//thread
The slow death of on-chain defi.
Intents, solvers and more problems.
1/15
In my opinion, the major reason behind the rise of intents, solvers and CeDeFi protocols is a failure of most defi primitives - pooling funds creates a high value target for hackers.
And the wide spread of upgradeable contracts introduced more attack vectors than eliminated
2/15
Once exciting "defi legos" narrative falls apart, facing the pressure of supply chain attacks - just one minor issue in one protocol could lead to huge losses spread over the ecosystem.
The leverage of network effects now do the bad, committing to the system's fragility.
3/15
So the solution in the middle-ground has been found: users hold assets on-chain but "solvers" or other 3rd parties do stuff with these assets on request, trading off their security for profits.
4/15
While being a very reasonable response to the original issue and having many other advantages like gasless settlement, UX & chain abstraction due to separation of the intent and the action, there are multiple issues people rarely talk about.
5/15
Issue #1. High dependence on centralized proprietary off-chain infrastructure.
One of the selling points of defi was that frontend or backend actually doesn't matter and the user could just interact with the contract in case when something goes down.
6/15 1.1 We used to believe that soon, all protocols would be powered by unstoppable world computers.
But now protocols are rapidly increasing their dependence on closed source off-chain backends and frontends, which not only affects app's liveness but leads to other problems.
7/15
Issue #2: Verticalization of the stack and trend to further centralization.
By delegating operations to off-chain components, app devs achieve full control over their user commitments, information flow and other parts of the supply chain.
8/15 2.1 By having that control, they secure their position and users, helping to extract as much value as they want, being only constrained by the competition and the market.
9/15 2.2 It also gives them long-term advantages over on-chain protocols that could be easily replicated and get a competitor with a new shiny incentive program, for example.
Which leaves me with a pessimistic view on the future market share of on-chain defi.
10/15
Issue #3: Composability through more centralization.
Constraints imposed by these protocols leave the only way to interoperate: to have what I call "omniscient solvers" - entities powerful enough to operate on each of them, increasing their margins, gaining even more power.
11/15 3.1 In the world of L2-centric roadmap, which not only splits users and liquidity across protocols but across L2s as well, I see this as one of the greatest currently existing centralized forces.
12/15
Issue #4: "slap an auction on the frontend and solvers will figure out the rest"
Such design undermines the role of the protocol compared to the "old" defi making it a web2 platform for traffic arbitrage.
Intermediaries that web3 should've been fighting are now powering it
13/15
We've come from fully on-chain vision of defi to centralized off-chain storefronts selling user orderflow to proprietary firms.
And I've seen almost no reflexive thoughts on this topic, most people just took it as given, not questioning the path our industry has taken.
14/15
Given the wide spread and acceptance by regular users, I'm afraid there might be no way back as this model provides better UX, stable revenue model, long-term "unfair advantage" and low maintenance costs for protocol using it.
15/15
I know there are other protocols addressing these issues but observing user behavior patterns and current political & development trends I could only hope they could make a noticeable impact on the industry.
P.S. Thanks for reading up to the end, hope you liked it!
As usual, here's a disclaimer: X isn't the best place for long articles and discussions so I've cut many nuances and the thread is super opinionated.
Would love to hear any feedback, especially if you disagree :)
And huge thanks to @DeanEigenmann for writing this post, which inspired me to write my thoughts on the topic.
First thing, that afaik, flashbots doesn't solve this. Only big miners have access to this market, flashbots doesn't give access for small independent miners.
And here's one of the reasons why they don't do it:
Flashbots meva is considered as blind auction, where all searchers have the same information. But here's a problem, there's a possibility of leakage of bundles from miners to friendly searchers.
Why some searchers (me included) don't like flashbots auctions.
A thread.
(1/11) Quick recap first:
Flashbots auctions are intended to move bot wars from mempool to some offchain environment, so they doesn't spam txpool and increasing gas prices while playing the game.
(2/11) The other objective was to "democratisize" mev, by making it lossless by allowing to not pay gas fees for reverts and faulty solutions.
I see that people like common truth, so here's another one: you can make some money on mev using flashbots pretty easily, but if you want to make crazy amounts of money, you're already late to the party and it's almost impossible to catch up imo.
A thread.
If you want to make sandwich/backrunning bots on eth using fb, you'll face an enormous competition, and as a result, high bribes to miners and low profit to yourself.
That's because fb will accept the highest-paying solution, and searchers will have a lot of time to compute their solutions using suboptimal methods.