full nodes were to get paid, then one could spoof as many full nodes as they like and get paid for each of their fake nodes. To prove they're full nodes you'd have to invent some kind of "proof of work" and add it to some kind of blockchain - congratulations you've invented mining and an altcoin blockchain.

Read "Red Balloons" by Zahar et al

I believe that if we can address node incentivization, like Satoshi suggested, "[I suspect] we need a better incentive for users to run nodes instead of relying solely on altruism". (IN AUGUST 2015)  Then we will have an easier time putting concerns about raising the block size at rest.

Perhaps the solution could present a Symbiotic relationship between nodes and miners, that would help with concerns regarding the propagation of larger blocks.

was

Full nodes need to get paid. Pruning can be definitely done to reduce costs. We do also need to encourage non pruned nodes as they are needed to help newer nodes to join the network

With blockexplorers like blockchain.info giving problems. I actually think it would be a good strategy to turn your full node into a blockexplorer. We can have a distributed network of blockexplorers and I think people would pay to use the blockexplorer / the API(having low downtime)

https://github.com/priestc/moneywagon is a good initiative to develop a universal blockchain API

From the Bitcoin-core Debug Window Peers;

73.179.230.242:59060
via x.x.x.x:8333

Direction            Inbound
Version              70002
User Agent         /Satoshi:0.11.2/
Services             NETWORK
Starting Height   237292
Sync Height        Unknown
Ban Score           0
Connection Time  10 h 4 m 39 s
Last Sent            5 s
Last Received      38 s
Bytes Sent          1 GB
Bytes Received    3 MB
Ping Time           336874 ms
Time Offset         1 s

That is a prime example of a peer I would like to charge for providing my full node services to it.

XT nodes were aborted stillborn by small block terrorists.

FWIW, he didn't say 1BTC per month, he says give him 1BTC and he'll run it as long as possible.  From what I can tell, he left the value of "as long as possible" underspecified.

ohh really?...and because of 0.5$ per montht even core developer publicly admit, that he is not running full node anymore..

"Block size has become too large for me to keep my home node running 24/7 anymore (it was affecting my phone call and videoconferencing quality too much). Now I just run it when I want to use Bitcoin."

source: http://bit.ly/1LGI9XG

1 BTC/month?  Are you kidding me?  That node probably costs you maybe $0.50/month to run, if you even notice a higher cost of living due to running it.

Myself and clients combined have dozens of full nodes running across the internet.  The $$$ to run a node never factors into anything.

You can setup your node to only use tor.  See this page for more details.

Changing the consensus rules is very difficult.
Successful examples called "revolutions"

It seems that the XT support is much lower than 50%. Maybe there will be no move to XT.

1. create a hardfork done.
2. move majority to it

We've completed 50% of our goals!

Fine with me, but the mechanism is the same. Everbody is free to use the altcoin or hardfork he wants.

I had a full node, i recieved DDoS all weeks... i dont like that mi ip are visible, if i could set on tor ip i could set my node again

Let us define some terms.
Altcoin - is a crypto with unique genesis block.
Hardfork - is crypto with the same genesis block but new rules.
This is not altcoin. This is hardfork, similar to BIP100/101/102...

There is a big number of coins which were lost because their owners forgot about them, lost privkeys, moved to provable unspendable outputs, etc.
And at least we would know - is Satoshi still has ability to move his coins? 

This rule doesn't change anything. All that will happen is people will move their coins from address to a new address once every 6 years.

Yes, everybody is able create an altcoin and everybody is free to use an altcoin created by others. Nothing new here.

This scenario is possible by changing current consensus rules.

In few words. The current code is something like (I am not sure, but do not care):

reward = ( 50 / ( 2^{( block_height / 210000 )}) ) + fees

You can suggest to change this formula to

reward = ( 50 / ( 2^{( block_height / 210000 )}) ) + fees + sum_of_all_utxo_from_block ( blockheight - 420000 )
and all utxo with the age of 420000 mark as non-spendable.

