I spoke to multicoin awhile back. No mention here but whatever.

It's still a good read.

To add on, here's some thoughts which are guiding the development / design principles for #SwitcheoTradeHub and Demex. https://twitter.com/TusharJain_/status/1303400507483009025

1/ Nothing built solely on Ethereum can become a decentralized bitmex. Anything about future developments is honestly just speculation at this point.

2/ What are the advantages of building solely on Ethereum?

"Hard security", censorship resistance, network of users.

3/ What do I mean by "hard security"? "Hard security" means that the system is provably secure without needing to examine complex game theory, effects of a byzantine L2 / sidechains, fault proofs, data availability etc. etc.

4/ If your smart contract has no bugs and your architecture makes sense, you're done. There are no what-ifs.

If there are, it's because they still do rely on external actors. Maker & Augur vs. Uniswap is a good example.

5/ "Hard security" is *not* important for a decentralized bitmex.

Traders just don't care.

6/ Case in point, "hard security" is made redundant the moment you use something like Chainlink, an external oracle.

If the oracle is byzantine, all your fraud proofs / STARKs in the world are worthless.

$4bn mcap and every team using Chainlink proves that noone cares.

7/ So it's kind of ironic when teams shoehorn themselves into building on Ethereum and end up using Chainlink.

You get stuck with high latency, slow liquidations, with no security benefits.

Design principle #1: Just build your own low latency chain and your own oracles.

8/ What *is* important though, is censorship resistance.

That is a massive priority for us - Being able to scale our trading engine on a large number of validators is our design principle #2.

9/ When launching Switcheo TradeHub, we were initially worried that we would not be able to get enough validators on board to secure the chain.

Turns out, we didn't need to worry.

10/ I have been keeping quiet on this, but I've actually been super stoked to see so many community / anonymous validators for Switcheo TradeHub (with minimal comms) popping up.

This contributes greatly to the resilience of the entire system against shutdowns in the long run.

11/ Further to that, I've been amazed by how strong communities can be, even during the recent $SUSHI saga.

We now have a new design principle.

#3: Let the community drive development.

12/ This is also why we brought forward launching our gov module to this month. We want to kickstart this as soon as possible.

There's a lot of ideas brewing now and we can't wait to share them.

13/ Network of users: Ethereum's lead in this aspect mustn't be forgotten.

The number of "woke" crypto users with Metamask installed is an edge that should be leveraged on.

These are the exact users who will try out a decentralized bitmex.

14/ Most non-Ethereum-only solutions will not allow a user to get started immediately with 0 onboarding. There is a setup phase involving a new wallet, creating a L2 account, etc.

On the flip side, approving every trade on Metamask is a pain.

15/ Design principle #4: allow traders already onboarded on Ethereum to get on the platform with 0 frustration.

We're still working on implementing this, but we know of a viable solution right now.

16/ Some additional random thoughts.

- CLOB / AMM: There's really no need to compare and contrast them. Just have both.
- Liquidity & Incentives: Sushi has shown us that liquidity isn't as sticky as it initially seems. This bodes well for newcomers / underdogs like us.

17/ Finally, the @demexchange public beta on testnet should be launching today.

Let me know what you think!