It seems that litecoin has had a fix against the attack that consists in creating many tiny outputs out of a given output to spam the network given the attacker is able to pay only a tiny fee for each tiny output, according to this.
Would be the same type of attack against Nu currently feasible? Why so?
1 Like
manysend has the same fee as a single send, it’s true, but that fee scales with the data size of the transaction. Even with spammers pushing it to the limit, that’s still 0.01 nbt/kb no matter what. With btc, those fees get recycled back to the miners. Here, those nbt are going to Nu via burning, which means they are not going back into circulation. With variable tx fees, we can spike the fees during high network activity and lower them again when the spam has subsided and Nu has taken money from the spammer…
As a side note, volume-dependent tx fees would charge for every output just like coblee is saying there about ltc.
Tks for your clarification. So in case of NuBits, if the fee is burnt, it never goes to the spammer so there is no financial incentive to spam…which decreases drastically the incentive to spam at all.
You would need to really want to hinder Nu and have the budget for that.
The point is not that the tx fee gets destroyed when executing NBT tx.
The recent Bitcoin spammers had the fee not destroyed, but it was out of their control as well, because it was paid to miners.
If all NBT are consumed by tx fees, new ones need to be granted, don’t you agree?
So destroying the fee is not what protects the network from spam. The mandatory fee is.
But only from spam that is intended to be at no cost.
The recent Bitcoin attack was done paying fees and not at no cost.
Spamming the Nu network with transactions would be quite cheap.
With 0.01 NBT/kB tx size, you can have 1 MB filled for a little bit more than 10 NBT (because you have to pay 0.01 NBT for each started 1 kB output; you will likely need more than 1024 outputs to fill a block of 1 MB size).
I don’t know what the maximum block size in the Nu network is, but to compare it with Bitcoin, I calculated the price per MB.
Ever since I thought about tx fees I’m a big fan of variable tx fees that adapt to the use level of the blockchain.
With variable mandatory tx fees you have an even bigger defense against abuse (than with fixed fees) at almost no cost (in regular times; fees only go up if the network is highly utilized or being spammed).
As the mandatory is the price that is paid for using the network it doesn’t sound to unfair if the price and the utilization are aligned with each other.
regarding the Nu network: can you pay a fee exceeding the mandatory fee to get your transaction expedited (meaning more likely included in a block than a transaction with only the mandatory fee)?
I don’t know how hard it is to make wallets aware of the use level (to suggest/require a proper fee), but my understanding is that wallets already have all they need as long as they are connected to the network.
Here is my take from more than one and a half years ago:
https://www.peercointalk.org/index.php?topic=1539.msg12777#msg12777
2 Likes
So the solution for bitcoin is to increase the fees when needed (when there is spam) and the same applies to NuBits.
The difference is that (interestingly enough) in Nu 2.0 shareholders will be able to vote for fees in real time directly into the wallet, whereas in bitcoin there is no mechanism to decide unequivocally about the fees, which makes bitcoin very slow to respond to emergency situations and Nu very fast and agile.
What a bright near future for Nu!
2 Likes