The Nifty Report’s excellent post on the MATH token and two projects inspired by it provide an interesting contrast with “Tokens Equal Text”:
MATH is an Ethereum smart contract that allows people to buy numbers, identified (and in identity with) ERC-721 non-fungible token IDs. Token ID 1 is the number one, and so on. To buy a number higher than 100 (the project’s creator owns the numbers up to 100), you must pay the owners of two existing numbers to generate it by adding their values. It’s a wonderfully absurd example of artificial scarcity and true ownership taken to the extreme. But it’s also an example of creating new forms of property on the blockchain, capturing things that could not previously be treated as commodities. Although people do try to assert ownership over numbers off of the blockchain as well.
Using token identifiers as the entire content and meaning of what they represent makes those tokens very secure – they require no off-chain resources to establish the significance and value that they assert merely by existing. They are self-describing or self-encoding.
The RGB token builds on MATH to create 16×16 coloured bitmaps by using the binary digits of three MATH numbers for their red, green and blue components and paying their owners for the privilege. It’s still an entirely on-chain token – no metadata from a web server or IPFS required – but it doesn’t encode the information directly into the token ID. The token ID is linked to each of those numbers elsewhere in the contract’s data instead. That makes sense as there isn’t enough room in the Ethereum number type used for token IDs to store three token IDs, but a monochrome bitmap of the same size or a much smaller coloured bitmap could be stored that way.
Likewise the WORDS contract also mentioned by The Nifty Report uses a similar join-and-pay scheme to that used by MATH in order to generate and purchase new words while storing them separately from their token IDs like RGB. As with bitmaps, words can be stored directly in a token ID. Depending on how long they are, several can be store in a single token ID as long as they don’t take up more than the 32 bytes of space available in the number type used to represent token IDs in ERC-721. Not doing this means that, like RGB, WORDS is a token with a significance that is purely on-chain but is not purely ID-based. Neither RGB nor WORDS are not self-encoding.
Tokens Equal Text’s tokens consist of words that are self-encoding. Its ERC-21 contract uses short fragments of text encoded as token ID numbers. It then assembles these these using an EC-998 composable token contract to create descriptions of imagined Vaporwave artworks. The colours for each token ID / each piece of the composition are generated by taking the first few hexadecimal digits of the hash of the token ID and treating that as an RGB colour, extracting surplus value although but surplus meaning from the code of the token IDs. This is a more complex structure than than MATH, RGB or WORDS but has a flatter creation structure – I have minted and composed all the tokens that will be available in the series and owning each token implies no residual rights.
Self-encoding tokens are both conceptually interesting and operationally robust. We’ve only started to see what they can do.
For more information about Tokens Equal Text see here:
Or to buy one of the pieces head over to OpenSea: