Don’t try to induce in error the readers with this statement: I was proposing a hybrid system with a feeless lane.
Dynamic Plasma implementation
- Feeless (no change)
- Hybrid (2 lanes: one feeless, one with fees)
- Fees
0
voters
1 Like
What’s the ultimate consequence of option 1) feeless (no change) in terms of scalability limitations?
This is a good point.
1 Like
Is there a counterpoint to the suggestion that the introduction of fees creates an incentive to spamming the feeless lane?
Or another perspective on the same concept: Won’t entities prefer to use the feeless lane rather than use the fee lane? And therefore wont the feeless lane conceivably be constantly maxed out and dominated as the result of not solving feeless dynamic plasma? It seems like this results in a defacto fee based L1 for all intensive purposes.
Does creating two lanes result in a hydra? (Two headed problem.)
If I read correctly, the tokenomic inflation/deflation theme is very prevalent - which makes me ask “Are we intending to solve for dynamic plasma, or are we hoping to redesign the token economy?”
I don’t induce anyone in error - I think everyone can read - and no, an article about how ETH became sound doesn’t counter the fact that is didn’t work on ETH and other cryptos for the benefit of holders. Check the charts.
1 Like
I think the issue is we don’t want fees, but no one has presented a solution to dynamic plasma that “works” without fees. Right now I’m keeping an open mind to the possibility a solution exists without fees.
1 Like
Thinking about this a little more. If we don’t have a fee lane, Trustwallet, for example, would need to incorporate PoW and/or QSR fusing into the wallet in order to support NoM. That is unlikely for a long time. Therefore, we would need to rely on a Metamask Snap if @vilkris can overcome the technical issues.
If we don’t have a fee BUT we solve the Metamask technical issues, won’t NoM work with CEX so long as they work with Metamask?
1 Like
The problem is that ordinary users will be pushed out of the system if the PoW is too difficult. By introducing another lane, the dynamic algo and market forces will balance the usage in a more uniform fashion.
By having a fee-based lane, you can spam the feeless lane as long as you want, but you won’t congest the network. That’s exactly the reason you’ll find a dynamic algo in my proposal that balances the two lanes by taking into consideration network usage and other parameters.
We can kill 2 birds with one stone: solving dynamic plasma and creating deflation for ZNN.
The ETH chart looks great. Hope I can say the same for ZNN.
Because dynamic plasma is about solving network spam. And to solve this issue, a holistic approach is needed where we need to think outside the box.
Let’s be honest, without aligning with the rest of the market won’t bring us closer to “mass adoption”. Imo we have another 6-8 months left to build before we enter the bull market. With a “standard” fee system we can start integrating with popular multi-chain wallets that will open up millions of potential buyers.
No, because CEX use custom backends, not Metamask to process user funds. Similar to multi-chain wallets, CEX prefer standard implementations. And we need CEX to bring even more users than multi-chain wallets.
You can see that I’m trying my best to answer everyone’s questions. @Shazz @Stark @Chadass @0x3639
Another positive outcome for the hybrid system:
An attacker will need both PoW and ZNN to be able to successfully overwhelm the network. This fact is greatly reducing the attack surface for a sophisticated attacker, as he needs to open 2 fronts in order to spam the network.
1 Like
I’d like to bring the focus back to an important topic: dynamic plasma requirements as a function of network utilization.
Despite 170 posts in this thread, there aren’t many about the problem we are trying to solve.
Instead, we’re debating the entirely separate topics of tokenomics, emissions, and price action. Anyone that’s passionate about these, please feel free to start another thread. Those changes belong in a separate ZIP.
After almost a year of deliberation, this is the only thread that proposes solutions to balance PoW and fused QSR plasma. I believe a few devs are still drafting their proposals, myself included.
Here’s an excellent summary of relevant data: KB: PoW and Plasma - ╰ Network - Forum
These variables/functions should be the primary focus of our discussion.
