BIP for deterministic pay-to-script-hash multi-signature addresses
7 messages, 12 Feb to 24 May 2015. Started by Thomas Kerin.
Thomas Kerin (3), Luke Dashjr, Peter Todd, Ruben de Vries, Eric Lombrozo
Hi all,
I have drafted a BIP with Jean Pierre and Ruben after the last discussion, related to a standard for deriving a canonical pay-to-script-hash address given a set of public keys and the number of signatures required. There have been two or three discussions about it on the mailing list to date, and various services already carry out this process. I hope a BIP to describe this process will allow services to declare themselves as BIPXX compliant, working towards interoperability of services and simplifying the derivation of scripts and their addresses by all parties.
BIP: XX Title: Deterministic Pay-to-script-hash multi-signature addresses through public key sorting Author: Thomas Kerin, Jean-Pierre Rupp, Ruben de Vries Status: Draft Type: Standards Track Created: 8 February 2015
==Abstract==
This BIP describes a method to deterministically generate multi-signature transaction scripts. It focuses on defining how the public keys must be encoded and sorted so that the redeem script and corresponding P2SH address are always the same for a given set of keys and number of required signatures.
==Motivation==
Most multi-signature transactions are addressed to P2SH (pay-to-script-hash) addresses, as defined in BIP-0016.
Multi-signature redeem scripts do not require a particular ordering or encoding for public keys. This means that for a given set of keys and number of required signatures, there are as many as 2(n!) possible standard redeem scripts, each with its separate P2SH address. Adhering to a an ordering scheme and key encoding would ensure that a multi-signature “account” (set of public keys and required signature count) has a canonical P2SH address.
By adopting a sorting and encoding standard, compliant wallets will always produce the same P2SH address for the same given set of keys and required signature count, making it easier to recognize transactions involving that multi-signature account. This is particularly attractive for multisignature hierarchical-deterministic wallets, as less state is required to setup multi-signature accounts: only the number of required signatures and master public keys of participants need to be shared, and all wallets will generate the same addresses.
While most web wallets do not presently facilitate the setup of multisignature accounts with users of a different service, conventions which ensure cross-compatibility should make it easier to achieve this.
Many wallet as a service providers use a 2of3 multi-signature schema where the user stores 1 of the keys (offline) as backup while using the other key for daily use and letting the service cosign his transactions. This standard will help in enabling a party other than the service provider to recover the wallet without any help from the service provider.
==Implementation==
For a set of public keys, ensure that they have been received in compressed form, sort them lexicographically according to their binary representation before using the resulting list of keys in a standard multisig redeem script. Hash the redeem script according to BIP-0016 to get the P2SH address.
==Compatibility==
* Uncompressed keys are incompatible with this specificiation. A compatible implementation should not automatically compress keys. Receiving an uncompressed key from a multisig participant should be interpreted as a sign that the user has an incompatible implementation. * P2SH addressses do not reveal information about the script that is receiving the funds. For this reason it is not technically possible to enforce this BIP as a rule on the network. Also, it would cause a hard fork. * Implementations that do not conform with this BIP will have compatibility issues with strictly-compliant wallets. * Implementations which do adopt this standard will be cross-compatible when choosing multisig addressses. * If a group of users were not entirely compliant, there is the possibility that a participant will derive an address that the others will not recognize as part of the common multisig account.
==Test vectors== The required number of signatures in each case is 2.
