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

[americandesi]

Quick question.! Do I have to run my own bitcoind client on the server to be able to run this script.?

[OutCast3k]

Quick question.! Do I have to run my own bitcoind client on the server to be able to run this script.?

No, its not required.

[michaelpaulson]

good job with this. it's very useful. i'll be sending a donation your way shortly.

is there anyway we can modify the script to bypass the requirement of being online for the transaction creation (i believe it looks up your balance)? I would like to use this offline.

Thanks!

[fbueller]

I don't mean to nag, but is there any sign of upcoming support for compressed keys? I think it's pretty vital, as when any party to a multisig address uses a key from bitcoind, or electrum from the next release on, will encounter the issue I had with the scriptSig reconstruction (leaving out all signatures where the key was compressed).

[imicaihi]

Great work on this!

I've tried the service and a 2of3 worked great.
however, I tried something that I actually need right now in the form of a 6-of-10 and... no go
Getting a tx rejected -22 and log error says "nonstandard transaction: scriptsig-size" which I assume relates to the 500b limit mentioned earlier in the thread.
Is there a way to make this work?
Reading around in some other discussions I saw being mentioned that even a 15 of 15 should work, so I'm a bit confused as to why 6 of 10 doesn't.

Also OutCast, on a side note, it would be really great if you could provide the non-minified bitcoinjs code (either as a download here or in the git repo) as part of the opensource approach.


Thanks!

[imicaihi]

as a followup - I just tried doing a manual multisig transaction of a 6 of 10, using only compressed pub keys in the latest bitcoin core, but still getting tx rejected 22. the script is somewhat shorter than with the non-compressed, but definitely not by a huge amount.

Any ideas on how to get this to work?
I've seen people saying that with compressed pub keys even a 15 of 15 could fit within the 520b limit, so what am I missing here?

[waxwing]

as a followup - I just tried doing a manual multisig transaction of a 6 of 10, using only compressed pub keys in the latest bitcoin core, but still getting tx rejected 22. the script is somewhat shorter than with the non-compressed, but definitely not by a huge amount.

Any ideas on how to get this to work?
I've seen people saying that with compressed pub keys even a 15 of 15 could fit within the 520b limit, so what am I missing here?

Hi, I'm guessing you're the same guy as I spoke with on reddit? the answer is still: use Eligius. 6 of 10 is not standard (exceeds 500 limit). Was there some aspect of that explanation that didn't make sense? (Unless it wasn't you, in which case, sorry).

[jobijo]

Hi

This is very cool.
However, I don't understand everything.

Can someone make a complete tutorial for the average user to make simple 2of3 transactions?

Some questions:
- what is a redeem script?
- what should we put in that field?
- can't there be some simplifications like predifined 2of3 addresses and transactions?
- in  the transaction, what should I use as inputs?

Sorry for so many questions.
It would be so great if it could be usable by average users!

[jobijo]

I finally could test succesfuly with real money.
answering my own questions: the redeem script is given when you create the multisig address.
It's quite simple!!

Have made a tutorial in french.
Will translate it in english and post here.

[jobijo]

A few remarks:

There seem to be a bug: in both "create transaction" and "verify transaction", the input transaction is unknown to blockchain. Is it normal?

Also, in the signing page, could there be a "verify" button or an automatic verify step before the transaction is signed?

I use chome in linux.
In create new transaction, when you paste the redeem script with the middle button, the script is not activated. When you add a space or any caracter (or remove a caracter), the address goes wrong.
There is no check on the redeem script integrity. This can be dangerous I think.
There should be a verify step.

In the verify tab, could you add a "sign" button for transactions, or a "create transaction" for a redeem script? This would simplify the whole process.

Thanks for reading

[lucserre]

My 3 of 5 is being rejected... The "verify script" shows all in order. Blockchain.info pushtx:

Script not of right size, expecting 2 but got 5

I couldn't use Blockchain.info's pushtx for this. *shrugs*

I just used the one on OPs website.

Coinb.in is also failing... I also just tried a 3 of 4...

3 of 4 worked fine for me... note that if you try to resubmit the same transaction it will say your txn is invalid.

My 3 of 4 just got confirmations. 3 of 5 still being rejected by both coinb.in and blockchain.info.

