Re: [bitcoin-dev] BIP proposal, Pay to Contract BIP43 Application

[ZmnSCPxj <ZmnSCPxj@protonmail.com>]

Good morning Omar,

BIP32 includes this text:

> In case parse_256(I_L) >= n or K_i is the point at infinity, the resulting key is invalid, and one should proceed with the next value for i.

This seems to suggest that it is possible for an attacker with sufficient compute power to find two contracts whose derivations alias each other if we "proceed with the next value for i".


More generally, have you considered the possibility of multiple separate contracting systems?

It may be possible to have a particular sequence of bytes that has a valid interpretation under one contracting system, that also has a valid interpretation under another contracting system.
I bring this up here: https://github.com/rgb-org/spec/issues/61
and: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2018-September/016354.html

It would then be possible to fool some victim into thinking it has committed to some innocuous contract in one contracting system, only to reveal later that the same sequence of bytes encoding that innocuous contract also corresponds to a more vicious contract in another contracting system.

Regards,
ZmnSCPxj


Sent with ProtonMail Secure Email.

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Tuesday, March 12, 2019 1:53 PM, Omar Shibli via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:

> Dear Gregory,
>
> First of all, I would like to express my deep appreciation to your entire craft in the FOSS ecosystem, specially in Bitcoin, even more In Blockstream.
> I think you are a brilliant engineer and very principled leader. your efforts are an inspiration for many, a truly enduring forever mark in history of FOSS.
>
> I've submitted fixes to your concerns here:
> https://github.com/bitcoin/bips/commit/b63ed0e17e872b7e7b8634591b0ddfa3dedfdc73#diff-deacf3a22d788a10ce12e4d92ee814ff
>
> Would appreciate your review.
>
> On other note, I still think that this security fix is redundant, I believe CKD function (BIP32) does encapsulate sufficient amount of entropy, but due to lack of formal knowledge and assistance, I've not managed to get formal proof, so I fallback'ed to add this patch for security reasons.
>
> Best regards,
> Omar
>
> On Fri, Sep 1, 2017 at 10:16 AM Omar Shibli <omarshib@gmail.com> wrote:
>
> > Hello Gregory,
> >
> > Thanks for you feedback.
> >
> > The BIP has been updated to explicitly specify the multiparty key derivation scheme which hopefully addresses your concerns.
> >
> > Please have a look at the updated draft of the BIP at the link below:
> >
> > https://github.com/commerceblock/pay-to-contract-protocol-specification/blob/master/bip-draft.mediawiki
> >
> > Any feedback is highly appreciated.
> >
> > Regards,
> > Omar
> >
> > On Tue, Aug 15, 2017 at 7:40 PM, omar shibli <omarshib@gmail.com> wrote:
> >
> > > Thank you for your time Gregory, I really appreciate that.
> > >
> > > What we are describing here is a method to embed cryptographic signatures into a public key based on HD Wallets - BIP32.
> > > In a practical application, we should have two cryptographic signatures from both sides, I don't think in that case your scenario would be an issue.
> > >
> > > More specifically in our application, we do the following construction:
> > >
> > > contract base: m/200'/0'/<contract_number>'
> > > payment base (merchant commitment): contract_base/<merchant_contract_signature>
> > > payment address (customer commitment): contract_base/<merchant_contract_signature>/<customer_contract_signature>
> > >
> > > payment address funds could be reclaimed only if the customer_contract_signature is provided by the customer.
> > >
> > > In terms of durability, our app is pretty simple at this point, we don't store anything, we let customer download and manage the files.
> > >
> > > I will update the BIP to address your concerns.
> > >
> > > On Tue, Aug 15, 2017 at 8:12 AM, Gregory Maxwell <greg@xiph.org> wrote:
> > >
> > > > This construction appears to me to be completely insecure.
> > > >
> > > > Say my pubkey (the result of the derivation path) is P.
> > > >
> > > > We agree to contract C1.   A payment is made to P + G*H(C1).
> > > >
> > > > But in secret, I constructed contract C2 and pubkey Q and set P = Q + G*H(C2).
