“Solana is the Promised Land for Blockchain” ⁠— An Exclusive Interview with Solend Founder Rooter

In June 2021, when Solend finished in first place at the Solana Season hackathon, the team had only built a prototype of its lending protocol. Fast forward to August, the open-source project officially launched on the Solana mainnet and has now amassed over $100 million in user deposits and $40 million in borrowings.

Rooter is the anonymous developer leading the Solend team and we recently caught up with him to discuss their rapid growth since the hackathon. In this interview, he shares the challenges and joys of building on Solana, and his thoughts on the recent Solana downtime and technical challenges.

We also dived deep into Solend’s operations, covering oracle issues, token listings, and the upcoming launches on their roadmap. He also answered the most pressing question for Solend users – wen token?

First, though, we talked about how Solend was born.

“We were looking at the ecosystem since Solana is a very powerful blockchain and we felt that Ethereum is kind of for whales and big companies now. It’s not really for the average person anymore; they’ve been priced out.

“We were looking at Solana and lending was a very clear thing that needed to be built. Given the size of Aave and Compound on Ethereum, we knew that a lending protocol would be pretty successful on Solana. It was a very clear opportunity, and although there are so many other competitors doing the same thing, I think Solend is doing pretty well and giving everything.”

What Are the Challenges With Building on Solana?

“I had a lot of experience building things on Ethereum before. I built some smart contracts and stuff. But then of course Solana is new. For most of the people on our team, Solana was the first blockchain development that they did.” 

“So, I think it’s worked out decently well because no matter what, no one is experienced with Solana because of how new it is. Some companies would normally look for someone with five years experience, but it is not even possible to get on Solana. We’d have to train people and people have to learn anyway which people on our team are doing, so they can use my experience with smart contract development in the past to advise and coach them a little.”

“One of the differences between Solana and Ethereum is that the developer tooling ecosystem is not as material. There are some smaller things missing like, when we first started there wasn’t a really good multisig wallet solution or DAO management software. I think those will come soon enough, but what it does is slow us down a little bit because we have to build it ourselves or wait for someone else to build it.”

What is the Best Thing About Building on Solana?

“One of the most amazing moments for me was when we were building the hackathon. We built the prototype of Solend and it basically was on devnet. I was using the app that we built ourselves in just a couple of weeks and I was really amazed by how well it worked. Transactions just got confirmed instantly and it was cheap and cost almost nothing. 

It was really like a moment of magic for me because this is what we’ve been wanting with Ethereum all these time.

It’s kind of like what Ethereum was like in 2018 and 2019 when it was still really cheap back then and you could pay like ten cents and your transactions would get confirmed in like ten seconds. Building on Solana is like that but even better. 

“I only had that experience with Ethereum because I was using it years ago, but most of the people who started using Ethereum recently (that is most people in general), they only know Ethereum as a really slow and expensive beast.

I think in some ways it feels like Solana is the promised land of blockchain. 

Thoughts on the Recent Solana Downtime

There have been two issues with Solana lately. First, the 17-hour downtime and even before and after that there are issues with the RPC nodes when users access Solana or any other Solana app, the transactions get sent to their RPC nodes which are overloaded and they just get like lost and don’t go through and there’s just was like a denial of service. 

Rooter said, “The whole downtime was not an ideal thing to be happening for a network of this size now because there’s quite a bit of value locked (around $4 billion), it’s not the little leagues anymore where there’s nothing at stake.” 

“I think we were pretty lucky that the market didn’t move too much.  We didn’t have many liquidations that should have happened that didn’t. The worst case scenario would have been SOL crashes by 20-30%. No liquidations could happen for a whole day and then things come back online and then everything is liquidated all at once and then maybe causing cascading liquidations and a market crash.”

“At Solend, we increased the liquidation threshold for all of our assets temporarily. We did it as a precaution although it turned out it wasn’t necessarily needed because the market came back to where it was. Overall it was definitely quite distressing and concerning and I definitely hope that this kind of thing doesn’t happen again.”