Vector 1 * List ** 02ff12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f8 ** 02fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f * Sorted ** 02fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f ** 02ff12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f8 * Script ** 522102fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f2102ff12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f852ae * Address ** 39bgKC7RFbpoCRbtD5KEdkYKtNyhpsNa3Z
Vector 2 (Already sorted, no action required) * List: ** 02632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed0 ** 027735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e77 ** 02e2cc6bd5f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b404 * Sorted: ** 02632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed0 ** 027735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e77 ** 02e2cc6bd5f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b404 * Script ** 522102632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed021027735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e772102e2cc6bd5f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b40453ae * Address ** 3CKHTjBKxCARLzwABMu9yD85kvtm7WnMfH
Vector 3: * List: ** 030000000000000000000000000000000000004141414141414141414141414141 ** 020000000000000000000000000000000000004141414141414141414141414141 ** 020000000000000000000000000000000000004141414141414141414141414140 ** 030000000000000000000000000000000000004141414141414141414141414140 * Sorted: ** 020000000000000000000000000000000000004141414141414141414141414140 ** 020000000000000000000000000000000000004141414141414141414141414141 ** 030000000000000000000000000000000000004141414141414141414141414140 ** 030000000000000000000000000000000000004141414141414141414141414141 * Script ** 522102000000000000000000000000000000000000414141414141414141414141414021020000000000000000000000000000000000004141414141414141414141414141210300000000000000000000000000000000000041414141414141414141414141402103000000000000000000000000000000000000414141414141414141414141414154ae * Address ** 32V85igBri9zcfBRVupVvwK18NFtS37FuD
Vector 4: (from bitcore) * List: ** 022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da ** 03e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e9 ** 021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc18 * Sorted: ** 021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc18 ** 022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da ** 03e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e9 * Script ** 5221021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc1821022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da2103e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e953ae * Address ** 3Q4sF6tv9wsdqu2NtARzNCpQgwifm2rAba
==Usage & Implementations== * BIP45 - Structure for Deterministic P2SH Multisignature Wallets - https://github.com/bitcoin/bips/blob/master/bip-0045.mediawiki#address-generation-procedure * Bitcore - https://github.com/bitpay/bitcore/blob/50a868cb8cdf2be04bb1c5bf4bcc064cc06f5888/lib/script/script.js#L541 * Haskoin - https://github.com/haskoin/haskoin/blob/master/Network/Haskoin/Script/Parser.hs#L112-122 * Armory - https://github.com/etotheipi/BitcoinArmory/blob/268db0f3fa20c989057bd43343a43b2edbe89aeb/armoryengine/ArmoryUtils.py#L1441 * Multisignature Brainwallet - http://ms-brainwallet.org/
For now, the BIP will live here: https://github.com/afk11/bips/blob/bip0090/bip-0090.mediawiki/
Looking forward to any feedback and discussions that arise!
- -- Thomas Kerin - -------------------------
My PGP key can be found here <http://pgp.mit.edu/pks/lookup?op=get&search=0x3F0D2F83A2966155>
Where is the Specification section?? Does this support arbitrary scripts, or only the simplest CHECKMULTISIG case?
On Thursday, February 12, 2015 9:42:23 PM Thomas Kerin wrote:
Quoted text (171 lines)
> Hi all, > > I have drafted a BIP with Jean Pierre and Ruben after the last > discussion, related to a standard for deriving a canonical > pay-to-script-hash address given a set of public keys and the number of > signatures required. There have been two or three discussions about it > on the mailing list to date, and various services already carry out this > process. I hope a BIP to describe this process will allow services to > declare themselves as BIPXX compliant, working towards interoperability > of services and simplifying the derivation of scripts and their > addresses by all parties. > > > BIP: XX > Title: Deterministic Pay-to-script-hash multi-signature addresses > through public key sorting > Author: Thomas Kerin, Jean-Pierre Rupp, Ruben de Vries > Status: Draft > Type: Standards Track > Created: 8 February 2015 > > ==Abstract== > > This BIP describes a method to deterministically generate > multi-signature transaction scripts. It focuses on defining how the > public keys must be encoded and sorted so that the redeem script and > corresponding P2SH address are always the same for a given set of keys > and number of required signatures. > > ==Motivation== > > Most multi-signature transactions are addressed to P2SH > (pay-to-script-hash) addresses, as defined in BIP-0016. > > Multi-signature redeem scripts do not require a particular ordering or > encoding for public keys. This means that for a given set of keys and > number of required signatures, there are as many as 2(n!) possible > standard redeem scripts, each with its separate P2SH address. Adhering > to a an ordering scheme and key encoding would ensure that a > multi-signature “account” (set of public keys and required signature > count) has a canonical P2SH address. > > By adopting a sorting and encoding standard, compliant wallets will > always produce the same P2SH address for the same given set of keys and > required signature count, making it easier to recognize transactions > involving that multi-signature account. This is particularly attractive > for multisignature