[coinb.in] Open Source, Multi Signature, HD Wallet, SegWit/Bech32 and more! - page 18.

RagnarDanneskjold

full member

Activity: 144

Merit: 100

Quote from: em on February 08, 2014, 06:43:05 AM

Quote from: OutCast3k on February 07, 2014, 08:06:21 AM

You've mentioned that the sellers maybe put at a disadvantage here because the buyer may not sign the transaction, well, there are many tried and tested sites where this type of system works. Such as; localbitcoins, bitmit, coingig or even all those .onion sites - but these sites require you to fully trust them, as they expect the coins to be deposited in their wallet and the buyer presses a "release button", as opposed to a multisig address, where no single person can steal them or claim they've been hacked, etc etc..

Exactly: here we are trying to avoid the need for trusted third parties policing the transactions. And often they rely on a separate trust metrics for assessing the participants: for example, Bitcoin.de has a system of mutual ratings among users based on the number of successful settlements, and each user's rating is visible to all other users. But this requires a centralized database and someone to maintain it, and we don't want that.

Quote from: OutCast3k on February 07, 2014, 08:06:21 AM

Anyway, unless I've miss understood your suggestion, it is not possible to do as payments can not be "cancelled".

However (please correct me if I'm wrong), I guess it's possible to make a transaction that will pay part of the amount sent to the script hash to one address (the Seller) and part to another (the Buyer). If the Buyer reneged on his agreement and refused to sign and broadcast the transaction after receiving the goods, he would have to forfeit the bond.

there are a number of ways you could arrange a multisig contract to address what you call bond forfeiture. Just to be sure, in its basic form, if the multisig address requires more signatures than are available to sign the tx, then funds will remain locked, yes. So there is extra burden on all parties to be sure keys are safe & available on time.... or, depending on the details & complexity of agreement - additional signer keys could be used, shared, split or other tricks to build in failsafes.

here is an example of an exchange implementation using Outkast's multisig [thanks again - much fun]
http://isx.io

your question 'is it possible to implement this protocol with p2sh primitives?" - p2sh could be html javascripted just like this one if thats what you mean.

See also:
https://github.com/bitcoin/bips/blob/master/bip-0016.mediawiki
https://github.com/bitcoin/bips/blob/master/bip-0017.mediawiki
https://github.com/bitcoin/bips/blob/master/bip-0012.mediawiki
http://bitcoin.stackexchange.com/questions/2749/what-is-an-explanation-of-the-p2sh-voting-in-laymans-terms

em

newbie

Activity: 4

Merit: 0

Quote from: OutCast3k on February 07, 2014, 08:06:21 AM

You've mentioned that the sellers maybe put at a disadvantage here because the buyer may not sign the transaction, well, there are many tried and tested sites where this type of system works. Such as; localbitcoins, bitmit, coingig or even all those .onion sites - but these sites require you to fully trust them, as they expect the coins to be deposited in their wallet and the buyer presses a "release button", as opposed to a multisig address, where no single person can steal them or claim they've been hacked, etc etc..

Exactly: here we are trying to avoid the need for trusted third parties policing the transactions. And often they rely on a separate trust metrics for assessing the participants: for example, Bitcoin.de has a system of mutual ratings among users based on the number of successful settlements, and each user's rating is visible to all other users. But this requires a centralized database and someone to maintain it, and we don't want that.

Quote from: OutCast3k on February 07, 2014, 08:06:21 AM

Anyway, unless I've miss understood your suggestion, it is not possible to do as payments can not be "cancelled".

However (please correct me if I'm wrong), I guess it's possible to make a transaction that will pay part of the amount sent to the script hash to one address (the Seller) and part to another (the Buyer). If the Buyer reneged on his agreement and refused to sign and broadcast the transaction after receiving the goods, he would have to forfeit the bond.

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: HostFat on February 06, 2014, 11:14:27 AM

I would be cool if someone could make an easy android app to arrange and start a poker match.
So everyone will be able to play poker everywhere with just their smartphone and empty pockets

Nice idea... final project could be pretty awesome.

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: em on February 07, 2014, 07:02:46 AM

Maybe I'm missing something, but I think that with this method the Seller is at a disadvantage here, because the Buyer might renege on his promise and not sign the transaction after receiving the goods (e.g., unless the Seller ship an extra amount of goods). True, a mediator could break the tie if two signatures over three were sufficient to release the payment, but it would be much better if recourse to trusted third parties could be minimized.
A scheme that might work better would run as follows:

- After reaching a verbal agreement, Buyer issues a "Conditional payment" for Seller (which at this stage s/he may cancel at any time)
- Seller accepts the payment by "posting a bond" of, say, X% of the amount. The payment's status becomes "Committed", and Buyer can't cancel it anymore.
- Seller ships the goods
- Buyer receives the goods, and "releases the payment": the initial amount is paid to Seller, and at the same time the bond is paid back to Buyer.

Is it possible to implement this protocol with p2sh primitives?

Hey,

Thanks for the feedback.

You've mentioned that the sellers maybe put at a disadvantage here because the buyer may not sign the transaction, well, there are many tried and tested sites where this type of system works. Such as; localbitcoins, bitmit, coingig or even all those .onion sites and the mediators do just fine - but these sites require you to fully trust them, as they expect the coins to be deposited in their wallet and the buyer presses a "release button", as opposed to a multisig address, where no single person can steal them or claim they've been hacked, etc etc..

I also suspect many highly trusted sellers would just accept payment directly anyway.

Anyway, unless I've miss understood your suggestion, it is not possible to do as payments can not be "cancelled".

All the best

em

newbie

Activity: 4

Merit: 0

Maybe I'm missing something, but I think that with this method the Seller is at a disadvantage here, because the Buyer might renege on his promise and not sign the transaction after receiving the goods (e.g., unless the Seller ship an extra amount of goods). True, a mediator could break the tie if two signatures over three were sufficient to release the payment, but it would be much better if recourse to trusted third parties could be minimized.
A scheme that might work better would run as follows:

- After reaching a verbal agreement, Buyer issues a "Conditional payment" for Seller (which at this stage Buyer may cancel at any time)
- Seller accepts the payment by "posting a bond" of, say, X% of the amount. The payment's status becomes "Committed", and Buyer can't cancel it anymore.
- Seller ships the goods
- Buyer receives the goods, and "releases the payment": the initial amount is paid to Seller, and at the same time the bond is paid back to Buyer.

Is it possible to implement this protocol with p2sh primitives?

HostFat

staff

Activity: 4270

Merit: 1209

I support freedom of choice

I would be cool if someone could make an easy android app to arrange and start a poker match.
So everyone will be able to play poker everywhere with just their smartphone and empty pockets

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: vm_mpn on February 02, 2014, 01:22:28 PM

I just have to say - It's amazing what you guys are doing...

Thank You!

Thanks

Quote from: RagnarDanneskjold on January 31, 2014, 04:43:27 AM

Quote from: fbueller on January 29, 2014, 09:15:54 PM

This is an awesome project, very glad I saw this thread.

+1 million!

Great stuff outkast - i will use this EVERY day from now on

Thanks

vm_mpn

hero member

Activity: 605

Merit: 500

I just have to say - It's amazing what you guys are doing...

Thank You!

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: fbueller on January 31, 2014, 10:23:01 AM

Just ran through a test transaction, this works great! Kudos again. My shitty laptop apparently can't sign transactions, the script crashes, but I've tested it successfully on another machine, works a treat. 3cac39fc4aefc65aaa5419a60ab52aebc01d3c55f178a8ebe274650916cf38d6

Good to hear it worked.

fbueller

sr. member

Activity: 412

Merit: 287

Just ran through a test transaction, this works great! Kudos again. My shitty laptop apparently can't sign transactions, the script crashes, but I've tested it successfully on another machine, works a treat. 3cac39fc4aefc65aaa5419a60ab52aebc01d3c55f178a8ebe274650916cf38d6

RagnarDanneskjold

full member

Activity: 144

Merit: 100

Quote from: fbueller on January 29, 2014, 09:15:54 PM

This is an awesome project, very glad I saw this thread.

+1 million!

Great stuff outkast - i will use this EVERY day from now on

fbueller

sr. member

Activity: 412

Merit: 287

This is an awesome project, very glad I saw this thread. I'm working on implementing multisig into Bitwasp, and currently support for multisig sucks in the clients. I was hoping someone would come up with a stateless signing tool. Looking forward to your marketplace, I haven't seen another open source one before

ktorn

newbie

Activity: 32

Merit: 0

Quote from: OutCast3k on January 28, 2014, 01:38:40 PM

Fixed

you can now send back to a multisig address.

