Let’s just move ahead with the most pragmatic solution that allows us to start thinking about what to build to create actual demand.
You dismissed my game theory points entirely.
Let’s move on from this deadlock and discuss the other criteria for dynamic plasma.
Until virtual voting, this is not guaranteed.
My point was: consensus nodes will be incentivized to prioritize and only accept fee’d transactions.
No, it’s because nobody gains anything from such an attack. The only threat is irrational spam.
Why not? It’s called the feeless paradigm for a reason.
This is one of the only attributes that sets our network apart.
Nano, IOTA, Shimmer, Decred, Banano, etc…
You’re free to discuss tokenomics in a separate thread. This is outside the scope of dynamic plasma.
2 Likes
“The ETH chart looks great. Hope I can say the same for ZNN.”
You’re cherry picking the one that fits out of 99978876 cryptos. This says a lot about your knowledge when it comes to economy in this space. I don’t doubt your coding skill, but here you’re just straight wrong and dishonest. ETH just got an ETF, probably not priced in its chart right?
The entire economic point you make in order to support your views on the fee / no fee debate is moot and nonsensical. No, fee won’t benefit holders in a meaningful way.
2 Likes
Let’s not undermine what makes ZNN different.
4 Likes
This continues to sound like a reasonable and effective way to deal with congestion whether its from spam or regular use.
1 Like
I don’t care about the price. I care about getting this right from the first shot, because if we have limited resources.
Dynamic Plasma is the anti-spam strategy of the network. With a hybrid design, we can kill two birds with one stone.
Because if we want to solve this issue, we need to get to the root of it: an attacker can spam easier without fees.
The byproduct of generating PoW, fusing QSR or burning ZNN (fee) will be Plasma.
They will render the network unusable. RandomX will alleviate this issue, but not entirely. A hybrid design is the definitive answer.
A feeless paradigm can coexist with a fees to create a hybrid paradigm.
This network has so much more potential.
IOTA is a failed project (Shimmer included). Nano (Banano included) is suffering from spam from time to time.
Designing components for a decentralized network may overlap with tokenomics and other “non-technical” areas quite often. Bitcoin is more than the tech stack.
I’m pointing out ETH because they have a working system that burns fees and we can all see it in action.
Burning fees will benefit everyone in the long term. More activity => more fees burned => less ZNN circulating supply.
Let’s strive for a more resilient system that takes into account all the variables.
Fees + Feeless > Feeless (it currently makes more sense than ZNN + BTC > BTC)
This hybrid design sets us apart and offers a competitive advantage over every other network out there.
Let’s suppose I’m an attacker and I want to spam the network:
- I can acquire beforehand QSR in large quantities at a reasonable price
- I can rent cheap PoW (even with RandomX I can overpower ordinary users)
- I can create a random ZTS token by burning 1 ZNN
- I can create unlimited accounts without any costs by sending my token (I don’t need ZNN or QSR)
Is there really any proof of this?
There are already many important ways for ZNN to be removed from circulating supply like locked up infra which is already a huge chunk. If this isnt enough for any rational investor then burning a few znn in fees surely wont change anything.
Can you explain how these points are an attack? The dynamic aspect of the algo would adjust the cost and more plasma/pow will be required. The incentive is to work within the capacity of the network.
I do not see how adding a lane for fees is a competitive advantage in any way, cant that be for the extension chain?
1 Like
No, there are none. Quite the opposite actually. Most cryptos burn fees and the impact on the circulating supply is hardly impacting anything else. The data are out there and can be accessed easily, @aliencoder cherry picked ETH because it fits his views and willingly ignored everything else. This is complacency and would result in poor decision; cherry picking data never worked well in research. Beside, here we have a dataset of 1 data point (wow), which is extremely questionable as ETH PA is not the result of the fee actor alone so arguing that it benefits holder is a bit rushed, at best.
I suggest we stop wording it as introducing transaction fees.
Aliencoder is suggesting we add an additional method to generate Plasma: Burning ZNN. Doesn’t sound too unreasonable, when stated this way.