hierarchical-deterministic wallets, as less state is > required to setup multi-signature accounts: only the number of required > signatures and master public keys of participants need to be shared, and > all wallets will generate the same addresses. > > While most web wallets do not presently facilitate the setup of > multisignature accounts with users of a different service, conventions > which ensure cross-compatibility should make it easier to achieve this. > > Many wallet as a service providers use a 2of3 multi-signature schema > where the user stores 1 of the keys (offline) as backup while using the > other key for daily use and letting the service cosign his transactions. > This standard will help in enabling a party other than the service > provider to recover the wallet without any help from the service provider. > > ==Implementation== > > For a set of public keys, ensure that they have been received in > compressed form, sort them lexicographically according to their binary > representation before using the resulting list of keys in a standard > multisig redeem script. Hash the redeem script according to BIP-0016 to > get the P2SH address. > > ==Compatibility== > > * Uncompressed keys are incompatible with this specificiation. A > compatible implementation should not automatically compress keys. > Receiving an uncompressed key from a multisig participant should be > interpreted as a sign that the user has an incompatible implementation. > * P2SH addressses do not reveal information about the script that is > receiving the funds. For this reason it is not technically possible to > enforce this BIP as a rule on the network. Also, it would cause a hard > fork. > * Implementations that do not conform with this BIP will have > compatibility issues with strictly-compliant wallets. > * Implementations which do adopt this standard will be cross-compatible > when choosing multisig addressses. > * If a group of users were not entirely compliant, there is the > possibility that a participant will derive an address that the others > will not recognize as part of the common multisig account. > > ==Test vectors== > The required number of signatures in each case is 2. > > Vector 1 > * List > ** 02ff12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f8 > ** 02fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f > * Sorted > ** 02fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f > ** 02ff12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f8 > * Script > ** > 522102fe6f0a5a297eb38c391581c4413e084773ea23954d93f7753db7dc0adc188b2f2102f > f12471208c14bd580709cb2358d98975247d8765f92bc25eab3b2763ed605f852ae * > Address > ** 39bgKC7RFbpoCRbtD5KEdkYKtNyhpsNa3Z > > Vector 2 (Already sorted, no action required) > * List: > ** 02632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed0 > ** 027735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e77 > ** 02e2cc6bd5f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b404 > * Sorted: > ** 02632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed0 > ** 027735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e77 > ** 02e2cc6bd5f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b404 > * Script > ** > 522102632b12f4ac5b1d1b72b2a3b508c19172de44f6f46bcee50ba33f3f9291e47ed021027 > 735a29bae7780a9755fae7a1c4374c656ac6a69ea9f3697fda61bb99a4f3e772102e2cc6bd5 > f45edd43bebe7cb9b675f0ce9ed3efe613b177588290ad188d11b40453ae * Address > ** 3CKHTjBKxCARLzwABMu9yD85kvtm7WnMfH > > Vector 3: > * List: > ** 030000000000000000000000000000000000004141414141414141414141414141 > ** 020000000000000000000000000000000000004141414141414141414141414141 > ** 020000000000000000000000000000000000004141414141414141414141414140 > ** 030000000000000000000000000000000000004141414141414141414141414140 > * Sorted: > ** 020000000000000000000000000000000000004141414141414141414141414140 > ** 020000000000000000000000000000000000004141414141414141414141414141 > ** 030000000000000000000000000000000000004141414141414141414141414140 > ** 030000000000000000000000000000000000004141414141414141414141414141 > * Script > ** > 522102000000000000000000000000000000000000414141414141414141414141414021020 > 000000000000000000000000000000000004141414141414141414141414141210300000000 > 000000000000000000000000000041414141414141414141414141402103000000000000000 > 000000000000000000000414141414141414141414141414154ae * Address > ** 32V85igBri9zcfBRVupVvwK18NFtS37FuD > > Vector 4: (from bitcore) > * List: > ** 022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da > ** 03e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e9 > ** 021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc18 > * Sorted: > ** 021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc18 > ** 022df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da > ** 03e3818b65bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e9 > * Script > ** > 5221021f2f6e1e50cb6a953935c3601284925decd3fd21bc445712576873fb8c6ebc1821022 > df8750480ad5b26950b25c7ba79d3e37d75f640f8e5d9bcd5b150a0f85014da2103e3818b65 > bcc73a7d64064106a859cc1a5a728c4345ff0b641209fba0d90de6e953ae * Address > ** 3Q4sF6tv9wsdqu2NtARzNCpQgwifm2rAba > > ==Usage & Implementations== > * BIP45 - Structure for Deterministic P2SH Multisignature Wallets - > https://github.com/bitcoin/bips/blob/master/bip-0045.mediawiki#address-gene > ration-procedure * Bitcore - > https://github.com/bitpay/bitcore/blob/50a868cb8cdf2be04bb1c5bf4bcc064cc06f > 5888/lib/script/script.js#L541 * Haskoin - > https://github.com/haskoin/haskoin/blob/master/Network/Haskoin/Script/Parse > r.hs#L112-122 * Armory - > https://github.com/etotheipi/BitcoinArmory/blob/268db0f3fa20c989057bd43343a > 43b2edbe89aeb/armoryengine/ArmoryUtils.py#L1441 * Multisignature > Brainwallet - http://ms-brainwallet.org/ > > For now, the BIP will live here: > https://github.com/afk11/bips/blob/bip0090/bip-0090.mediawiki/ > > Looking forward to any feedback and discussions that arise!
