convention/standard for sorting public keys for p2sh multisig transactions
16 messages, 14 to 16 Jan 2015. Started by Ruben de Vries.
Eric Lombrozo (3), Jean-Pierre Rupp (3), Ruben de Vries (2), devrandom, Jeffrey Paul, Pavol Rusnak, Jeff Garzik, Matt Whitlock, +3 more
For p2sh multisig TXs the order of the public keys affect the hash and there doesn't seem to be an agreed upon way of sorting the public keys.
If there would be a standard (recommended) way of sorting the public keys that would make it easier for services that implement some form of multisig to be compatible with each other without much hassle and making it possible to import keys from one service to another.
I'm not suggesting forcing the order, just setting a standard to recommend, there doesn't seem to be much reason for (new) services to not follow that recommendation.
Ryan from BitPay broad this up before ( https://sourceforge.net/p/bitcoin/mailman/message/32092958/) and in bitcore they've implemented lexicographical sorting on the hex of the public key. In a short search I can't find any other library that has a sorting function, let alone using it by default, so bitcore is currently my only reference.
Ruben de Vries CTO, BlockTrail
At CryptoCorp we recommend to our customers that they sort lexicographically by the public key bytes of the leaf public keys. i.e. the same as BitPay.
On Wed, 2015-01-14 at 17:37 +0100, Ruben de Vries wrote:
Quoted text (36 lines)
> For p2sh multisig TXs the order of the public keys affect the hash and > there doesn't seem to be an agreed upon way of sorting the public > keys. > > > If there would be a standard (recommended) way of sorting the public > keys that would make it easier for services that implement some form > of multisig to be compatible with each other without much hassle and > making it possible to import keys from one service to another. > > > I'm not suggesting forcing the order, just setting a standard to > recommend, there doesn't seem to be much reason for (new) services to > not follow that recommendation. > > > Ryan from BitPay broad this up before > (https://sourceforge.net/p/bitcoin/mailman/message/32092958/) and in > bitcore they've implemented lexicographical sorting on the hex of the > public key. > In a short search I can't find any other library that has a sorting > function, let alone using it by default, so bitcore is currently my > only reference. > > > > > Ruben de Vries > CTO, BlockTrail > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ Bitcoin-development mailing list Bitcoin-development@lists•sourceforge•net https://lists.sourceforge.net/lists/listinfo/bitcoin-development
I think everyone is pretty much following this standard now.
- Eric On Jan 14, 2015 12:58 PM, "devrandom" <c1.sf-bitcoin@niftybox•net> wrote:
Quoted text (63 lines)
> At CryptoCorp we recommend to our customers that they sort > lexicographically by the public key bytes of the leaf public keys. i.e. > the same as BitPay. > > On Wed, 2015-01-14 at 17:37 +0100, Ruben de Vries wrote: > > For p2sh multisig TXs the order of the public keys affect the hash and > > there doesn't seem to be an agreed upon way of sorting the public > > keys. > > > > > > If there would be a standard (recommended) way of sorting the public > > keys that would make it easier for services that implement some form > > of multisig to be compatible with each other without much hassle and > > making it possible to import keys from one service to another. > > > > > > I'm not suggesting forcing the order, just setting a standard to > > recommend, there doesn't seem to be much reason for (new) services to > > not follow that recommendation. > > > > > > Ryan from BitPay broad this up before > > (https://sourceforge.net/p/bitcoin/mailman/message/32092958/) and in > > bitcore they've implemented lexicographical sorting on the hex of the > > public key. > > In a short search I can't find any other library that has a sorting > > function, let alone using it by default, so bitcore is currently my > > only reference. > > > > > > > > > > Ruben de Vries > > CTO, BlockTrail > > > ------------------------------------------------------------------------------ > > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > > GigeNET is offering a free month of service with a new server in Ashburn. > > Choose from 2 high performing configs, both with 100TB of bandwidth. > > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > > http://p.sf.net/sfu/gigenet > > _______________________________________________ Bitcoin-development > mailing list Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > > -- > Miron / devrandom > > > > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development >
We in Haskoin do the same.
On 14/01/15 17:39, devrandom wrote:
Quoted text (3 lines)
> At CryptoCorp we recommend to our customers that they sort > lexicographically by the public key bytes of the leaf public keys. i.e. > the same as BitPay.
Quoted text (5 lines)
> On 20150114, at 09:39, devrandom <c1.sf-bitcoin@niftybox•net> wrote: > > At CryptoCorp we recommend to our customers that they sort > lexicographically by the public key bytes of the leaf public keys. i.e. > the same as BitPay.
To clarify: the raw bytes of the public key itself, not the ascii base58 representation of the pubkey hash - right?
-jp
On 14/01/15 20:27, Jeffrey Paul wrote:
Quoted text (1 line)
> To clarify: the raw bytes of the public key itself, not the ascii base58 representation of the pubkey hash - right?
Could you give an example of two pubkeys where the following condition is met?
raw(pubkey1) > raw(pubkey2) and base58(pubkey1) < base58(pubkey2)
Sounds like this warrants a micro-BIP just to get everybody on the same page.
On Wed, Jan 14, 2015 at 11:37 AM, Ruben de Vries <ruben@blocktrail•com> wrote:
Quoted text (37 lines)
> For p2sh multisig TXs the order of the public keys affect the hash and > there doesn't seem to be an agreed upon way of sorting the public keys. > > If there would be a standard (recommended) way of sorting the public keys > that would make it easier for services that implement some form of multisig > to be compatible with each other without much hassle and making it possible > to import keys from one service to another. > > I'm not suggesting forcing the order, just setting a standard to > recommend, there doesn't seem to be much reason for (new) services to not > follow that recommendation. > > Ryan from BitPay broad this up before ( > https://sourceforge.net/p/bitcoin/mailman/message/32092958/) and in > bitcore they've implemented lexicographical sorting on the hex of the > public key. > In a short search I can't find any other library that has a sorting > function, let alone using it by default, so bitcore is currently my only > reference. > > > Ruben de Vries > CTO, BlockTrail > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > >
I would highly recommend NOT using Base58 for anything except stuff that is to be copy/pasted by the enduser.
Internally, pubkeys are DER-encoded integers.
- Eric On Jan 14, 2015 2:54 PM, "Jeffrey Paul" <jp@eeqj•com> wrote:
Quoted text (31 lines)
> > > On 20150114, at 09:39, devrandom <c1.sf-bitcoin@niftybox•net> wrote: > > > > At CryptoCorp we recommend to our customers that they sort > > lexicographically by the public key bytes of the leaf public keys. i.e. > > the same as BitPay. > > To clarify: the raw bytes of the public key itself, not the ascii base58 > representation of the pubkey hash - right? > > -jp > > -- > Jeffrey Paul EEQJ > jp@eeqj•com https://eeqj.com > +1-800-403-1126 (America) +1-312-361-0355 (Worldwide) > 5539 AD00 DE4C 42F3 AFE1 1575 0524 43F4 DF2A 55C2 > > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development >
Ciphrex was using this convention well before BitPay...and BitPay's BIP32 implementation was at least partly taken from ours.
- Eric On Jan 14, 2015 8:03 PM, "Andy Alness" <andy@coinbase•com> wrote:
Quoted text (70 lines)
> Doing same (BitPay convention) for our multisig support. > > On Wed, Jan 14, 2015 at 3:53 PM, Eric Lombrozo <elombrozo@gmail•com> > wrote: > > I would highly recommend NOT using Base58 for anything except stuff that > is > > to be copy/pasted by the enduser. > > > > Internally, pubkeys are DER-encoded integers. > > > > - Eric > > > > On Jan 14, 2015 2:54 PM, "Jeffrey Paul" <jp@eeqj•com> wrote: > >> > >> > >> > On 20150114, at 09:39, devrandom <c1.sf-bitcoin@niftybox•net> wrote: > >> > > >> > At CryptoCorp we recommend to our customers that they sort > >> > lexicographically by the public key bytes of the leaf public keys. > i.e. > >> > the same as BitPay. > >> > >> To clarify: the raw bytes of the public key itself, not the ascii base58 > >> representation of the pubkey hash - right? > >> > >> -jp > >> > >> -- > >> Jeffrey Paul EEQJ > >> jp@eeqj•com https://eeqj.com > >> +1-800-403-1126 (America) +1-312-361-0355 (Worldwide) > >> 5539 AD00 DE4C 42F3 AFE1 1575 0524 43F4 DF2A 55C2 > >> > >> > >> > >> > ------------------------------------------------------------------------------ > >> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > >> GigeNET is offering a free month of service with a new server in > Ashburn. > >> Choose from 2 high performing configs, both with 100TB of bandwidth. > >> Higher