Multi-signature with no address reuse . How coordination is done?

Question

I would like to have both 2-of-3 multi-signature and no address reuse. For example,

3 BTC belong to multi-sig user1 (Officer),user2 (Manager), user3 (CEO) (2of3 multisig) with pubkey1,pubkey2,pubkey3. They send 1 BTC to user 5. Assume zero fees. The change of 2 BTC should be sent to user1,user2,user3 (multi-sig) with pubkey1.1, pubkey2.1, pubkey3.1

Well, this is easy to say here! but

In real life, how do the users in multi-sig coordinate their next pub key? Would it be a coordination nightmare? What happens when one of the user (CEO) always not available for any communication? We will be stuck with an old Public Key for (user3)? Is there any automatic coordination like BIP32 ? etc.

Answer 1

Some multisig wallets use BIP 45 or otherwise do something similar if they aren't using those specific derivation paths. Such wallets will have the cosigers share their extended public keys with each other. From those xpubs, they can derive each others child keys by themselves. They then agree that all the keys derived at index 0 are for address 0, index 1 for address 1, etc.

They may then also use BIP 67 to determine the order of the keys. BIP 67 just sorts the keys so they are in lexicographic order.

Alternative to BIP 67 is to just have the pubkeys in a predetermined order. I.e. the key derived from person 1's xpub is 1st. The key from person 2 is 2nd. And so on.

This way the participants have determined which keys to use for a particular index, and the order to put the keys. They can now construct the redeemScript and thus the multisig address. All of this can occur without communicating again.