On Thu, Feb 12, 2015 at 10:13:33PM +0000, Luke Dashjr wrote:
Quoted text (2 lines)
> Where is the Specification section?? Does this support arbitrary scripts, or > only the simplest CHECKMULTISIG case?
It might be enough to rewrite this BIP to basically say "all pubkeys executed by all CHECKMULTISIG opcodes will be in the following canonical order", followed by some explanatory examples of how to apply this simple rule.
OTOH we don't yet have a standard way of even talking about arbitrary scripts, so it may very well turn out to be the case that the above rule is too restrictive in many cases - I certainly would not want to do a soft-fork to enforce this, or even make it an IsStandard() rule.
The idea is more like BIP44/45 to have a 'standard' that software can comply by and express they do so that it makes a step towards compatibility between (wallet) software.
On Fri, Feb 13, 2015 at 8:53 AM, Peter Todd <pete@petertodd•org> wrote:
Quoted text (19 lines)
> On Thu, Feb 12, 2015 at 10:13:33PM +0000, Luke Dashjr wrote: > > Where is the Specification section?? Does this support arbitrary > scripts, or > > only the simplest CHECKMULTISIG case? > > It might be enough to rewrite this BIP to basically say "all pubkeys > executed by all CHECKMULTISIG opcodes will be in the following canonical > order", followed by some explanatory examples of how to apply this > simple rule. > > OTOH we don't yet have a standard way of even talking about arbitrary > scripts, so it may very well turn out to be the case that the above rule > is too restrictive in many cases - I certainly would not want to do a > soft-fork to enforce this, or even make it an IsStandard() rule. > > -- > 'peter'[:-1]@petertodd.org > 000000000000000013cf8270118ba2efce8b304f8de359599fef95c3ab43dcb1 >
On 12/02/15 22:13, Luke Dashjr wrote:
Quoted text (2 lines)
> Where is the Specification section?? Does this support arbitrary scripts, or > only the simplest CHECKMULTISIG case?
The BIP is a process for deriving only the type of scripts you would encounter doing addmultisigaddress. More complicated scripts would require more metadata to be shared, but the only case we describe is when given public keys and the number of signatures required.
You're right, we're missing a Specification. I have tweaked the document to cover this now.
On 13/02/15 07:53, Peter Todd wrote:
Quoted text (8 lines)
> It might be enough to rewrite this BIP to basically say "all pubkeys > executed by all CHECKMULTISIG opcodes will be in the following > canonical order", followed by some explanatory examples of how to > apply this simple rule. OTOH we don't yet have a standard way of even > talking about arbitrary scripts, so it may very well turn out to be > the case that the above rule is too restrictive in many cases - I > certainly would not want to do a soft-fork to enforce this, or even > make it an IsStandard() rule.
It would be interesting, but I agree it should not be brought into these validation rules - just a convention for people to follow for now. I think it's fair that implementers are free to order them however they please. But I think there is good reason for wallets to opt in to the convention and declare this, for ease of recovery, and for interoperability reasons.
I wonder are there any other blockers or modifications that need to be made for this to be merged?