redundancy.Lower latency.Increased capacity.Completely compliant. > >> http://p.sf.net/sfu/gigenet > >> _______________________________________________ > >> Bitcoin-development mailing list > >> Bitcoin-development@lists•sourceforge•net > >> https://lists.sourceforge.net/lists/listinfo/bitcoin-development > > > > > > > ------------------------------------------------------------------------------ > > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > > GigeNET is offering a free month of service with a new server in Ashburn. > > Choose from 2 high performing configs, both with 100TB of bandwidth. > > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > > http://p.sf.net/sfu/gigenet > > _______________________________________________ > > Bitcoin-development mailing list > > Bitcoin-development@lists•sourceforge•net > > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > > > > > > -- > Andy Alness > Software Engineer > Coinbase > San Francisco, CA >
On Wednesday, 14 January 2015, at 3:53 pm, Eric Lombrozo wrote:
Quoted text (1 line)
> Internally, pubkeys are DER-encoded integers.
I thought pubkeys were represented as raw integers (i.e., they're embedded in Script as a push operation whose payload is the raw bytes of the big-endian representation of the integer). As far as I know, DER encoding is only used for signatures. Am I mistaken?
In BIP45 it mentions "lexicographically sorting the public keys".
https://github.com/bitcoin/bips/blob/master/bip-0045.mediawiki#Address_Generation_Procedure
On 15 January 2015 at 03:32, Jeff Garzik <jgarzik@bitpay•com> wrote:
Quoted text (64 lines)
> Sounds like this warrants a micro-BIP just to get everybody on the same > page. > > > On Wed, Jan 14, 2015 at 11:37 AM, Ruben de Vries <ruben@blocktrail•com> > wrote: > >> For p2sh multisig TXs the order of the public keys affect the hash and >> there doesn't seem to be an agreed upon way of sorting the public keys. >> >> If there would be a standard (recommended) way of sorting the public keys >> that would make it easier for services that implement some form of multisig >> to be compatible with each other without much hassle and making it possible >> to import keys from one service to another. >> >> I'm not suggesting forcing the order, just setting a standard to >> recommend, there doesn't seem to be much reason for (new) services to not >> follow that recommendation. >> >> Ryan from BitPay broad this up before ( >> https://sourceforge.net/p/bitcoin/mailman/message/32092958/) and in >> bitcore they've implemented lexicographical sorting on the hex of the >> public key. >> In a short search I can't find any other library that has a sorting >> function, let alone using it by default, so bitcore is currently my only >> reference. >> >> >> Ruben de Vries >> CTO, BlockTrail >> >> >> ------------------------------------------------------------------------------ >> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >> GigeNET is offering a free month of service with a new server in Ashburn. >> Choose from 2 high performing configs, both with 100TB of bandwidth. >> Higher redundancy.Lower latency.Increased capacity.Completely compliant. >> http://p.sf.net/sfu/gigenet >> _______________________________________________ >> Bitcoin-development mailing list >> Bitcoin-development@lists•sourceforge•net >> https://lists.sourceforge.net/lists/listinfo/bitcoin-development >> >> > > > -- > Jeff Garzik > Bitcoin core developer and open source evangelist > BitPay, Inc. https://bitpay.com/ > > > ------------------------------------------------------------------------------ > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > >
A public key is a point in the elliptic curve. As such it has an X and a Y component. Its serialization is described very succintly here:
https://en.bitcoin.it/wiki/Protocol_specification#Signatures
On 15/01/15 01:17, Matt Whitlock wrote:
Quoted text (1 line)
> I thought pubkeys were represented as raw integers (i.e., they're embedded in Script as a push operation whose payload is the raw bytes of the big-endian representation of the integer). As far as I know, DER encoding is only used for signatures. Am I mistaken?
Since we only need the sorting for creating the scriptPubKey, wouldn't it make the most sense to sort it by the way it represented in that context?