I’m way more interested in actual numbers and simulations:
-
How much Plasma would burning 1 ZNN Generate? Moonbaze suggested a flat number of transactions per ZNN burned, instead of a Plasma amount.
-
After the tx is confirmed, plasma levels go back up. How would this work with burning ZNN? When QSR gets unfused, account plasma is lost. When burning ZNN, do we only get temporary Plasma for a single transaction? If it’s not temporary, it will allow an user to basically acquire an “infinite” number of transactions by burning ZNN (provided network plasma threshold to be included in a momentum remains the same).
1 Like
Ethereum is the most obvious proof, but there are others nevertheless.
Locking != removing
With 1000 participants in the network you won’t notice the difference, but at >100k the numbers will add up.
Point 1 is as easy to accomplish if a dynamic algo is in place.
Again, I’m not arguing about price increase. Most cryptos use fees because it is a standard way to deter spam and that’s it.
Exactly:
- Burn ZNN to generate plasma.
- Fuse (lock) QSR to generate plasma.
- Generate PoW to generate Plasma.
Whatever works for you.
We can borrow his base fee level: 1 ZNN burned = 200 txs
Of course temporary. You burn 1 ZNN, you get 200 txs if the network activity is at a baseline. If the network gets congested, you’ll need to burn more (subject to the dynamic algo).
2 Likes
But that doesnt solve Dynamic Plasma, its just another way to transact.
It is an empirical problem maybe thats the approach we need to take.
Thank you for the replies but this is not proof, just an example of it.
I’m still struggling to understand why we cannot prevent spam with PoW and QSR fusing alone. Maybe we stop calling it spam and just call it valid TXs. They are the same.
2 lanes | PoW and QSR
PoW lane get saturated. PoW adjustment goes up and take longer to generate PoW. Setup RandonX to work on one address at a time. 1 CPU = one PoW generator per address. If you spam the network you need lots of CPUs. It’s costly.
QSR lane gets saturated. Amount of QSR needed to process a TX goes us. As usage goes way up (in years) it’s possible users will need to rent QSR to process a TX because the threshold is too high.
Lane width - Dynamically adjust lane width to ensure both lanes are saturated. When both are saturated, we can adjust them to make sure the width is “fair” (TBD).
Issues
- what is someone fuses 1m QSR on one address and hogs the network while the PoW lane is full?
– Response: Limit QSR per address via governance. - what if someone fuses 1000 QSR (the threshold established) across 1000 addresses.
– Response: That becomes the new limit and everyone needs to fuse 1000qsr and it’s first in first out. What if we also require a very small PoW from a CPU when QSR is saturated? That would mean an attacker would need 1000 CPUs or would need to process the spam sequentially from fewer CPUs slowing them down.
3 Likes
Wouldn’t it be possible to implement a PoW + QSR lanes solution and stresstest it?
2 Likes
yes, I think it can be simulated in python. Would not be hard to simulate.
We know the max throughput from feedback from Moonbaze. 3800 send / receive account blocks per momentum.
1 Like
assume 3800 account blocks every 10 seconds. We can model this in excel and python. Or we can use @risk.
That would be interesting so we would have empirical data to make decisions.
2 Likes
Gr8 we should try this and see how it goes 1st.
We’d need to see how the chain react with unpredictable factors aka computational power from Pillars, bandwidth, memory usage, etc no?
1 Like
Because spamming the network must incur economic punishment for the attacker, while offering a net-positive benefit for the whole network. Burning ZNN achieves this and can be complemented by PoW generation and QSR fusion in a tri-lane system.
Paper: https://arxiv.org/pdf/2304.06014.pdf
- Significant later results include [Chung and Shi(2023)] which show that in general, no transaction fee mechanism can satisfy user incentive compatibility, miner incentive compatibility and be off-chain agreement proof all at once.
- A conjecture can be made that in a perfect, rational system, the variance between transaction fees will be uniform; that is, users whose transaction are included into the block will pay the exact same fee. This has no correlation to the urgency of a transaction, since transactions are still restricted by the block interval - simply paying more than others will not make their transaction be confirmed any faster. Also, if the block size is limited, assuming there are enough transactions submitted, the blocks will always be full and identical for each (rational) miner.
1 Like