I repeat. This looks like a joke. But this is not a joke. The majority can implement any rules. And this rule is not worse (and not better) than other.

That will save a lot of money. First it will reduce the requirement for our hard drive. Second it will reduce the supply of bitcoin by deleting old coins.

Let us prune the 6 year old utxo! 

LOL
Of course, the majority can change everything.
But do not say "we can" without proving it.

We can change bitcoin protocol so that 0.1 bitcoin extra is issued per day. That is the small price to pay for the security of bitcoin.

People have the option to carry many different credit cards. The card issuing businesses make a point of giving away their payment tools. History tells them that people will make use of their credit cards.

Ignoring the desirability of using credit or credit cards, making Bitcoin easy to use and removing as many cost barriers to entry as possible should always be a priority. Any trade-offs on free to use or very low cost to use need to have a good reason to exist.

Who knows, splitting transaction fees between nodes and miners might be an option. I doubt the original design considered that the function of running a node would be separated from mining, so it would be an option to realign the original design to some extent.

I am invested and do indeed run a full node despite the lack of compensation.  If I sell off my investment then I will stop unless it is made to be worth my while.

*If* I were going to be compensated for running a full node then I would gladly pay other full nodes for the services they provide me.  If the full nodes are just going to pay each other then that is way less interesting than having SVP clients, etc., pay for services.

Ah, how about if we "enhance" the code to limit the amount of services provided proportional to the amount of services received?

Where do these rewards come from?

All nodes should be rewarded in some way, such as 0.0001 Bitcoin per day or 0.1bitcoin shared by all the nodes.

IMO, those who makes profit from bitcoin will have the biggest incentive to setup enterprise level full nodes: Exchanges, mining pools, mining farms, ASIC producers, wallet providers, payment processors, etc...  even those heavily invested in bitcoin should have some full node setup to indirectly support their investment

You can operate a full node* with pruning, what you're going on to describe there is UTXO commitments which are totally unneeded for pruning and are of unclear cost, unclear value, and unclear implications.


Unclear cost: rehashing the utxo set for every block, even if maintained in a high overhead efficient data structure is a multiplicative IO cost on running a full node.  Making the initial cost cheaper is essential, but at what runtime penalty is it acceptable?  is 20x acceptable?  If the utxo set is constantly changing, how do people synchronize it from you?  do you have to hold old versions? does a peer have to fetch it only from a single other peer?

Unclear value:  starting off a utxo commitment means you have only SPV security. If you were happy with SPV security; why not run SPV in the first place and dispense with the intermediate step. The answer is that it's only SPV of the history, but where is the dividing line? When someone today says how do you know that the system's creator didn't secretly mint himself a trillion coins-- the answer is because your own node verified it wasn't so.

Unclear implications: with no incentive to not run a committed bootstrapped node, is there any reason to believe the old history would even be _available_ to someone who wanted to do a full security initialization?

(*And in Git master Bitcoin Core relays blocks and transactions, and gives full wallet functionality, minus rescan when pruned.)

Currently, pruning is not the solution. You should be able to operate a full node with pruning. That means you need to be able to exchange UTXO information and an UTXO hash has to be stored in the block header.

I do think the discussion is interesting.  In the original description of Bitcoin, everyone who has a wallet is running a node.  Clearly this wasn't practical as the blockchain continues to scale and people wanted to run wallet software on their phones and cetera.  It would be cool if somehow the miners were tipping the nodes or something like that.  Perhaps this is akin to a waiter tipping out their busboy.  I dunno.  I guess I'm just here to follow along with the discussion for now.

I do believe in Bitcoin and want to see it flourish; I'm going to continue to run my full node despite no payment.  I was just hoping to get people to think about and discuss the topic.