Considerations:
- bandwidth constraints / momentum capacity
- difficulty adjustment periods
- fluctuating transaction / data costs based on recent network conditions
- fused QSR limitations
- the ratio of PoW plasma to fused QSR plasma
- exceptions, such as guaranteed tx priority for certain tx types
- RBF, but with more plasma
A few points about the work we need to accomplish:
- We need both node and client PoCs for RandomX integration. Vilkris has started that work here.
- We need a node PoC for transaction ordering based on transaction plasma. Even if we can’t enforce it, I’d expect most pillars will use this code.
- We need to define an algorithm that balances plasma requirements based on network conditions. Afterwards, we’ll need a node PoC.
In terms of feeless spam control:
- Prioritize fused QSR over PoW bandwidth
- In the absence of fused QSR tx, PoW tx should be able to use all available bandwidth
- In the presence of an abundance of fused QSR tx, PoW tx bandwidth should be limited.
- Allow fused QSR to be frozen and seized via governance
- This threat alone should deter actors from spamming senselessly
- I can elaborate in a future post if anyone wants to know more
- A penalty is not censorship
- Account-based plasma requirements
- If an account submits many tx in a short period of time, their subsequent txs cost more plasma during a cooldown period.
- I’m not sure if this is feasible
Are there other feeless ways to limit spam?
4 Likes
The problem we are trying to solve is deeper than you think. I’ve brainstormed many scenarios that include only the feeless design and a hybrid system is the only way to really get the best of both worlds. Otherwise we’re running in circles.
We are debating only tokenomics related to anti-spam mechanisms. Burning ZNN indirectly benefits everyone from this ecosystem.
Bitcoin is successful not only because it has great tech, but it also has powerful incentive structures.
I agree: burn ZNN fee > Fuse QSR fee > PoW fee
I won’t support anything like this and it’s dangerous to even propose it (I’m against this proposal, not your opinion).
I agree and I’ve proposed this earlier.
Without a fee based mechanism, one cannot limit a sophisticated attacker to spam the network. By combining all of those mechanisms in a hybrid design, we can achieve a system that will deter even a determined attacker.
We’re already doing this because three people decided fees are the only way to counter spam.
Let’s talk about the drawbacks that will result from the introduction of fees.
- Increased risk that consensus nodes accept/refuse certain transaction types
- people are worried that my proposal introduces censorship risk, when this is arguably worse
- Increased risk of PoW and fused QSR spam in order to force the average user to buy and burn ZNN
- we should probably brainstorm the game theory for the incentive changes
- George warned that fees could introduce perverse incentives
- Reputational damage from years of marketing a feeless paradigm, only to concede to the status quo and force users to pay transaction fees
- inb4 “fees are optional” tg sticker packs
- Weakened QSR utility will likely have a negative effect on its future price action
- Risk that feeless zapps won’t be feasible under certain conditions
Your reasoning for fees is:
- PoW and QSR are relatively inexpensive and permissionless, therefore it will be trivial to spam that bandwidth
- Fees are progressively expensive; therefore, as a measure of last resort, malicious actors won’t be able to spam the network infinitely
My reasoning for feeless is:
- PoW is subject to hardware, energy, and difficulty constraints. Difficulty adjustments will already limit the threat of PoW-related congestion.
- Transactions will be subject to dynamic (fused QSR) plasma requirements, potentially even at the account-level.
- If we can algorithmically limit individual addresses from spamming, then we’re probably in good shape to scale.
- If not, then we can rely on governance to intervene and forcibly limit their throughput.
- inb4 “slippery slope”. Sure, but we’re already trusting pillars to act in the network’s best interest.
- Won’t Sentinels and PoW links already limit every address’ throughput anyways?
You also haven’t addressed most of the concerns that actually pertain to dynamic plasma, the ones I listed. The introduction of fees as a congestion control is just one part of the solution we need.
Let’s set aside this debate of fees vs no fees so we can move on from this deadlock and focus on the rest of the items.
