Over 5 years after the UNI drop (and thank you, btw), the biggest legacy of Uniswap in DeFi, I feel, is the Uniswap v2 contract. For one simple reason. If you see just about any other smart contract on any EVM compatible chain, you have to think: “Well, where’s the gotcha.” Now, that chain might have gotchas! But that contract is going to do what it says on the tin. That’s beautiful, and what we, five years later, are missing now.
-
Governance is suggestions: 5 years of DeFi and I am convinced more than ever that decentralization is an ideal that is strived for at best, and given lip service at worst. Governance votes are never guarantees, always just suggestions. The examples of DeFi protocol insiders, the five or less actual “devs” who actually “do the thing,” making “emergency measures” despite token votes are numerous. At the end of the day, there’s always some human typing “git commit” and they may or may not do it based on the votes.
-
Back ends aren’t transparent and fixed: Then there are the “immutable” contracts themselves. Either they are completely obscured as to their source code, and full of admin keys and “upgradability,” or they have it simply a few layers deep, in another contract that is referenced or in some off-chain computational resource. Only a Uniswap v2 contract will do what it says on the tin. Every time. I mean, there may be more examples, but none off the top of my head, certainly not with the historical impact. And still current relevance! Someone spins up a new EVM compatible chain and someone two seconds later is going to come up with a Uniswap v2 clone with Uniswap v2 contracts that work exactly like they did, years ago. (But again, the chain itself could have admin keys, upgradability, 20 validators that are really just one dude, etc., etc.)
-
Front ends aren’t transparent and fixed: Of course, there are also frontends. Uniswap was also fairly historic (in a less positive way) for being the first high profile case of a “DeFi” protocol censoring their frontend for fear of regulatory attack. Those who knew, of course, could find a way to the Uniswap v2 contracts without using the Uniswap front end. And that’s beautiful. Because, at the time, there was basically no regulation—good or bad—there was only fear of de facto registration through legislation (hence the front-end censorship), but nothing actually written into law as far as I can recall. Which is okay, I think, if—if I can use the old maxim—Code is law. In /that world/ with no regulation, Code is Law works for me. I don’t know about this world. Personally, I was much less scared of smart contracts five years ago than just about everything that is going on now.
(My personal view, if I may slightly digress, just so there’s no misunderstanding as to motives: Crypto should not be connected in any way to banks in any way where ultimately the general public should have to bail out the degens. FDIC-insured banks with “fractional reserves” should not be impacted if some crypto 1000x leveraged prediction market rug pulls or death spirals. That means, in my ideal world, yes, code is law, the MEV extractor or the MEV extractor’s extractor (per a recent case) is on equal terms—all the way up until the proverbial fist hits the proverbial nose, which to me is the onramps and offramps. Do what you want with your bits of 0s and 1s but when you try to convert them to dollars and cents… Well, where the buck starts, is where the buck stops. Just a bit of disclosure on my personal views. I also don’t like one-dollar-equals-one-vote and the outsized power of A16Z, etc., but that’s another matter for another day.)
So, I don’t think we’ll ever get away from a human has to type in “git commit” sometimes. Robots aren’t running the cooling towers in data centers. Humans will be a part. But we /can/ make it about as “immutable” and decentralized as possible when proposals are enacted, and I think this suggestion needs to incorporate this if we want to believe Uniswap is striving not lip-servicing.
-
Governance implementation is as human-less as possible: Basically, you have the high-level proposal, but you also have a very, very detail implementation plan connected to it. One that is clearly audited HEAVILY. But basically the plan is a really detailed computer program that is waiting to read the Ethereum blockchain (presumably) for the result of this vote as basically a parameter that will run the entire program. It details exactly what will happen to the contracts, the infrastructure, the websites, the front, the back, you name it and it happens all programmatically (ideally). If there’s ANY part with a human in it, make them have skin in the game, like Hayden Adams has to himself push some button, fine, but if he doesn’t push the button per the vote, then he has a legally-binding contract where he donates half his wealth to open-source development (or something).
-
Back-ends are transparent and fixed. Basically every trick they have now, remove it. No admin keys. No upgradability. As little as possible off-chain. Truly decentralized infrastructure would be great… Dunno if it exists. I mean, it does in name, the number of DePIN projects out there is insane. But how many are truly decentralized? But at the very, least there’s probably a way to use disinterested third-parties, maybe a lot of different DePINs and possibly classical infrastructure providers along with some kind of kill switch if it gets messed with and a High Availability solution to go to the others if it happens. In other words, Disaster Recovery assuming that a human wanting to go against the vote is /a disaster/. Not saying it will be easy, but I am pretty sure a solution can be made that is so whack-a-mole it won’t go down, won’t get changed.
-
Front-ends that are transparent and fixed. Permaweb is probably the easiest way to solve most of that. But just like back-ends, you got the High Availability, whack-a-mole solutions, that include P2P, Tor, you name it. I don’t doubt if there’s the money for it, there’s a solution for it. Point is, even if some powerful country says that memecoin is against our national interest, not much anyone can do about it coming (front end) or going (back end).
Of course, the biggest downside risk is you could spend a bunch of money on something, and after all those audits, it still ships with a bug that bricks it instantly. (Okay, that and a government decides everyone human who made a vote and/or enabled the enaction of that vote should be put in jail. That would be pretty bad.) BUT, at least everyone would know the result would be TRUE (as true as can be) Decentralized Finance.
[EDIT: 10 seconds later, formatting.]