I'm currently running a full node and it's taking approximately 1% CPU and approximately 80MB of RAM (obviously these numbers fluctuate).   My incentive for running it is that I want to strengthen the network.  This is more than just a "geeky good feeling".  If you're a part of a community, and you want that community to do well, and that community succeeds, then your efforts are rewarded.  I'm not saying that I don't want tips, I'm just saying that posting a topic to beg for a tip in return for keeping your node going is probably not going to result in a tip, at least I don't expect it to.  Good luck tho.

One is being developed. It may take a while to implement it, but every journey begins with a first step:

http://spreadcoin.info/news/proof-of-bitcoin-node-white-paper/

Then maybe in the future we would need something like bitnodes incetive program, but at a large scale.

A protocol level change to incentivize nodes would be very difficult, if even possible, for so many reasons.  Anything with incentive will be hit hard by scam attempts.  Also, consensus for changes becomes more difficult as bitcoin project grows.  You can be sure that any BIP to incentivize full nodes would be debated as much as changing block size limits has been.  I don't see it happening, personally.

FWIW, the bitnodes incentive program is now going to be continued past the original expiration date, thanks to 21.

Well, enable pruning and get rid of all the useless history!
(It's like cleaning your garage and throw away all the stuff you didn't use last 10 years)
Let someone else keep the full chain for future data archeology.

 
 
I was just about to make a topic about this when I saw this one. 
 
We absolutely need to compensate full nodes. To not do so is as foolish as to say "people will mine for the good of the network; no need to compensate miners".

I've been running a full node for quite a while now.  It runs me about $90/80 euros per year, it is usually in the top 10-15 on that leaderboard, although I am not sure if that is at all relevant or very useful. (4 Cores, 4GB RAM (5GB burst), 150GB RAID10, etc). Running a full node doesn't have to be expensive and it doesn't have to be difficult.  Just good to keep security updates applied regularly.

I don't believe that the fees would impact my decision whether to run it one way or another though.  Sure, if it was on my phone (give it 5 years and a phone might be able to handle it), but unless the fees are more significant, for me, they wouldn't play a role in me running or not running a node.  I'm sure not everyone would have the same opinion.

After 5 years, I may keep it going or I may not.  It isn't the cost, it is just a question of time to check it every so often and the lack of fees doesn't play a role.

(I doubt any fee for full nodes will happen any time soon from a consensus perspective anyway, though.)

 

my words^^

I can even provide IP publicly for everyone to check, that node is up, running and processing requests. point is, that nobody will give you anything, so amount of nodes will continue in decreasing trend + there is still aspect of increasing blockchain size.

I bet, that this will be sometimes subject of change, because serious financial network (tool) simply can't rely on geeks volunteering activities and good hearts of random internet strangers..

Transactions can contain any number of inputs and any numbers of outputs.  If the SPV node could request a bitcoin address from a full node then its transaction could include an output of a small amount to that address.  This would provide an incentive for this node to relay this transaction.  But between nodes transactions have to be relayed for free.  As it already is, nodes and clients must communicate whether they are full-nodes when they connect.  A client could lie though.

One solution is to run an Electrum server along with the bitcoin daemon and then leave a donation address in the server.  Electrum client users will see the address.  I know, you are begging here.  But at least there is an easy mechanism for people to support the node they will be sending their transaction to.

sdp

If a Bitcoin angel will give me 1 BTC now then I will run my full node as long as possible.

I already did it and recently switched to electrum, after running almost 2 years SSD based node with 100Mb/s symetric link. Simply, because there is no motivation to run it except some good geeky feeling, which is after years not enough.

the "Incentive Program" is just joke, which is based on luck and can't cover monthly expenses, even if you will win once every month...btw, I'm not talking about some massive fees, which can fund new ferrari, but just 5-10USD per month to cover basic costs..nothing else.

btw this is nothing new: http://www.coindesk.com/bitcoin-nodes-need/

I don't think the fee you receive will ever cover the full cost of running the full node. Running a node is "volunteer" work, you don't expect to be paid. I think Satoshi visioned most users in the future would use SPV wallets and only businesses or those who have financial interest in bitcoin to run full nodes.