Fees are part of the solution to counter spam. It’s a hybrid design.
We have a solution for this: threshold encryption to obfuscate transactions while they are in the mempool.
This risk already exists within the current system.
That is exactly what I’ve been doing for the last 2 months.
Fees alone are introducing “perverse incentives”, but Bitcoin is running fine with high fees nonetheless. But we can do even better.
Again, my proposal is a hybrid design.
I think we improve our reputation and adopt a cerebral solution: to create a hybrid design where we take the best of both worlds.
I think we all care about the future of this network, not random fudders.
QSR already has utility: fusing and spawning Pillar slots.
QSR is burned to create Pillar slots. It is also locked for fusing.
In a hybrid system, you can opt out of fees and only use PoW if you want.
My reasoning about introducing fees is that:
- Fees are a good deterrent when the network is spammed.
- A hybrid design is better than a feeless design or a design based only on fees.
- Burning ZNN represents the missing link in this ecosystem.
- If the network is spammed, one cannot fuse QSR.
- It is easier to attack a feeless design than a hybrid one.
If we start changing the code, we must be on the same page because we’re in the same team. I just want to be sure we have all the answers we need to proceed with the implementation.
Keep in mind though that znn burn from tx fees will unlikely have more than a marginal impact on the reduction of inflation. As Chadass said, we need usage, hype, adoption. Znn burn per txs is a bandaid when you have zero usage and demand.
Let’s just agree that the initial dynamic plasma solution won’t include the following:
- obfuscated transactions
- virtual voting
- sentinel support
No, it isn’t. Transaction spam will not result in any sort of fees for the average user.
They will only have to wait for their transactions to be confirmed.
I disagree; rational actors in fee-only systems will act in ways that benefit them.
In the hybrid system, rational and irrational adversaries can spam the network to force everyone to pay fees. They only need to reach a threshold where the average user is sufficiently deterred from waiting.
Additionally, consensus nodes can reinforce this initiative by only accepting fee’d transactions, because it’s in their financial interest to have people buy/burn ZNN.
In a feeless system, the risk of a rational actor behaving maliciously is slim, and it becomes nonexistent with governance intervention. Irrational QSR whales won’t stand to gain financially from spam. They risk losing money in that circumstance, as network participants leave the system due to frustration.
This is likely why no one has spammed NoM to date.
And I’m convinced a balanced PoW/QSR system with difficulty adjustments will sufficiently deter senseless PoW spam given enough time.
Keyword is weakened. I think its use case will be diminished if the proposed feeless lanes are constantly maxed out. People will resort to paying the ZNN fee, and even then, they may be forced to wait or pay a higher fee.
The average user won’t care to run infra.
So are:
- QSR penalties via governance
- PoW difficulty adjustments
- Account-based plasma adjustments
I know I’m repeating myself. That’s why I asked to move on to other discussion topics.
Better in which way? Flexibility? UX? Scalability?
I believe the product with the lowest cost to users and best UX will outperform all solutions.
Our chain was specifically designed to be feeless in order to compete in a landscape plagued with high fees.
This narrative is derived from an argument about tokenomics and emissions.
Dynamic plasma doesn’t need deflationary forces.
Not necessarily; I think we can design a creative solution to this problem.
Vilkris has been working on a provider/onboarding system, and expressed other ideas to support new users.
If we want a quick solution, of course, fees are the easiest way forward.
If it’s designed well, it should cost considerable resources to temporarily inconvenience other users.
Why don’t we design a system that models after Bitcoin’s sats/vbyte instead of lanes with dedicated/variable bandwidth?
I understand their nodes are incentivized to accept higher tx fees, but we will probably address this with virtual voting in the future.
1 Like
How about setting a very steep curve for QSR requirements for plasma? Maybe 10,000 QSR for high plasma. Plebs can use plasma faucets. QSR maxis can start a PaaS economy. Keep the base layer streamlined.
