Signing message with multisig wallet #2408
Comments
|
I know of no standard library for doing such a thing. So if it were implemented in Electrum, only Electrum would be able to verify it until every other wallet supported it. |
|
you do not sign a message with an address, but with a public key. |
|
How can you easily verify your multisig wallet partner still has access to their keys? I was hoping the signature interface would allow me to sign something with the individual private key of one party, in such a way that the other party could verify it. Is that not possible? |
|
@fresheneesz There is currently no way in the GUI/CLI to sign/verify a message with a privkey/pubkey; only "with" an address. If someone implements this, it could get merged, most certainly for the CLI. It does not look difficult to do. If you want a workaround right now, you can encrypt a message to your cosigner's pubkey, and if they can decrypt it and tell you the contents, they still have the key. |
|
That sounds like a reasonable workaround. Doesn't look like an intuitive workflow tho. In particular, I'm seeing the following usability issues:
The last problem there is the biggest blocker, since at this point I don't know how to encrypt something with this multisig wallet. |
I see. That's a bug. Please open a separate issue for it. |
* Miscellaneous fixups to synchronizer.py and wallet.py - Updated copyright in wallet.py and synchronizer.py - Fixed a FIXME in synchronizer.py - Fixed miscellanous python formatting/style issues found by PyCharm - Added some typing annotations to make code more readable * Refactor of synchronizer.py to support tracking of change addresses This is an initial step in order to allow us to only subscribe to N change addresses at a time. * Fixed a comment that was too long * Added mechanism to network.py to unsubscribe from addresses New method in class Network: unsubscribe_from_scripthashes. If all callbacks for a scripthash are gone, will send a blockchain.scripthash.unsubscribe message to the server. This will be used later to reap old change addresses. * Unsubscribe from individual addresses on wallet close This should keep the subs count down for unused addresses when wallets are closed. If the refct for a scripthash drops to 0, we send blockchain.scripthash.unsubscribe to the server. Note that an individual address may be "seen" or used by multiple wallets so we only unsub when a scripthash's refct drops to 0 (meaning no wallet cares about it anymore). * Minor optimization in synchronizer.py subscribe_to_addresses Don't build a list and then convert it to a dict. Build the dict directly. This saves CPU. * Get rid of checking if history is sorted properly This check wastes CPU cycles and doesn't actually amount to any functional change in Electron Cash. Moreover, the check is not guaranteed to work correctly in the case of mempool tx's appearing in history, since the mempool txns have an unspecified order. * Added tracking of change addresses and expiration candidates to synchronizer.py These are the data structures we will use to expire change address subscriptions. * Implemented expiring old change address subs in synchronizer.py If Synchronizer is constructed with positive, nonzero limit_change_subs, then it will take an alternate code path to expire old change addresses. It still runs through the entire change address list, and does briefly subscribe to all change addresses to detect changes, however it then unsubscribes when limit_change_subs is exceeded for change addresses. * Added ability to enable/disable retiring of change in QT GUI Default is enabled with a threshold of older than: 1000 * Refactor: Precede "private" methods in synchronizer.py with '_' * Added periodid check that we are not over chagne addr subs limit * Small nit/optimization Create the huge int once for the sorter in _check_change_subs_limits (should save CPU cycles). * Added custom messaging when changing change retire settings Tell user that they need to close and reopen the wallet (rather than all of electron cash), in order for settings to take effect. * Optimization: Persist the retired change addresses to wallet storage This allows us to have very fast startup times in the happy case where the retired change addresses have no new history. Note that now it is possible for the wallet to not realize if it was sent funds to a "retired" change address. Such is life. User can disable the feature to revert to old behavior or they can select "rebuild wallet history" in that corner case. * Update address tab GUI code to gray out "retired" change addresses This at least provides the user with some visual indicator of what is happening. * Only load the retired change set if using the limit_change option Otherwise don't bother loading the set as it may cause trouble. * Fixed undefined variable bug
Hi,
I would like to sign message from a multisig wallet.
Ideal solution would be using the multisig "3" address and passing the partially signed message around cosigner (the sameway transaction are signed)
Less ideal solution would be to sign with a single sig address "1" derived from every cosigner xpub.
I tried to derive /0/0 from my xpub but when I try to sign a text message with that address I receive "Address not in wallet"
Is it somehow possible to sign text messages with a multisig wallet?
The text was updated successfully, but these errors were encountered: