[PULL] private key and wallet export/import

[jackjack]

You forgot the --pwpath argument
Should be .\ on Windows I guess

[krepta3000]

I also tried it with that option.

Code:

--pwpath=.\

Same error.  I also tried it with the long form of the path.

Code:

--pwpath=c:\progra~1\bitcoin\pywallet\

Same error.

[jackjack]

Ok I think I see the problem
Run '.\pywallet.py --dumpwallet --datadir m:\bitcoin\old\'
Does it looks like that: (don't post it)

Code:

        {
            "addr": "1lqksjiqsidj", 
            "hexsec": "kkjqsdsd", 
            "reserve": 1, 
            "sec": "5vkkqsjdnqs"
        }
    ], 
    "tx": "unsupported", 
    "version": 32500
}

If not, run 'python pywallet.py --dumpwallet --datadir m:\bitcoin\old\'

[krepta3000]

Yes, it looks like that.

[jackjack]

Which one?
The first?

[krepta3000]

Well, anyway, thanks to pywallet.py, whatever branch, I've got a wallet that isn't broken with 0/unconfirmed transactions that refuse to be processed by the bitcoin network, Yay!   And now I will run bitcoin more often, and backup my wallet more often too.

[jackjack]

Thanks for the report, I'll investigate

[krepta3000]

Which one?
The first?

Yes, it looks like the first one.

[BitcoinLocator]

What is the status/eta of this pull request?  Unless integration is eminent we'd be willing to offer a bounty to help accelerate it's release. We really need this feature, our current workaround involving multiple bitcoind instances, and relaunch rescan per user session is catastrophically cumbersome.

Thanks!

[Gavin Andresen]

What is the status/eta of this pull request?  Unless integration is eminent we'd be willing to offer a bounty to help accelerate it's release. We really need this feature, our current workaround involving multiple bitcoind instances, and relaunch rescan per user session is catastrophically cumbersome.

It will get pulled/released faster if you help test it.  Pull the patch, compile a custom bitcoind, then run it and report any bugs.  If you don't find any bugs, add a short note either here or in the PULL request.

Testing is currently the bottleneck for getting new features...

[BitcoinLocator]

It will get pulled/released faster if you help test it.

Not sure how competent we are at that, but we'll give it a shot.

If anyone else reading this thread wants to help with testing, we will gladly throw you some bitcoins to make it worth you while.

[riush]

I get an error with removeprivkey on encrypted wallets. Works fine on non-encrypted.

foo@bar:~/work/bitcoin/src$ ./bitcoind getnewaddress
mgVu7WXEdbCHnKiQW8q1g7ND5e6jR97NGX
foo@bar:~/work/bitcoin/src$ ./bitcoind dumpprivkey mgVu7WXEdbCHnKiQW8q1g7ND5e6jR97NGX
93AywmB22cb5H59m6vttmPUFCPf98sB7NEwj48q9RvbaMGQHzbd
foo@bar:~/work/bitcoin/src$ ./bitcoind removeprivkey 93AywmB22cb5H59m6vttmPUFCPf98sB7NEwj48q9RvbaMGQHzbd
foo@bar:~/work/bitcoin/src$ ./bitcoind importprivkey 93AywmB22cb5H59m6vttmPUFCPf98sB7NEwj48q9RvbaMGQHzbd
foo@bar:~/work/bitcoin/src$ ./bitcoind encryptwallet password
foo@bar:~/work/bitcoin/src$ ./bitcoind walletpassphrase password 1000
foo@bar:~/work/bitcoin/src$ ./bitcoind getnewaddress
msKK1iq6LYLxz8iUvovWMqh37T3U9fha4M
foo@bar:~/work/bitcoin/src$ ./bitcoind dumpprivkey msKK1iq6LYLxz8iUvovWMqh37T3U9fha4M
92isYRmvbjH7TcYnqKU3YZUmQvUatRCmYSPYsAoEku6Xv8s7qhy
foo@bar:~/work/bitcoin/src$ ./bitcoind removeprivkey 92isYRmvbjH7TcYnqKU3YZUmQvUatRCmYSPYsAoEku6Xv8s7qhy
error: {"code":-1,"message":"GetAllReserveKeyHashes() : unknown key in key pool"}

[riush]

What is the status/eta of this pull request?  Unless integration is eminent we'd be willing to offer a bounty to help accelerate it's release. We really need this feature, our current workaround involving multiple bitcoind instances, and relaunch rescan per user session is catastrophically cumbersome.

Maybe one bitcoind handling multiple (separately encrypted) wallets would even better suit your needs? I'd chip in on that too.
Any word on how realistic that would be?

[nelisky]

What is the status/eta of this pull request?  Unless integration is eminent we'd be willing to offer a bounty to help accelerate it's release. We really need this feature, our current workaround involving multiple bitcoind instances, and relaunch rescan per user session is catastrophically cumbersome.

It will get pulled/released faster if you help test it.  Pull the patch, compile a custom bitcoind, then run it and report any bugs.  If you don't find any bugs, add a short note either here or in the PULL request.

Testing is currently the bottleneck for getting new features...

I have been running bitcoind, in the recent past master/HEAD with patches mentioned by sipa on bitcoin's github  and more recently sipa's own showwallet branch. Other than the one bug it had with the construction of the privkey for printing on dump (a typo, already fixed) and the fact that the import of each key takes multiple minutes (it scans all blocks for transactions, so that's expected, I would like to see a batch import in the future so a single sweep could verify multiple addresses) I can tell you it is working very well.

I'm using it quite heavily these days to organize offline wallets with a set amount kept on each address and then keeping the addresses and priv keys on paper in a safe.

My workflow only touches dumpprivkey and importprivkey. There is a glitch with the address balance listing after importing but it's a known bug and doesn't affect the global wallet balance consistency.

I would LOVE to see this pulled into bitcoin proper, and if you need any specific unittest written (there's a framework now, right?) I can try to find an hour to give it a shot soon.

[samr7]

I'd like to report a minor problem.

If I send coins to a specific address, and then import the private key for that address, the UI updates and reports the transaction as "Payment to yourself."  However, the total coin balance doesn't seem to update until after bitcoin is restarted.

[casascius]

I have tested it, and it works for me (aside from, I have encountered the aforementioned wallet dump glitch which is said to be a typo and now fixed).

[Eli]

This is a bit ridiculous...

The idea is obviously useful enough that dozens of people are shown their support, and only a single few expressed their discontempt with reasons like "it's not safe for stupid users".

Right, I agree that it's not safe for stupid users, and I agree that sometimes you have to protect them. But creating this artificial entry barrier just to deter developers who are not familiar with compiling and patching C code is just plain evil, and bordering stupidity.

It's obviously a needed and wanted feature, you don't want people having this feature turned on by mistake, right? how about:

Code:

$ bitcoind -server -rpcuser=... -rpcpassword=... [b]-advanced[/b]

With this small flag, -advanced, users will be able to actively activate the feature, and only those who think that they know what they're doing will be able to harm themselves.

Now please, for the love of peace, pull this patch to the main client, and make it official. I'll write the documentation if needed.

[joulesbeef]

the paper wallets and casascius's real bitcoin or bitbillswhich would be great for promotion of bitcoin, could use something like this.

[thirdlight]

No[w] please, for the love of peace, pull this patch to the main client, and make it official.

+1

[molecular]

No[w] please, for the love of peace, pull this patch to the main client, and make it official.

Testing is currently the bottleneck for getting new features...

... maybe we should test and post our results in the comments on github: https://github.com/bitcoin/bitcoin/pull/220