At the moment it’s just a way to attract rational investors that will see real demand for ZNN.
We’re thinking long term (like an investor would), aren’t we?
I agree.
They also wait for their 1 sat/vbyte transactions.
We’re talking about a hybrid system here.
Pillars will see that a transaction has x amount of Plasma and decide to include it into a momentum.
No one spammed NoM because no one knows NoM exists in the first place.
Let the market forces and the dynamic algorithm decide which lane is more used.
If they want to pay fees, let them pay fees. Don’t try to impose something they are not familiar with just for the sake of being different. This is the model they already know. And this is how 99% of crypto currently works. We won’t be able to list anywhere if we only pursue the feeless narrative. Again, I’m not trying to replace the feeless design, but complement it.
Flexibility and UX:
- Users outside our bubble are already familiar with fees
- CEX listings will be easier in the future
- Integration with multi-chain wallets such as TrustWallet, Exodus, Coinomi will be possible
- Market forces will decide which lane is more used
- ZNN will become deflationary and this is a big selling point (especially in a bullmarket)
Remember that we’re in Phase 0. For Phase 1 NoM needs to evolve. Isn’t it a self-evolving tech?
The ecosystem desperately needs deflationary forces.
You always forget we’re talking about a hybrid system, where fees are only part of the solution.
I need to repeat myself too: to spam a hybrid design an attacker needs to open 2 fronts.
Because I think we can do better.
Virtual voting is consensus related. When we’ll have threshold encryption for the mempool, we won’t need to worry about which transactions get in the momentum. The pre-processing will be performed by Sentinels before Pillars will include the transactions.
@aliencoder I think you make some good points. Setting aside the discussion of Dynamic Plasma, I hope we can introduce deflationary forces in the network. Maybe the Sidechain and other scaling solutions will help.
If you were to take the other side of the debate, and support feeless TXs on the base layer, how would you try to design the system to accomplish that and what are the issues with that system?
My thoughts:
We can establish two lanes (PoW and Fused QSR). The width of the lanes can be fixed or adjusted programmatically based on saturation. If someone is attacking PoW (which could be easy) a user could circumvent that by fusing QSR. If someone is attacking PoW and QSR, they need to find a lot of QSR. It’s cheap, but not easy to buy. And there are “honest” QSR whales that could help overcome any QSR spammer.
If someone has 1M QSR to spam the network they can fuse 100qsr to 10,000 addresses. Users can overcome that by fusing 101QSR. The system will eventually come into balance. And, if we eventually create a QSR marketplace to “rent” QSR, attacking the network with 1m QSR will have a cost.
Back to PoW. Can we limit the amount of PoW that can be generated by one address? So one address cannot overcome (block) the network? If we can do that, the spammer would need to spread out their PoW across several addresses. They will just be another user processing valid TXs and other valid transactions would just take longer to process. If the PoW lane is saturated and Fused is not, we can adjust the lane width. Over time it could be “costly” (time consuming or lots of QSR) to process TXs on the L1. This will be for moving large amounts of digital gold (over time). This pushes TXs to the deflationary layers - L2, L3 and Sidechains.
1 Like
We already have them for QSR. We need them for ZNN too.
But without fees we have an incomplete solution. Let the market forces decide if fees or feeless is the way.
Programatically with a fixed minimum size.
You don’t want to rely on network participants for designing an architecture. Only on game theory and programmable approaches.
I see that everyone think “fees” are the bogeyman. The hybrid solution is our best shot to 10x this ecosystem in the next 6 to 8 months.
Feels like we should solve for Dynamic Plasma without regard to our collective desire to increase price. We all want to increase demand, price, and dev interest. But I think (hope) solving Dynamic Plasma is a different issue than solving economics.
As a thought experiment, if we separate those two desires and focus exclusively on solving the problem of Dynamic Plasma, why can’t my big picture thoughts solve the problem without fees?
2 Likes