Latest version of the document: https://github.com/afk11/bips/blob/213e8a27a3a2eaaf44f79221a9f9f888af002801/bip-0067.mediawiki
On 13/02/15 23:43, Thomas Kerin wrote:
Quoted text (3 lines)
> > On 12/02/15 22:13, Luke Dashjr wrote: >> Where is the Specification section?? Does this support arbitrary
scripts, or
Quoted text (3 lines)
>> only the simplest CHECKMULTISIG case? > > The BIP is a process for deriving only the type of scripts you would
encounter doing addmultisigaddress. More complicated scripts would require more metadata to be shared, but the only case we describe is when given public keys and the number of signatures required.
Quoted text (2 lines)
> > You're right, we're missing a Specification. I have tweaked the
document to cover this now.
Quoted text (5 lines)
> > > > On 13/02/15 07:53, Peter Todd wrote: >> It might be enough to rewrite this BIP to basically say "all pubkeys
executed by all CHECKMULTISIG opcodes will be in the following canonical order", followed by some explanatory examples of how to apply this simple rule. OTOH we don't yet have a standard way of even talking about arbitrary scripts, so it may very well turn out to be the case that the above rule is too restrictive in many cases - I certainly would not want to do a soft-fork to enforce this, or even make it an IsStandard() rule.
Quoted text (2 lines)
> > It would be interesting, but I agree it should not be brought into
these validation rules - just a convention for people to follow for now. I think it's fair that implementers are free to order them however they please. But I think there is good reason for wallets to opt in to the convention and declare this, for ease of recovery, and for interoperability reasons.
Quoted text (6 lines)
> > > -- > Thomas Kerin > ------------------------- > My PGP key can be found here
<http://pgp.mit.edu/pks/lookup?op=get&search=0x3F0D2F83A2966155>
Quoted text (3 lines)
> > >
------------------------------------------------------------------------------
Quoted text (2 lines)
> Dive into the World of Parallel Programming. The Go Parallel Website, > sponsored by Intel and developed in partnership with Slashdot Media,
is your
Quoted text (9 lines)
> hub for all things parallel software development, from weekly thought > leadership blogs to news, videos, case studies, tutorials and more. Take a > look and join the conversation now. http://goparallel.sourceforge.net/ > > > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development
- -- Thomas Kerin - -------------------------
My PGP key can be found here <http://pgp.mit.edu/pks/lookup?op=get&search=0x3F0D2F83A2966155>
A few months back, William Swanson and I had worked on a more general script template format. Unfortunately, other work has prevented us from being able to fully complete it - but here’s the start:
https://docs.google.com/document/d/1nGF6LjGwhzuiJ9AQwKAhN1a1SXvGGHWxoKmDSkiIsPI <https://docs.google.com/document/d/1nGF6LjGwhzuiJ9AQwKAhN1a1SXvGGHWxoKmDSkiIsPI>/
- Eric Lombrozo
Quoted text (28 lines)
> On Feb 12, 2015, at 11:53 PM, Peter Todd <pete@petertodd•org> wrote: > > On Thu, Feb 12, 2015 at 10:13:33PM +0000, Luke Dashjr wrote: >> Where is the Specification section?? Does this support arbitrary scripts, or >> only the simplest CHECKMULTISIG case? > > It might be enough to rewrite this BIP to basically say "all pubkeys > executed by all CHECKMULTISIG opcodes will be in the following canonical > order", followed by some explanatory examples of how to apply this > simple rule. > > OTOH we don't yet have a standard way of even talking about arbitrary > scripts, so it may very well turn out to be the case that the above rule > is too restrictive in many cases - I certainly would not want to do a > soft-fork to enforce this, or even make it an IsStandard() rule. > > -- > 'peter'[:-1]@petertodd.org > 000000000000000013cf8270118ba2efce8b304f8de359599fef95c3ab43dcb1 > ------------------------------------------------------------------------------ > Dive into the World of Parallel Programming. The Go Parallel Website, > sponsored by Intel and developed in partnership with Slashdot Media, is your > hub for all things parallel software development, from weekly thought > leadership blogs to news, videos, case studies, tutorials and more. Take a > look and join the conversation now. http://goparallel.sourceforge.net/_______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development