> > > >
> > > > Now I can take that payment (paid to Q + G*(C1) + G*H(C2)) and assert
> > > > it was in act a payment to P' + G*H(C2).   (P' is simply Q + G*H(C1))
> > > >
> > > > I don't see anything in the proposal that addresses this. Am I missing it?
> > > >
> > > > The applications are also not clear to me, and it doesn't appear to
> > > > address durability issues (how do you avoid losing your funds if you
> > > > lose the exact contract?).
> > > >
> > > > On Mon, Aug 14, 2017 at 6:05 AM, omar shibli via bitcoin-dev
> > > > <bitcoin-dev@lists.linuxfoundation.org> wrote:
> > > > > Hey all,
> > > > >
> > > > > A lot of us familiar with the pay to contract protocol, and how it uses
> > > > > cleverly the homomorphic property of elliptic curve encryption system to
> > > > > achieve it.
> > > > > Unfortunately, there is no standard specification on how to conduct such
> > > > > transactions in the cyberspace.
> > > > >
> > > > > We have developed a basic trade finance application that relies on the
> > > > > original idea described in the Homomorphic Payment Addresses and the
> > > > > Pay-to-Contract Protocol paper, yet we have generalized it and made it BIP43
> > > > > complaint.
> > > > >
> > > > > We would like to share our method, and get your feedback about it, hopefully
> > > > > this effort will result into a standard for the benefit of the community.
> > > > >
> > > > > Abstract idea:
> > > > >
> > > > > We define the following levels in BIP32 path.
> > > > > m / purpose' / coin_type' / contract_id' / *
> > > > >
> > > > > contract_id is is an arbitrary number within the valid range of indices.
> > > > >
> > > > > Then we define, contract base as following prefix:
> > > > > m / purpose' / coin_type' / contract_id'
> > > > >
> > > > > contract commitment address is computed as follows:
> > > > > hash document using cryptographic hash function of your choice (e.g. blake2)
> > > > > map hash to partial derivation path
> > > > > Convert hash to binary array.
> > > > > Partition the array into parts, each part length should be 16.
> > > > > Convert each part to integer in decimal format.
> > > > > Convert each integer to string.
> > > > > Join all strings with slash `/`.
> > > > > compute child public key by chaining the derivation path from step 2 with
> > > > > contract base:
> > > > > m/<contract_base>/<hash_derivation_path>
> > > > > compute address
> > > > > Example:
> > > > >
> > > > > master private extended key:
> > > > > xprv9s21ZrQH143K2JF8RafpqtKiTbsbaxEeUaMnNHsm5o6wCW3z8ySyH4UxFVSfZ8n7ESu7fgir8imbZKLYVBxFPND1pniTZ81vKfd45EHKX73
> > > > > coin type: 0
> > > > > contract id: 7777777
> > > > >
> > > > > contract base computation :
> > > > >
> > > > > derivation path:
> > > > > m/999'/0'/7777777'
> > > > > contract base public extended key:
> > > > > xpub6CMCS9rY5GKdkWWyoeXEbmJmxGgDcbihofyARxucufdw7k3oc1JNnniiD5H2HynKBwhaem4KnPTue6s9R2tcroqkHv7vpLFBgbKRDwM5WEE
> > > > >
> > > > > Contract content:
> > > > > foo
> > > > >
> > > > > Contract sha256 signature:
> > > > > 2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae
> > > > >
> > > > > Contract partial derivation path:
> > > > > 11302/46187/26879/50831/63899/17724/7472/16692/4930/11632/25731/49056/63882/24200/25190/59310
> > > > >
> > > > > Contract commitment pub key path:
> > > > > m/999'/0'/7777777'/11302/46187/26879/50831/63899/17724/7472/16692/4930/11632/25731/49056/63882/24200/25190/59310
> > > > > or
> > > > > <contract_base_extended_pub_key>/11302/46187/26879/50831/63899/17724/7472/16692/4930/11632/25731/49056/63882/24200/25190/59310
> > > > >
> > > > > Contract commitment pub key:
> > > > > xpub6iQVNpbZxdf9QJC8mGmz7cd3Cswt2itcQofZbKmyka5jdvQKQCqYSDFj8KCmRm4GBvcQW8gaFmDGAfDyz887msEGqxb6Pz4YUdEH8gFuaiS
> > > > >
> > > > > Contract commitment address:
> > > > > 17yTyx1gXPPkEUN1Q6Tg3gPFTK4dhvmM5R
> > > > >
> > > > >
> > > > > You can find the full BIP draft in the following link:
> > > > > https://github.com/commerceblock/pay-to-contract-protocol-specification/blob/master/bip-draft.mediawiki
> > > > >
> > > > >
> > > > > Regards,
> > > > > Omar
> > > > >
> > > > > _______________________________________________
> > > > > bitcoin-dev mailing list
> > > > > bitcoin-dev@lists.linuxfoundation.org
> > > > > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
> > > > >