I run a bitcoin node (ranked usually in the top 200 on the leaderboard: https://getaddr.bitnodes.io/nodes/leaderboard/ ).  But it is a bit of a beast and I often seen 400+ connections.  And the last non-stress test from coinwallet.eu when they released the free btc tripled the load on my server.

I also run a masternode on Dash (https://dashninja.pl/masternodes.html ), but that requires a 1K Dash investment for the masternode to have any remuneration.  I make about 2.5 Dash per 5 days.  That covers my cost for the Dash masternode and bitcoin node.  It is nice to get something back.

Read it before; read it again at your prompting.  I want to see Bitcoin succeed; would it kill us to pay full nodes a tiny fee?

I have pretty good internet speed; 90Mb/s down and 11Mb/s up.

I'm in bitnodes at 73.190.2.60 but haven't once received anything.

You are conflating issues. Requiring good bandwidth and large amounts of storage is a technical implementation issue. It is a barrier because it is not optimised rather than set in stone. Again, not financial. The current requirement for a full node is simply that they can verify transactions to provide consensus. They don't need miners' levels of processing power, electrical or hardware. The issue here is that the current implementation is onerous on bandwidth and storage for small devices. They are addressing the storage and one way to address the bandwidth is by spreading the load over more nodes. If every client is a node, each node needs less bandwidth. If you only have two nodes, then just those two are required to have sufficient bandwidth to do all transactions on the network within 10 minutes. The solution, again, is to make every client a node thus forcing all participants to support the network.

I support the idea of small fee for full nodes. After all the full nodes are helping to store all the data which is growing very fast- when I first started the blockchain was only <10GB, now it is approaching 40GB.

Bitcoin is meant to provide currency/exchange freedom.
By requiring everyone to run a full node, you are limiting the amount of potential users, which is harmful for a peer to peer system.
Not everybody has access to good bandwidth or storage.

No you haven't understood correctly. I'm saying that if the model behind the theory of what financial incentives would make a gazillion people start running nodes all of a sudden was true, then people would only buy the above mentioned stuff if it made them money.

What it will actually do is make money for a small number of people who know what a node is and already run them. It is rentier wannabes wanting to get in on the miner monopoly by creating their own little fiefdom and who can blame them? Why should miners get all the cash?

If you really want people to run nodes, then make running a node an inherent part of participating in the network-no full node, no access to bitcoin. That is the technical solution and doesn't rely on assumed greed or allow gaming of the system.

No? If I understand your argument correctly, it cannot be applied to running a node since it can't possibly "make us" money.

I wish people would stop peddling financial solutions to technical problems. We all buy the latest car, iPhone, laptop or TV because of the money it makes for us, right? We all update our OS, firmware and disks because the newer it is, the more money it makes for us, yes?

What would happen if everyone did? Food for thought: https://en.bitcoin.it/wiki/Full_node#Why_should_you_run_a_full_node.3F

It doesn't matter how much money it costs you for others.
Having good hardware doesn't have much of an impact on the quality of your node either.
Good internet speed does though.

You can check how https://getaddr.bitnodes.io/nodes/incentive/ as it might interest you.

It costs me real money to run a full node.  Besides the extra wear and tear on my gear, the full node software consumes memory, disk space, bandwidth, and processing power which generates heat that has to be dealt with driving up my cooling costs.  I would like to be compensated for this expense.

I have a pretty powerful processor with fast and large memory and SSD.  I have a pretty high speed link.  Clients of my full node benefit.

Perhaps there is no hope of getting compensated.  If not then maybe I would be wise to switch to an SVP wallet instead.

Not sure why you bumped a topic from 2 years ago.

As for your service, I don't think it offers more than what is currently available for free.
If you could explain the difference between a "high quality" node and a normal node that could help me understand why you want to charge people.

How would one go about getting the software implemented to charge for the high quality full node services I provide?