add BitID support

[mackuba]

https://docs.google.com/presentation/d/19iODrgKsVigsA17csNF3HU75UG9eiI5Z6cX2OczvIsw/pub
https://github.com/bitid/bitid
http://bitid-demo.herokuapp.com/

[EricLarch]

Cool :)
Let me know if I can be of any help.

Here is a video of the BitID UX on the Bitcoin Android wallet :
https://www.youtube.com/watch?v=3eepEWTnRTc

Some notes about BitID metada for wallet implementation :
https://github.com/bitid/bitid/blob/master/bitid_metadata.md

[ghost]

@mattatgit @weilu @jenbennings @haustraliaer @javgh @jsuder

Need to research this and decide if this is something we are seriously considering implementing across all of the products.

[weilu]

@EricLarch any thoughts on how BitID could work with a bip32(HD) wallet?

[EricLarch]

It is just a question of convention about which branches to use as
derivation to build keys associated to a website.
For instance /i/[hash of domain]
BTChip is a HD hardware wallet implementing BitID ; @btchip do you have any
recommendations to share?
We can work all together on a simple spec I'd publish on the BitID repo.
Le 28 juil. 2014 05:35, "Wei Lu" notifications@github.com a écrit :

@EricLarch https://github.com/EricLarch any thoughts on how BitID could
work with a bip32(HD) wallet?

—
Reply to this email directly or view it on GitHub
#367 (comment).

[btchip]

I'm doing something simple today - any derivation containing 0xB11D in a branch is considered as a BitID address and skips the second factor check.

In order to be compliant with the recent standards, I'd suggest to use BIP 43 and define 0xB11D purpose for BitID - then BitID addresses could be encoded as follows to avoid creating a very long path

m / 0xB11D' / account' / IEEE 802.3 CRC 32 of website origin' / account sub index (for possible collisions)

[mackuba]

As far as I can tell, this hasn't really taken off so far. I don't remember seeing any further posts about it since the announcement. In fact, I've seen a competing solution announced by BitPay http://blog.bitpay.com/2014/07/01/bitauth-for-decentralized-authentication.html (which also doesn't seem to be implemented by anyone yet). I would leave this as an idea for the future for now.

[weilu]

@EricLarch @btchip By the time BIP43 came out Hive's wallet part of the code was already stablized. We are currently using the default branch & accounts (m/0'/0 and m/0'/1) as specified in bip32. Is there any way we can make BitID work with the current keychain structure?

[EricLarch]

Right now BitID is implemented only in Mycelium (beta version), Dark Wallet and BTChip. There are numerous contributions on the server side, but few from the wallet perspective. BitAuth is a different approach on the subjet, aiming to connected identities to wallets (BitID is just about proving you own an address, connected or not to a blockchain transaction depending on the context).
Just if you consider BitID as a 2FA replacement of Google Authenticator, having this option on a wallet could be a huge incentive to switch to it.

[btchip]

@wellu the most transparent way would probably be to use a dedicated account for BitID (0xb11d), then iterate an index per new account created on any website.

[mackuba]

I'll think about this if it ever becomes a popular thing, I don't think it makes sense to waste time on this if it ends up never becoming anything more than a proof of concept.

[btchip]

It's a bit of a chicken and egg problem - it's not going to be popular if no wallet supports it :) and it shouldn't be that long to implement if you already support message signing and URI callbacks. But talk is cheap and pull requests are better, I know.