Most people don't realize there are tokens on Ethereum that predate the ERC-20 standard itself.
In February 2016 — less than a year after Ethereum launched — Alex Van de Sande from the Ethereum Foundation deployed a token called Unicorns. If you donated 2.014 ETH to the Foundation's tip jar, you automatically received a Unicorn. It was one of the first token contracts ever deployed on Ethereum.
A month later, he created Unicorn Meat and the Grinder Association DAO — essentially the first onchain token exchange. You put Unicorns in, you got 1000 Unicorn Meat out. One-way grinder.
These were demo contracts for the Mist browser. They were never meant to have value. But they've survived 10 years on mainnet, and there are still ~185 addresses holding Unicorn tokens.
The problem: Because these tokens have 0 decimal places and non-standard transfer functions, modern DEXes can't handle them. Uniswap V3's fee math literally rounds to zero for 0-decimal tokens. You can wrap them (wrapped versions exist as standard ERC-20s), but the underlying issue persists.
The solution: An onchain orderbook purpose-built for these tokens. No backend, no matching engine, no admin keys. Escrowed limit orders with partial fills, all settled onchain.
The interesting engineering challenge was handling the decimal mismatch between pairs. When a 0-decimal token trades against a 3-decimal token, the price math needs exact precision. The contract uses OpenZeppelin's mulDiv with Rounding.Ceil to ensure the taker always pays at least fair value.
Broader question: There are probably hundreds of non-standard tokens stuck on Ethereum with no trading infrastructure. AMMs just weren't designed for them. Is a lightweight, auditable orderbook the right primitive for this class of assets?
There are tokens older than ERC-20 still sitting on Ethereum mainnet — and they're now tradeable
byu/gorewndis inCryptoTechnology
Posted by gorewndis