I've tested and it works for me as well: https://blockchain.info/tx/13326d4e0aa8ed620d4d28d26810f4ac9ce75e508d84b218f281c681389abdbd

Great work OutCast3k! Cheesy

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: ktorn on January 27, 2014, 01:34:46 AM

Quote from: OutCast3k on January 26, 2014, 06:30:21 AM

@ktorn I've been looking into it, its a little trickier than I first thought but should have it resolved over the next couple of days.

No worries, take your time.

Fixed

you can now send back to a multisig address.

One of my test cases http://blockchain.info/tx/832679f95685a8178bc7128dd3450fc29d0552a3ee6bcc3f1254f48d0ba4e0c4

ktorn

newbie

Activity: 32

Merit: 0

Quote from: OutCast3k on January 26, 2014, 06:30:21 AM

@ktorn I've been looking into it, its a little trickier than I first thought but should have it resolved over the next couple of days.

No worries, take your time.

ABISprotocol

sr. member

Activity: 278

Merit: 252

ABISprotocol on Gist

OutCast3k:

thank you, you are a good entity

much thanks

very excellent

etc

OutCast3k

hero member

Activity: 714

Merit: 601

Just a small update.

If anyone is running my script offline, I've fixed a lot of bugs so i'd suggest getting an update from github.

Quote from: btcmsia on January 24, 2014, 12:07:04 PM

latest bitcoin-qt client seems only generate address from compressed public key.
so all my addresses are associated with compressed publickey (prefix with 02/03, not 04).
and when compressed publickeys are used, their private keys also in different format, (prfix with K/L instead of 5)

@btcmsia I will look into this, but will have to get back to you before I promise anything. You can always generate a pubkey from the new keys tab, https://coinb.in/multisig/#newKeys You could also use bitaddress.org or brainwallet.org to generate keys pretty quickly.

Quote from: ktorn on January 24, 2014, 01:58:53 AM

Now that P2SH output support that we talked about earlier would be the icing on the cake Wink

Should I nudge the bitcoinjs folks about the issue or do you want to have a go at it?

@ktorn I've been looking into it, its a little trickier than I first thought but should have it resolved over the next couple of days.

Many thanks to everyone who has been testing Cheesy

btcmsia

newbie

Activity: 2

Merit: 0

Quote from: OutCast3k on January 22, 2014, 05:06:02 AM

Unfortunately, you can't use an address in replace of a pubkey.

An address is a one way hash of a pubkey. Its not possible to use an address because it cant be used to generate a pubkey, and its the pubkeys (which is included in the redeem scripts) that are needed to validate the signatures in a transaction when you release/send the coins from your mutlsig address.

I'm talking about compressed public key, not address.

latest bitcoin-qt client seems only generate address from compressed public key.
so all my addresses are associated with compressed publickey (prefix with 02/03, not 04).
and when compressed publickeys are used, their private keys also in different format, (prfix with K/L instead of 5)

there will be a compatible issue if your apps only support uncompressed pkey.

ktorn

newbie

Activity: 32

Merit: 0

Quote from: OutCast3k on January 24, 2014, 03:37:48 AM

I'll look into supporting the ability to send back to a multisig address later today, it shouldn't be that big of a deal. I don't think you'll get a response from the bitcoinjs guys.

Yeah, I've noticed that their github repo hasn't be touched in a while. Ping me when you patch the code and I'll give it a test on this side.

OutCast3k

hero member

Activity: 714

Merit: 601

Quote from: ktorn on January 24, 2014, 01:58:53 AM

Quote from: OutCast3k on January 23, 2014, 07:41:16 PM

This issue has now been fixed

the order that signatures are added no longer matters.

I've just tested it twice, switching the order of the signing keys between tests, and both transactions were successful.
Your fix works!

Now that P2SH output support that we talked about earlier would be the icing on the cake Wink

Should I nudge the bitcoinjs folks about the issue or do you want to have a go at it?

I'll look into supporting the ability to send back to a multisig address later today, it shouldn't be that big of a deal. I don't think you'll get a response from the bitcoinjs guys.

Topic: [coinb.in] Open Source, Multi Signature, HD Wallet, SegWit/Bech32 and more! - page 18. (Read 74845 times)