It might be over 500 bytes... in which case (I'm told) it won't get confirmed by anyone.

The same thing is happening to me -   the 3-of-5 isn't working.  I had a 2-of-3 work just fine this morning, but this afternoon, there is no way to get the 3-of-5 to work.  Is coinbin not coding the transactions correctly? (Tried with blockchain.info to push too)
Is my money lost for ever??? (  Thanks coinb.in...)  I hope that is not the case...
Thanks,
-Luc

[OutCast3k]

The same thing is happening to me -   the 3-of-5 isn't working.  I had a 2-of-3 work just fine this morning, but this afternoon, there is no way to get the 3-of-5 to work.  Is coinbin not coding the transactions correctly? (Tried with blockchain.info to push too)
Is my money lost for ever??? (   Thanks coinb.in...)  I hope that is not the case...
Thanks,
-Luc

The hex encoded transaction you have given me via email appears to use non existent transaction IDs, as I've explained back to you. I don't think your money is lost, but if you could try again and note down each step you take, and the output and then contact me if it fails.

I finally could test succesfuly with real money.
answering my own questions: the redeem script is given when you create the multisig address.
It's quite simple!!

Have made a tutorial in french.
Will translate it in english and post here.

Thanks, glad you liked it.

Great work on this!

I've tried the service and a 2of3 worked great.
however, I tried something that I actually need right now in the form of a 6-of-10 and... no go
Getting a tx rejected -22 and log error says "nonstandard transaction: scriptsig-size" which I assume relates to the 500b limit mentioned earlier in the thread.
Is there a way to make this work?
Reading around in some other discussions I saw being mentioned that even a 15 of 15 should work, so I'm a bit confused as to why 6 of 10 doesn't.

Also OutCast, on a side note, it would be really great if you could provide the non-minified bitcoinjs code (either as a download here or in the git repo) as part of the opensource approach.


Thanks!

Can you drop me a email with the details, and i'll take a look. I'm not sure right now.

[dexX7]

Could it be possible that only non compressed public keys are allowed to be added during the multisig creation?

[lucserre]

The same thing is happening to me -   the 3-of-5 isn't working.  I had a 2-of-3 work just fine this morning, but this afternoon, there is no way to get the 3-of-5 to work.  Is coinbin not coding the transactions correctly? (Tried with blockchain.info to push too)
Is my money lost for ever??? (   Thanks coinb.in...)  I hope that is not the case...
Thanks,
-Luc

The hex encoded transaction you have given me via email appears to use non existent transaction IDs, as I've explained back to you. I don't think your money is lost, but if you could try again and note down each step you take, and the output and then contact me if it fails.

Thanks for reply. I think I know why.  When I put my RedeemScript in your area to construct the transaction, it shows me the three inputs, but the numbers are mixed up.
For example, the first one you listed from the decode transaction section is: 3fbd...120a   but is listed like this in the construct transaction section: 0a12...bd3f.
It looks like the 8 bit groupings are grouped together, but in some sort of reverse order (0a is at the start in one, and at the end in the other).
Correct me if I'm wrong, but it looks like (to me) there is an error somewhere in your code that flipped the txid's. ?
Is this a little-endian/big-endian thing?
I'm not sure why it would have worked for 2-of-3 but now for 3-of-5 it's being encoded differently?
-Luc

[queptiv]

summary of how this works?

[devthedev]

summary of how this works?

The page uses javascript to generate a multi signature address, the buyer sends the funds to the address. Once the agreement between all parties has completed a transaction can be created using the redeem script which is then signed by the participants. You can then broadcast the signed transaction into the network, releasing the funds to the chosen address.
Posted from Bitcointa.lk - #puVNlkBcvhsA7bDq

[fbueller]

Could it be possible that only non compressed public keys are allowed to be added during the multisig creation?

Yes. It will also 'forget' signatures added by compressed public keys. Makes it useless when everyone uses compressed already..

[elbill]

It's down

[OutCast3k]

It's down

Will be back up soon, sorry about this. Some jerk is ddosing the server, and demanding he's paid X amount of bitcoins before he stops. As if!

[froggyfeels]

Will this be coming back?
Looks like an SSL error, I remember getting onto the site fine before, and then there was the ddos attack, it looks like something else is wrong now.
I would like to try it on testnet.
Hope to see this project continue on.