Sushi Swap and Web3

Hello, I represent a project ( which is aiming to bring web3 to the masses. I define web3 as the dweb (p2p, ipfs, tor, bittorrent) plus economy/blockchain.

I have been looking at some of the top Ethereum defi projects since the Tornado event/incident and I am finding that almost no “defi” project is ready to be a part of web3 and be as decentralized as it claims to be for the long term.

While unrelated, to prove this point, compound finance is planning on censoring transactions on their frontend, even on IPFS.

For the Sushi project, I am looking to see what is required for the project to:

  1. Get a web3 domain, or use existing on ENS and other web3 blockchain registries. Handshake also comes to mind.
  2. Get an IPFS, Skynet, or even Arweave deployment of the frontend (maybe all 3?) and have the DAO maintain this as part of DevOps. I will openly state that I am biased toward Skynet, but ANY decentralized storage/hosting is better than an ICANN domain & VPS server.
  3. Have the domain(s) point to one of these deployments to benefit the Defi community and prevent censorship of finance.

Lume Web is currently testing a prototype Firefox-only extension (v2 for the project, v1 was in march or so) that will enable access to all of the above (arweave in the future) in a purely decentralized way via web3 technology being actively developed. This includes access to 6 different blockchain registries/domains on 2 content/storage networks at present. If anyone is interested in assisting in testing on that, you can join the server at (top right)

My goal here is to be proactive and assist all major defi protocols that wish to actually be decentralized and not turn into Cefi or even what I have recently heard fairfi as a marketing spin by some lawyers of projects.

Mods, if this is better as an informal proposal/something else, feel free to move, but I see this as simply an introduction and, ideally, an open discussion on real decentralization and what I view as the true web3.



Sushi owns its own infrastructure, provides RPC access via

Web3 Domain is not necessary, however there are existing eth domains used, but not used for public purposes.

Sushi is the only DApp that maintains is own RPC services independent of any other RPC Provider. The Relay does proxy non state changing rpc calls to alchemy to offload requests and to improve cache hits.

Let me know if you have additional questions,


Thanks, however, this is where I feel you are wrong. For web3 to be mainstream, it needs to use domains. Expecting ordinary users to understand or access IPFS or other type URL’s or know where to look limits defi and web3 to geeks/power users and is a bad mentality. We want defi to grow and be more mainstream, like BTC, not be a limited community due to a high entry bar.

Using an ETH domain or any web3 domain pointing to an IPFS or Skynet deployment will ensure actual web3 access to Sushi the day that you are forced to take offline the ICANN version. Waiting until then is also reactive and not proactive, aka caught with your pants down.

I am, however, glad sushi is running its own RPC for the project. Lume Web relies on pocket networks. Though the concern is web3 access as in dweb, not RPC. My definition of web3 is not a blockchain or eth.


Skynet is no longer actively developed, they shut down a few weeks ago.

My issues are with IPFS: sure sushi can publish to IPFS etc, however this doesn’t help secure it, only makes it slightly more available.

ICANN revocation by warrant doesnt solve your issue with Web3 domain name, it is still reliant upon DNS. Unless you are suggesting pointing it at the .eth contract level for the pinned version, the trouble remains you have no way of verifying that it is an authentic deployment, in a easily done way, though we have some ideas, it will have to wait until the new frontend is launched.

The main issue with the frontend is that the translations for sushis dapp add considerable overhead, both by forcing it to use outdated compiling methods and being just overall a shitty implementation. When the nee frontend is ready an automated release process can be implemented putting it on IPFS. However this still doesn’t provide any authentication assurances to the end user that they are dealing with a legitimate deployment.

at the end of the day the relay also hosts its own copy of the the frontend dapp, however this is the latest stable beta of the frontend,

Labs is shutting down Skynet is not. It was a VC company that is spending a few more months on dev, but is bankrupt otherwise. They have not shut down yet. The sia foundation (503c) will be highly likely to continue the development, and they are funded by a blockchain tax. Lume Web uses Skynet technology and continues to build as it can be viewed as a Layer 3. I will admit that Skynet and Sia get a lot of confusion, but Skynet is not dead and is not dying. Skynet is going to be community-developed moving forward.

however this doesn’t help secure it, only makes it slightly more available. Which is the point of much of what defi is for, easy accessibility to trade without censorship.

I will state that skynet will very soon be the 1st and only content network that enables client verification of the files you download without needing the whole file 1st (like IPFS does).

ICANN revocation by warrant doesn't solve your issue with Web3 domain name, it is still reliant upon DNS. The problem is that I agree with you to an extent that a ton of the infrastructure we rely on is centralized and thus not much better than ICANN. Lume Web is solving this, and a big aspect of that is using pocket network and other RPC services if needed that are not infura-like services to be censored. It also means not relying on, as lume web will support decentralizing this.

Unless you are suggesting pointing it at the .eth contract level for the pinned version, yes, point to an IPNS or Skynet resolver Skylink (Skynet version of IPNS).

You can trust a group’s version of a deployment, but the key here is that the RPC to do the DNS cannot be censored, and the code cannot be censored. Lume Web is working on both of these problems. Outside that, it is up to social consensus to build trust in an IPFS hash, user identity, or a web3 domain.

Lastly, to clarify, the intent of all this is not to validate that the app you have is legit, even if that can be important to an extent (though Skynet solves this); it is censorship resistance and does not need to be a geek or technical user to figure out how to gain access.