On Thu, Jan 15, 2015 at 2:03 PM, Wladimir <laanwj@gmail•com> wrote:
Quoted text (23 lines)
> On Thu, Jan 15, 2015 at 1:17 AM, Matt Whitlock <bip@mattwhitlock•name> > wrote: > > On Wednesday, 14 January 2015, at 3:53 pm, Eric Lombrozo wrote: > >> Internally, pubkeys are DER-encoded integers. > > > > I thought pubkeys were represented as raw integers (i.e., they're > embedded in Script as a push operation whose payload is the raw bytes of > the big-endian representation of the integer). As far as I know, DER > encoding is only used for signatures. Am I mistaken? > > OP_CHECKSIG (and OP_CHECKSIGVERIFY) takes a DER-encoded pubkey and a > DER-encoded signature on the stack. > > Possibly you're confused with OP_HASH160 <hash160> OP_EQUALVERIFY as > used in outputs, which compares the 160-bit hash of the pubkey against > the given hash (usually taken from a bitcoin address). > > It doesn't help understanding to consider either as integers. They are > binary blob objects with either a fixed format (DER) or a fixed size > (hashes). > > Wladimir >
It would - it assumes you have the set of keys and are sorting before you derive and send funds to such a P2SH address.
It seems there is scope for further narrowing down how a multisig scripthash address should be determined - what do people think of anticipating only compressed keys for scripts?
It's possible to cause confusion if one put forward a compressed key at some time, and an uncompressed key at another. A different script hash would be produced even though there is no difference to the keys involved. The client will not search for this.
Having spoken with Jean-Pierre and Ruben about this for quite some time now, there is 100% the need for a BIP outlining this. Everyone has had the idea at some point, and some of us already using it, but people shouldn't have to go digging in BIP45 for the two lines which mention it. All we need is a place to put the docs.
I am building up a list of implementations which currently support sorting, and briefly describing a motivation for such a BIP.
On 16/01/15 10:16, Ruben de Vries wrote:
Quoted text (2 lines)
> Since we only need the sorting for creating the scriptPubKey, > wouldn't it make the most sense to sort it by the way it represented
in that context?
Quoted text (3 lines)
> > > On Thu, Jan 15, 2015 at 2:03 PM, Wladimir <laanwj@gmail•com
<mailto:laanwj@gmail.com>> wrote:
Quoted text (2 lines)
> > On Thu, Jan 15, 2015 at 1:17 AM, Matt Whitlock
<bip@mattwhitlock•name <mailto:bip@mattwhitlock.name>> wrote:
Quoted text (4 lines)
> > On Wednesday, 14 January 2015, at 3:53 pm, Eric Lombrozo wrote: > >> Internally, pubkeys are DER-encoded integers. > > > > I thought pubkeys were represented as raw integers (i.e.,
they're embedded in Script as a push operation whose payload is the raw bytes of the big-endian representation of the integer). As far as I know, DER encoding is only used for signatures. Am I mistaken?
Quoted text (30 lines)
> > OP_CHECKSIG (and OP_CHECKSIGVERIFY) takes a DER-encoded pubkey and a > DER-encoded signature on the stack. > > Possibly you're confused with OP_HASH160 <hash160> OP_EQUALVERIFY as > used in outputs, which compares the 160-bit hash of the pubkey against > the given hash (usually taken from a bitcoin address). > > It doesn't help understanding to consider either as integers. They are > binary blob objects with either a fixed format (DER) or a fixed size > (hashes). > > Wladimir > > > > > -- > BlockTrail B.V. > Barbara Strozzilaan 201 > 1083HN Amsterdam > The Netherlands > > Phone:+31 (0)612227277 > E-mail:ruben@blocktrail.com <mailto:ruben@blocktrail.com> > Web:www.blocktrail.com > <http://www.blocktrail.com/> > Github:www.github.com/rubensayshi <http://www.github.com/rubensayshi> > > BlockTrail B.V. Is registered with the Dutch Chamber of Commerce in
Amsterdam with registration No.:60262060 and VAT No.:NL853833035B01
Quoted text (3 lines)
> > >
------------------------------------------------------------------------------
Quoted text (11 lines)
> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > > > _______________________________________________ > 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>
I see no reason to restrict compressed/uncompressed. Strings don't have to be the same length to sort them lexicographically. If a multi-sig participant provides an uncompressed key, they are declaring that the key that they use and it will only be used uncompressed. Clients don't have to go looking for all combinations of compressed & uncompressed.
