Allow users to generate a random private key (recovery phrase) on-device once triggered via rabbit-hole. Allow for submission of a 12/24 word recovery phrase if desired (such as from another r1).
Start with Bitcoin support, allow users to scan BTC address or BTC invoice QR codes with the camera and say “send XXX Bitcoin to this address”/“Send Bitcoin to this invoice”.
They should also be able to generate a QR code to send BTC that is held on the R1 (or rather, can be sent using the R1, as the R1 holds the keys).
Users can type in their device pin to authorize the transaction.
Later, enable support for the BTC lightning network, allowing for point-of-sale payments in countries where BTC usage is normal (such as El Salvador).
If security is a concern, you could have it be a 2/2 multi-signature wallet, with one key held on the user’s device, and another key held on Rabbit servers.
I.m.h.o. this feature request opens a huge number of risks. E.g. recovery of the wallet in case of hardware issues. What happens when the r1 is reset? Would the wallet would be lost? Who is then liable for the lost crypto in the wallet.
Also, if the rabbithole becomes the backup of the on-device wallet, then the rabbithole also becomes a serious target for hacks.
For me this is maybe the most important reason: Maintaining crypto wallet functionality will also cost lots of development resources, resources that should be prioritized for features that will be used by a majority of the community.
Rabbithole doesn’t become a backup of the on-device private key. The user is responsible for backing up the device’s key like with any other crypto wallet. Recovery to a new r1 or external wallet is possible by recovering the private key.
Do you say that just because a feature is “hard” to develop, then it isn’t worth making? I could say the same thing about multi-language support (yet they did it).
There are many audited, open-source out-of-the-box libraries available for basic crypto wallets. And the r1 would not be a place you store hundreds of thousands of dollars — you move over a few hundred/couple thousand dollars, and use it like a checking account.
The difficulty to develop or the complexity to develop is already going to be high for features in development. The team has already mentioned that they are working on something such as 1Password integration and/or credit/debit card connections for LAM training. Do you say they shouldn’t do these things because they are high risk?
And saying a feature that “wouldn’t be used by a majority of the community” is short sighted. Just because you don’t or wouldn’t use it doesn’t speak for the world. BTC is near it’s all time high and debatably going higher with the new US administration. Many casual users do not already have experience with cryptocurrencies, and rabbit starting out as a BTC wallet for payments would be a good entry point to introduce both kids and adults to it. This would of course be an optional feature for advanced users.
TL;DR: the complexity of this feature is not much higher than any other feature that is currently being developed. And i think there are higher risk features that are being developed — such as giving rabbit/LAM access to your accounts, sharing your data (including passwords), the fact that all your queries are stored on their servers… this is not that out of the question.