Feature: Support Mixed-case Checksum address encoding

[sambacha]

See the EIP Specification here: https://github.com/ethereum/ercs/blob/master/ERCS/erc-55.md https://github.com/ethereum/EIPs/blob/master/EIPS/eip-1191.md

EIP-1191¹: This uses ChainId to encode mixed-case checksumming to addresses, allow the same address to be differentiated based on the encoding of character case based on ChainId

Footnotes

1. Correct EIP is 1191, not 55 ↩

[paulmillr]

This one? erc-55 doesn't mention anything about chainid

micro-eth-signer/src/address.ts

Lines 37 to 46 in 3208304

	addChecksum(nonChecksummedAddress: string): string {
	const low = addr.parse(nonChecksummedAddress).data.toLowerCase();
	const hash = bytesToHex(keccak_256(low));
	let checksummed = '';
	for (let i = 0; i < low.length; i++) {
	const hi = Number.parseInt(hash[i], 16);
	const li = low[i];
	checksummed += hi <= 7 ? li : li.toUpperCase(); // if char is 9-f, upcase it
	}
	return add0x(checksummed);

[sambacha]

whoops, meant this one sorry! https://github.com/ethereum/EIPs/blob/master/EIPS/eip-1191.md

[sambacha]

example:

EIP-1191 Checksum for mevETH

0x24aE2dA0F361AA4bE46B48EB19c91e02c5E4F27e

Normal Checksum

0x24Ae2dA0f361AA4BE46b48EB19C91e02c5e4f27E

[paulmillr]

That seems useful.

[paulmillr]

c8a86a6

[sambacha]

Hell yeah, thanks for getting this done so quickly!

[ricmoo]

I’ve just been made aware of these changes.

EIP-1191 is a derelict EIP, has been removed and discouraged from existing libraries and should not be introduced.

Web3.js and EthereumJS (back in 2021) had to wait until a major version to remove it, which I would suspect would be the same here.

The discussion-to link in the EIP covers much of the issues, as well as helps give context why it was never moved from last call to final. It was contention and breaks (entirely) backwards compatibility in the address format, which EIP-155 tried so desperately to maintain.

I would love to carry on this discussion further if necessary and interested though. I thought the EIP had finally been defeated, but now see people talking about it again, so wanted to speak up. :)

[paulmillr]

@sambacha thoughts on reverting this?