On 01/16/2015 11:34 AM, Thomas Kerin wrote:
Quoted text (94 lines)
> > > It seems there is scope for further narrowing down how a multisig > scripthash address should be determined - what do people think of > anticipating only compressed keys for scripts? > > It's possible to cause confusion if one put forward a compressed key > at some time, and an uncompressed key at another. A different script > hash would be produced even though there is no difference to the keys > involved. The client will not search for this. > > > Having spoken with Jean-Pierre and Ruben about this for quite some > time now, there is 100% the need for a BIP outlining this. Everyone > has had the idea at some point, and some of us already using it, but > people shouldn't have to go digging in BIP45 for the two lines which > mention it. All we need is a place to put the docs. > > I am building up a list of implementations which currently support > sorting, and briefly describing a motivation for such a BIP. > > > On 16/01/15 10:16, Ruben de Vries wrote: > > Since we only need the sorting for creating the scriptPubKey, > > wouldn't it make the most sense to sort it by the way it represented > in that context? > > > > On Thu, Jan 15, 2015 at 2:03 PM, Wladimir <laanwj@gmail•com > <mailto:laanwj@gmail.com>> wrote: > > > On Thu, Jan 15, 2015 at 1:17 AM, Matt Whitlock > <bip@mattwhitlock•name <mailto:bip@mattwhitlock.name>> wrote: > > > On Wednesday, 14 January 2015, at 3:53 pm, Eric Lombrozo wrote: > > >> Internally, pubkeys are DER-encoded integers. > > > > > > I thought pubkeys were represented as raw integers (i.e., > they're embedded in Script as a push operation whose payload is the > raw bytes of the big-endian representation of the integer). As far as > I know, DER encoding is only used for signatures. Am I mistaken? > > > OP_CHECKSIG (and OP_CHECKSIGVERIFY) takes a DER-encoded pubkey and a > > DER-encoded signature on the stack. > > > Possibly you're confused with OP_HASH160 <hash160> OP_EQUALVERIFY as > > used in outputs, which compares the 160-bit hash of the pubkey > against > > the given hash (usually taken from a bitcoin address). > > > It doesn't help understanding to consider either as integers. > They are > > binary blob objects with either a fixed format (DER) or a fixed size > > (hashes). > > > Wladimir > > > > > > -- > > BlockTrail B.V. > > Barbara Strozzilaan 201 > > 1083HN Amsterdam > > The Netherlands > > > Phone:+31 (0)612227277 > > E-mail:ruben@blocktrail.com <mailto:ruben@blocktrail.com> > > Web:www.blocktrail.com > > <http://www.blocktrail.com/> > > Github:www.github.com/rubensayshi <http://www.github.com/rubensayshi> > > > BlockTrail B.V. Is registered with the Dutch Chamber of Commerce in > Amsterdam with registration No.:60262060 and VAT No.:NL853833035B01 > > > > > ------------------------------------------------------------------------------ > > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > > GigeNET is offering a free month of service with a new server in > Ashburn. > > Choose from 2 high performing configs, both with 100TB of bandwidth. > > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > > http://p.sf.net/sfu/gigenet > > > > _______________________________________________ > > Bitcoin-development mailing list > > Bitcoin-development@lists•sourceforge•net > > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > > > > >
------------------------------------------------------------------------------
Quoted text (11 lines)
> New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > GigeNET is offering a free month of service with a new server in Ashburn. > Choose from 2 high performing configs, both with 100TB of bandwidth. > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > http://p.sf.net/sfu/gigenet > > > _______________________________________________ > Bitcoin-development mailing list > Bitcoin-development@lists•sourceforge•net > https://lists.sourceforge.net/lists/listinfo/bitcoin-development
It is better if the scheme is strongly deterministic.On 16 Jan 2015 17:09, Alan Reiner <etotheipi@gmail•com> wrote:
Quoted text (91 lines)
