Updating Filecoin’s value accrual mechanisms #1010
Replies: 2 comments 15 replies
-
|
Thanks @AxCortes for the write up. Quick question I had on initial pass:
But regardless of what the SP wants I think this just isn't what clients want. Have we had clients request this at all? If that's not the case why would we build something to incentivise this. I think better time could be spent incentivising other activity on the network that spurs adoption opposed to incentivising a payment mechanism that's going to cause clients/SP's problems. Not saying we can't accept FIl but do we need the incentive for the clients. Defi could surely help in enabling FIL/USD hedging or payment deals. Am I missing something there? |
Beta Was this translation helpful? Give feedback.
-
|
I actually totally agree with your question - I think the ideal scenario is when deals are paid in any currency (bigger markets) and the Filecoin network takes a fee (converts say USDC into FIL and burns it) |
Beta Was this translation helpful? Give feedback.
-
I think you are just describing Explicit proof fees, which I agree with too, and are the first point of the proposal, so agree we should definitely do that. It is also not mutually exclusive with the idea of also locking FIL for deals, it's a separate approach we can explore. I agree burning does more than locking for value accrual, but the effect of locking should not be dismissed either. |
Beta Was this translation helpful? Give feedback.
-
|
@AxCortes I think the resistance to part 2 of incentivising paid deals to use FIL is that it would seem to degrade the product offering and also probably not work. SPs can offer a better service to clients by trading in the tokens that are convenient for the client. If SPs are paid through the back door to offer a less convenient service, they'll get less business and it would seem that the network as a whole is providing less value (leaving less for FIL to capture too). As @mjroddy points out, restrictions or incentives would likely be worked around, hedged, intermediated etc to provide the service that the parties want anyway. Restrictions to thwart such arrangements (like requiring more locking) just further degrade the service offering.
The problem is well stated, but I take the opposite strategic direction from it. The best service with many users doesn't necessitate any use of FIL for medium-of-exchange, so how can we design Filecoin to thrive and capture a fair share of value despite this. Take it as a design axiom that we can't force use of FIL for deals, and design an economically sustainable system from there. I think part 1 of onboarding/service fees is a good, probably sufficient solution, so long as the network can effectively charge a fee associated with the service value (e.g. fee market). |
Beta Was this translation helpful? Give feedback.
-
@mjroddy Here's how I think about these things. What is the service that the Filecoin network provides?
What the Filecoin network provides is a coordination mechanism that helps clients and providers make a transaction. It establishes a standard protocol for creating and checking proofs, and pays thousands of computers around the world to check those proofs. It provides a tightly integrated programmable system for coding rules about service faults, payments etc. (It also provides some basic economic blocks like pledge and penalties, though arguably some of these built-in ones are not of value to the participants). This coordination system provides a clear baseline set of rules and a basis on which clients can trust that their data is replicated. Filecoin provides this coordination and trustworthy system to SPs, who can then use it to sell verifiable data storage services to clients. A client and SP could do all this themselves, without a blockchain. An SP could seal sectors and send proofs, including daily Window PoST, directly to the client, which could check them. This would be a bit of a nightmare to arrange, multiplied by many clients-provider pairs, and a huge barrier to use. The key value that Filecoin provides is one network that does this for everyone. And the more clients and SPs are using the same network, the better it gets for everyone: more secure, more technology and product development, etc etc. Imagine for a moment that blockchain capacity was unbounded and so gas was free (but many validators still secured the network for some reason). In this case we might say that no fee is technically required. But Filecoin should still charge a fee for the service that it provides to the SPs by practically enabling this verifiable storage. The fee can't be so high that it raises the cost to clients too much, but it can certainly be above the pure technology costs. |
Beta Was this translation helpful? Give feedback.
-
|
@anorth taking this a step further, should (is it possible) the filecoin network charge for these services in a flat USD figure, payable in FIL. It could burn the entire fee or cycle some back to a development fund for grants and funding etc. That means that even if gas was free SP's would still pay for the proof check to provide assurance about their own services to clients Zooming out the filecoin network becomes deflationary when the value of services to clients (measured through fees) is greater than the incentive for them to join.... |
Beta Was this translation helpful? Give feedback.
-
How can protocol tell if a deal payment is genuine as opposed to miners spamming to get quality multipliers? |
Beta Was this translation helpful? Give feedback.
-
|
It doesn't, but the point is a non-genuine payment has the same cost, the SP has to lock up the amount of FIL in the deal that otherwise a real client would have. So there are not really "fake deals", it is a real deal, where the SP is the paying client, which they are free to do. |
Beta Was this translation helpful? Give feedback.
-
|
FIL+ |
Beta Was this translation helpful? Give feedback.
-
|
we could just do PoS if locking up FIL is the value we are out for |
Beta Was this translation helpful? Give feedback.
-
|
I agree the idea is similar to just including an additional staking element to get more rewards. One point of the design should be that the additional rewards wouldn't be enough for the SP to just go and lock their own FIL. The finances would make sense if it is the client's payment. Suppose a client wanted to pay an SP and spread the payment equally throughout the Sector lifetime, They are already locking some amount of funds with a vesting schedule. We'd just want to make it favorable for those funds to be FIL, but not necessarily incentivice locking funds that were not going to be locked anyway. |
Beta Was this translation helpful? Give feedback.
-
i might be missing something and don't quite get it. It says...
can't one just make fake deal payment to get multipliers? like paying back client the deal payment amount? |
Beta Was this translation helpful? Give feedback.
-
Background
One important goal in Filecoin is to drive real user adoption. Specifically we want to onboard more real user paid data deals. There are, however, no direct mechanisms in place through which real user adoption would lead to FIL value accrual.
The only way value accrues onto Filecoin as user demand rises is through base fee burn. This is appropriate for other blockchains whose main product users consume is their block space, but it is not appropriate for Filecoin where the main product offered is provable data storage. This issue has been previously presented at length #557, where a suggested solution is to introduce explicit proof fees, where value would accrue to Filecoin every time someone uses a storage proof, regardless of the amount of gas the proof consumes.
Recently a lot of discussion has begun on the possibility of introducing new types of proofs, enabling different types of storage, such as Proof of Data Possession (PDP) enabling Filecoin hot storage. The value of these proofs to the users would be even more disconnected from their respective gas usage, since these proofs by necessity must be less gas-expensive, while providing more value to the users, than cold storage proofs.
A separate issue is that, while we expect a rise in paid deals would lead to a rise in demand for FIL (to be used as payment for such deals), there are no actual incentives that make users prefer to pay for their deals in FIL. In fact large part of deal payments may happen completely off chain.
Proposed solutions
While initial discussion on explicit proof fees lost some momentum following their introduction in #557, we propose actively restarting this line, specially in light of the hot storage and PDP/ new proofs discussion. As we diversify the products and services Filecoin offers, the disconnect between the value they provides to users, and the base fee burn will grow.
In particular, if new hot storage capabilities as are being discussed in #988 are introduced, these should be accompanied with appropriate proof fees, such that wide adoption of Filecoin hot storage actually leads to value accrual to Filecoin.
As was noted in the original proof fees proposal, having appropriate proof fees, would also enable us to remove the controversial Batch Balancer mechanism, which aims to address similar issues, but has significant flaws.
We can explore mechanisms to incentivize deal payments to actually happen in FIL and on chain, such that increased demand for storage translates to increased demand for FIL. One direct mechanism is to award sector quality multipliers based on amount of FIL deal payment shown in the sector. This could be organized such that it increases competition amongst sectors to include larger FIL payments on chain.
Beta Was this translation helpful? Give feedback.
All reactions