12:19:25 <br-m> <rbrunner7> I never wondered about this, until today: In the CLI wallet app, if you create a new wallet by entering a filename that does not yet exist, you can't specify a seed offset passphrase right away. You have to restore to introduce one, as far as I can see. Which is both not comfortable, and not obvious.
12:21:09 <br-m> <rbrunner7> Is this a bug or a feature, so to say? Would there be fear that if we present the seed offset passphrase feature right away it gets too obvious, with bad people noticing too easily that there is such a feature, and that therefore there may be "decoy wallets"?
12:24:08 <br-m> <rbrunner7> Or, people not understanding, entering something, and later never get their wallet restored properly again? People mistaking it for a second password, or the password yet again, or even worse misunderstandings?
12:30:24 <br-m> <tobtoht> It's a power user feature and should only be requested if the user explicitely requests it. E.g. via a --set-passphrase command line flag.
12:32:10 <br-m> <rbrunner7> Interactive wallets have more options here. Feather Wallet has an Options button on the create new wallet dialog.
12:32:30 <br-m> <rbrunner7> Or, more possible ways how to offer it, but not too obviously and "in your face"
12:38:34 <br-m> <tobtoht> It's unfortunate that "passphrase" has become the common term for this feature. It is too similar to password. In Electrum, it is described as: "extend seed with custom words" which makes it much easier to intuit what it does.
12:39:15 <br-m> <rbrunner7> Agree.
13:52:54 <br-m> <rbrunner7> I took the following technical question from Monday's meeting: Assume we do not store passphrases in the core software's Polyseed implementation. Can we, with reasonable effort, still make sure that m_passphrase as currently written by monero_c based wallet apps survives a load and write by the core software, and is still available to query?
13:55:05 <br-m> <rbrunner7> I tried, and the answer is "yes". It takes only a few lines of code and declarations to read and write the monero_c style passphrase in a compatible way, without making wipeable_string serializable, and have it available e.g. as account_keys::m_monero_c_passphrase that the core software otherwise totally ignores.
14:18:15 <br-m> <sneedlewoods_xmr:matrix.org> @tobtoht: AFAICT there is no flag like this, it's like rbrunner7 said, you'd have to use something like --restore-from-seed if I'm not mistaken
14:19:24 <br-m> <rbrunner7> Yup, that would be new startup parameter. Maybe something for a future PR ...
14:20:09 <br-m> <rbrunner7> We have so few of them, could somebody please add few more :)
14:20:39 <br-m> <jpk68:matrix.org> It would be nice to have though ;)
14:20:49 <br-m> <tobtoht> Does the current Polyseed PR not ask for or allow setting a passphrase?
14:21:11 <br-m> <rbrunner7> No, not at wallet creation. Symmetric between legacy seed and Polyseed: No question.
14:22:06 <br-m> <rbrunner7> And no "setting". That would change the spend private key, no? Only restore with passphrase.
14:26:47 <br-m> <tobtoht> Ok, restoring with passphrase is possible, that's fine. Not sure if we really need an extra flag just to be able to do it at creation.
14:27:26 <br-m> <sneedlewoods_xmr:matrix.org> in before jpk68 PR, I'd vote: no :D
14:28:01 <br-m> <jpk68:matrix.org> Hopefully it can be something like Set a passphrase? [y/N] with 'no' selected by default
14:28:17 <br-m> <jpk68:matrix.org> Rather than it unconditionally asking the user to set a passphrase, and just leaving it blank for none
14:28:45 <br-m> <jpk68:matrix.org> But that's just my opinion
14:28:46 <br-m> <tobtoht> Users will mistake this for the password even if you point it out explicitely.
14:30:21 <br-m> <jpk68:matrix.org> What about "Extend this seed with a custom offset?"
14:32:08 <br-m> <rbrunner7> Everybody is an advanced user. Just ask them :)
14:32:27 <br-m> <jpk68:matrix.org> True :D