> > I see no reason to restrict compressed/uncompressed. Strings don't have to be the same length to sort them lexicographically. If a multi-sig participant provides an uncompressed key, they are declaring that the key that they use and it will only be used uncompressed. Clients don't have to go looking for all combinations of compressed & uncompressed. > > On 01/16/2015 11:34 AM, Thomas Kerin wrote: > > >> >> >> It seems there is scope for further narrowing down how a multisig scripthash address should be determined - what do people think of anticipating only compressed keys for scripts? >> >> It's possible to cause confusion if one put forward a compressed key at some time, and an uncompressed key at another. A different script hash would be produced even though there is no difference to the keys involved. The client will not search for this. >> >> >> Having spoken with Jean-Pierre and Ruben about this for quite some time now, there is 100% the need for a BIP outlining this. Everyone has had the idea at some point, and some of us already using it, but people shouldn't have to go digging in BIP45 for the two lines which mention it. All we need is a place to put the docs. >> >> I am building up a list of implementations which currently support sorting, and briefly describing a motivation for such a BIP. >> >> >> On 16/01/15 10:16, Ruben de Vries wrote: >> > Since we only need the sorting for creating the scriptPubKey, >> > wouldn't it make the most sense to sort it by the way it represented in that context? >> >> >> > On Thu, Jan 15, 2015 at 2:03 PM, Wladimir <laanwj@gmail•com <mailto:laanwj@gmail.com>> wrote: >> >> > On Thu, Jan 15, 2015 at 1:17 AM, Matt Whitlock <bip@mattwhitlock•name <mailto:bip@mattwhitlock.name>> wrote: >> > > On Wednesday, 14 January 2015, at 3:53 pm, Eric Lombrozo wrote: >> > >> Internally, pubkeys are DER-encoded integers. >> > > >> > > I thought pubkeys were represented as raw integers (i.e., they're embedded in Script as a push operation whose payload is the raw bytes of the big-endian representation of the integer). As far as I know, DER encoding is only used for signatures. Am I mistaken? >> >> > OP_CHECKSIG (and OP_CHECKSIGVERIFY) takes a DER-encoded pubkey and a >> > DER-encoded signature on the stack. >> >> > Possibly you're confused with OP_HASH160 <hash160> OP_EQUALVERIFY as >> > used in outputs, which compares the 160-bit hash of the pubkey against >> > the given hash (usually taken from a bitcoin address). >> >> > It doesn't help understanding to consider either as integers. They are >> > binary blob objects with either a fixed format (DER) or a fixed size >> > (hashes). >> >> > Wladimir >> >> >> >> >> > -- >> > BlockTrail B.V. >> > Barbara Strozzilaan 201 >> > 1083HN Amsterdam >> > The Netherlands >> >> > Phone:+31 (0)612227277 >> > E-mail:ruben@blocktrail.com <mailto:ruben@blocktrail.com> >> > Web:www.blocktrail.com >> > <http://www.blocktrail.com/> >> > Github:www.github.com/rubensayshi <http://www.github.com/rubensayshi> >> >> > BlockTrail B.V. Is registered with the Dutch Chamber of Commerce in Amsterdam with registration No.:60262060 and VAT No.:NL853833035B01 >> >> >> > ------------------------------------------------------------------------------ >> > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. >> > GigeNET is offering a free month of service with a new server in Ashburn. >> > Choose from 2 high performing configs, both with 100TB of bandwidth. >> > Higher redundancy.Lower latency.Increased capacity.Completely compliant. >> > http://p.sf.net/sfu/gigenet >> >> >> > _______________________________________________ >> > Bitcoin-development mailing list >> > Bitcoin-development@lists•sourceforge•net >> > https://lists.sourceforge.net/lists/listinfo/bitcoin-development >> > > > > > > > > ------------------------------------------------------------------------------ > > New Year. New Location. New Benefits. New Data Center in Ashburn, VA. > > GigeNET is offering a free month of service with a new server in Ashburn. > > Choose from 2 high performing configs, both with 100TB of bandwidth. > > Higher redundancy.Lower latency.Increased capacity.Completely compliant. > > http://p.sf.net/sfu/gigenet > > > > > > _______________________________________________ > > Bitcoin-development mailing list > > Bitcoin-development@lists•sourceforge•net > > https://lists.sourceforge.net/lists/listinfo/bitcoin-development > >