All field notes

Blockchain · Retrospective

From coloured coins to ERC-1155: how I have described a token on different chains

Four different tokens in a two-by-two grid, each with the same small blue dot at its centre.
Concept illustration · AI-generated

Ethereum's first block was mined on 30 July 2015, so the network has just turned ten. Anniversaries invite looking back, so here is mine: the ways I have described tokens, in protocol documentation, contract READMEs and announcements to holders. The words changed with every chain. Three questions underneath them did not. Who can issue more? Who can move it? Who can destroy it?

On a UTXO chain, the answer is in the asset ID

TriveAsset, the token layer on Trivechain, follows the coloured-coins protocol that Colu designed for Bitcoin. When I published its documentation in April 2020, the first question was answered on the page about asset IDs.

Every TriveAsset ID starts with L or U. A locked asset's ID is derived from the particular unspent output that its issuance transaction consumed. An output can be spent only once, so nobody can issue more of that asset, not even the issuer. An unlocked asset's ID is derived from the script of that output instead, so whoever controls the same address can issue more units later. The answer to "who can issue" is in the first character of the identifier.

The other two answers come from the transaction format. An asset travels with the output that carries it, so whoever can spend that output can move the asset. In a burn transaction, a transfer instruction aimed at output index 31 is read as destruction rather than delivery. Its aggregation policy even separated units that merge freely from units that keep their own metadata, the line Ethereum's standards draw between fungible and non-fungible tokens.

On an EVM chain, the answer is in the functions

When TRVC moved to Binance Smart Chain in May 2021, I wrote its BEP20 contract. Only the owner can call mint, and only up to a maximum supply of 80 million TRVC. Any holder can burn their own tokens, and burning also lowers that maximum, so burned supply cannot be minted again. Moving follows the usual pattern: the holder, or a spender the holder has approved. Because the contract's ownership can itself be transferred, "who is the owner" becomes a question of its own.

A month later I wrote a teaching token, DEVK, for a Developer Kaki session. It mints its whole supply in the constructor and has no mint function at all.

Writing those two contracts made something plain that the standard's name hides. ERC-20 does not define minting or burning. It says only that a contract creating tokens should emit a Transfer event from the zero address. Who can issue is whatever the code says, which is why verified source on an explorer matters more than the label on the token.

In a game, the question has two layers

Playermon's creatures were issued as ERC-721 tokens, and in December 2023 items such as breeding pods became tradable as ERC-1155. ERC-721 states that minting and burning are not included in the specification. ERC-1155 specifies only how a mint or a burn must be reported. Both leave issuance to the contract.

In a game, new tokens tend to appear because of something that happens in play. The contract decides which account may mint; the game decides when that account does. Reading the contract answers the first part. The second depends on the studio's keys and processes, so describing a game token honestly means describing both.

On Solana, the answers are fields

When PYM also went to Solana in March 2024, the vocabulary changed again. A Solana token is a mint account with a mint authority, which can create more tokens unless it has been set to none, and an optional freeze authority, which can freeze holders' token accounts. Each token account has an owner and may have a delegate, and either can burn from it. The questions become fields that anyone can read on an explorer, plus one my earlier contracts never had to answer: who can freeze?

A token on two chains adds another: where is the total? Our answer that March was to lock 100 million PYM in a public Safe on Polygon before the same quantity went into a pool on Solana.

What stayed the same

Who can issueWho can moveWho can burn
TriveAssetThe issuer, if the ID starts with U; nobody, if LWhoever can spend the output carrying itThe holder, with a burn instruction
TRVC (BEP20)The contract owner, up to the capThe holder, or an approved spenderThe holder, and the cap falls
DEVK (BEP20)Nobody after deploymentThe holder, or an approved spenderNobody: there is no burn function
Solana tokenThe mint authority, if one is setThe owner or delegate, unless the account is frozenThe owner or delegate

Ten years of Ethereum have produced a great deal of vocabulary. When I describe a token now, I put those three answers first, in plain words, with where to check each one: an identifier, a function or a field. A standard's name tells a reader how a token moves between wallets. It does not tell them who can make more of it.