FIP-0062: Fallback Method Handler for the Multisig Actor

[vyzo]

See filecoin-project/builtin-actors#1253
Implementation in filecoin-project/builtin-actors#1252

[arajasek]

What is our motivation to include payment channel actors too? There aren't a lot of situations today where we send to paychs, and the FIP doesn't really motivate this change.

[jennijuju]

approved with nits

[anorth]

A question for the discussion thread, when it exists:

• Will this change also remove the implementation of the FRC-0042 universal receiver hook from the multisig? It should note that by touching the payment channel, that actor will also become a valid receiver of such tokens but those tokens will be lost forever because it can't deal with them. Or, maybe don't touch the payment channel.

Editorially approving this draft, but I maintain the questions above wrt whether this is the right change.

[raulk]

@anorth

Will this change also remove the implementation of the FRC-0042 universal receiver hook from the multisig?

The blanket acceptance of the exported FRC-42 range makes the code for the universal receiver hook redundant. I will mention that here.

It should note that by touching the payment channel, that actor will also become a valid receiver of such tokens but those tokens will be lost forever because it can't deal with them. Or, maybe don't touch the payment channel.

Alternatively we could deny the universal receiver hook? I don't have a bias and I'm happy descoping the paych from the FIP entirely, if that makes us feel better. Payment channels are hardly used on mainnet, so the lack of support won't be limiting.

EDIT: descoped.

UpdateChannelState activity

Settle activity

Collect activity