Solend Uses Pyth Network Oracles (On Sept. 20, The Oracle’s BTC/USD Feed Saw a Flash Crash that Temporarily Quoted the Price of Bitcoin at $8k. We Asked Rooter 0x How Solend Managed the Incident

He said, “We’ve analyzed all of the transactions that happened within the time period. We’re still finishing up our investigations on it and it seems like only one user was affected, liquidated for around $100-$200 (A refund will be made through Solend’s in-house insurance fund). Once again, we’re still lucky that it didn’t affect a broader set of people. We’re still investigating why it happened that way and why no one else was affected.” 

“When we were building Solend, Pyth didn’t have time-weighted average price (TWAP oracles) and the default was to use spot prices. Now that they have TWAPs, we definitely want to look into upgrading our oracles to use a TWAP, although it is definitely hard to do because Solend is already live and it is very difficult to make protocol updates.”

“There is also the potential of using other oracles. There’s Switchboard.xyz as well and Chainlink is coming online pretty soon and they have a lot of security features that would be good to use. We’re gonna look into building with other oracles and also introducing other security features that might improve the Pyth oracle.” 

The Solend Insurance Fund

For a lending protocol, an active bug bounty and especially an insurance fund is a primary feature. It gives users confidence that they’d be refunded if there’s a bug or issue with the platform. Rooter 0x spoke to us broadly on that subject.

“There’s so many things that could go wrong with a lending protocol. For instance, we have to be very careful about any asset that we list. Even if it’s a legit project, their token can still have mintability which is behind the multisig. A hacker could steal the multisig and then mint an infinite amount of tokens and borrow everything out of Solend. Basically, every single asset we list could rekt the whole protocol, and this is why we are very careful about listings.

“There are multiple single points of failure which we have to manage, so we have to be extremely careful about. Everything works fine until it doesn’t.” 

“There’s a lot of other lending protocols on Solana and I suspect that some of them are working with a higher level of risk tolerance. Maybe they’re moving really quickly and not scrutinizing things as much as we are. It will work fine in the short-term and maybe they’ll even get ahead in some metrics.”

“But I think in the long-term, all it takes is just one event for the entire thing to get ruined. We want to be really careful and treat Solend development like rockets. If you’re sending a rocket into space you don’t just check things once or twice you check like ten or twenty times. That’s the kind of attitude we want to have.”

“The [Solend] insurance fund is smaller now since it is just the fees generated by the protocol, but we plan to leverage our token generation event like our DAO treasury to build up an insurance fund. The Solend Treasury is going to be the insurance fund, and the plan for the future is to grow an insurance fund for any type of issue. We plan to grow an insurance fund of around $10 million to $50 million.”

The Solend Roadmap

The next plan on the Solend roadmap is the launch of a token and liquidity mining program. The project also hopes to offer permissionless and isolated lending pools to allow listings of more assets and lowers risk on major pools.

(Solend Dashboard UI)

Unlike Solend’s current cross-chain pool where users can deposit one asset and borrow another, an isolated lending pool allows for users to lend and borrow a single asset. Under such conditions, if a hacker steals a multisig to mint an infinite amount of  a token, they can only remove everything from the token pool without putting all the lenders on Solend at risk.

He confirmed that Solend would be interested in listing Saber Protocol (SBR) as it is a token that a lot of people are interested in. However, the project does not have enough trading volume and liquidity for Solend to actively manage lending, borrowing and liquidation of the asset. The future launch of isolated lending pools would allow for such and even more listings to take place on Solend.

Wen Solend Token?

Unlike some DeFi projects, Solend has taken the hard path by choosing to have a working product before its token generation event.  A common question among users is when the token will launch and if there will be a retroactive airdrop for early adopters.

“It [the token] should be out by next month. Our initial goal was to have it out at the end of September, but we still have a bit of work to do, because some things are taking longer than we expected but basically, it should be out very soon “like next month.”

“Anyone who has been supplying or borrowing” on Solend would qualify for some retroactive rewards. Also, our liquidity program is gonna look a little different from what people are used to as with Compound, Aave, Sushi, and Yearn Finance.” 

“We view all of these as the first version of liquidity mining, which was really great for certain goals, but there are flaws with it. We want to address those flaws with our V2 which looks very different from the traditional one. We’re excited about it and it’s gonna be really positive overall and we are excited to share more information about that soon.”