Version 0.6 release candidate 1

[dunand]

On ubuntu 11.10 x64 I have this error when trying to run ./bitcoind -daemon

I was on 0.5.2 before.

EXCEPTION: 22DbRunRecoveryException       
DbEnv::open: DB_RUNRECOVERY: Fatal error, run database recovery       
bitcoin in AppInit()       

terminate called after throwing an instance of 'DbRunRecoveryException'
  what():  DbEnv::open: DB_RUNRECOVERY: Fatal error, run database recovery

[1714878497]

1714878497

[1714878497]

1714878497

[1714878497]

1714878497

[1714878497]

1714878497

[slothbag]

I have a wallet.dat file I have been using since v0.3.2, after trying v0.6rc1 using the same wallet.dat file I can no longer run v0.5.2, it crashes on startup. The new version must have made some changes to the wallet.dat that previous versions are not handling.

Not a big issue cause I have backups, but thought it might be worth mentioning.

[1Pakis]

On ubuntu 11.10 x64 I have this error when trying to run ./bitcoind -daemon

I was on 0.5.2 before.

EXCEPTION: 22DbRunRecoveryException      
DbEnv::open: DB_RUNRECOVERY: Fatal error, run database recovery      
bitcoin in AppInit()      

terminate called after throwing an instance of 'DbRunRecoveryException'
  what():  DbEnv::open: DB_RUNRECOVERY: Fatal error, run database recovery

I get this error too on xubuntu 11.04

[kjj]

I just used importprivkey to import a key, and the whole thing hung up for several minutes.  Is that normal?

The import appears to have been successful though.

[kwukduck]

Yesterday my pc crashed for no aparent reason (just  went like what seemed to be standby, then after 10 seconds rebooted). When i started bitcoin client it took about 2-3 minutes to load before the gui popped up.
After the gui shows it just freezes, can't click anything, task manager claims it's not responding, also RPC doesnt reply to commands.

I did a chkdsk but all seems fine there...

Also, all seems good when reverting back to 0.5.2.

Can't really explain it or reproduce the crash exactly, all i know is that 0.6rc1 now freezes whenever i start it. Used to work just fine before.

[blueadept]

I just used importprivkey to import a key, and the whole thing hung up for several minutes.  Is that normal?

The import appears to have been successful though.

Yes, the entire blockchain is rescanned for transactions to the key you imported.

[viper1092]

*Drool* 

Some nice looking features here in oh6RC1. Soon as I'm back to my MBP I'm gonna test.

[Pieter Wuille]

• How far is this "complete compatibility" with older versions? Can I move a wallet.dat from this new release that's created these new keys to an older release and have it still work?

Good point. It will probably work, but the compressed pubkeys will be considered non-compressed ones by older versions, so they will probably miss transactions. Over time, the ledger seen by older clients and newer clients may diverge. I'll try to fix this before 0.6.0 final is released.

This was fixed in commit 38067c1.

[Wordlet]

I just used importprivkey to import a key, and the whole thing hung up for several minutes.  Is that normal?

The import appears to have been successful though.

Yes, the entire blockchain is rescanned for transactions to the key you imported.

Does it make the UI unresponsive with no indication as to what is happening? Sounds like a bug to me.

[Diapolo]

I had one strange observation with this RC1 on Windows, which I can't reproduce but perhaps the info is important.

My system has Win7 x64 uses CGMINER for mining and runs the Bitcoin client while I mine. Yesterday I had a bluescreen ... rebooted and tried to start the BC client. I saw the loading screen and thought everything was good, but the GUI simple froze and did not recover from that state. I had the impression the database was corrupt, so I restored a wallet.dat backup and tried to restart the client another time. And again it froze after the loading screen ... interesting fact, I saw via ProcessExplorer, that there were threads working and my data directory grew, so it was in fact re-downloading the block-chain. But no chance, the GUI keept freezing.

So I reverted back to the latest stable release 0.5.2 and now everything is fine again. A further important info I have to give is, I use Microsoft EMET 2.1, which implements some security concepts for applications, which don't support them out of the box via a kind of additional layer, alternatively minimizes attack vectors, so that let's say a buffer-overflow is harder to use for an attack on the application or the host system. I disabled EMET for the BC client the same time I switched back to 0.5.2, so I can't really say if it was a bug with 0.6 or a problem with the bluescreen crash + EMET usage and a database problem. After the re-download of the blockchain I re-activated EMET, so that it's now again active and works without any problems (and it did that for 0.6 before that bluescreen, too ...).

Dia

[Diapolo]

Wallets created with this version of bitcoin will
use 33-byte 'compressed' public keys instead of
65-byte public keys, resulting in smaller
transactions and less traffic on the bitcoin
network. The shorter keys are completely
compatible with older versions.

Will older wallets (the wallet.dat?) get converted to the new format or can this be done somehow without a data-loss?

Dia

[Pieter Wuille]

Will older wallets (the wallet.dat?) get converted to the new format or can this be done somehow without a data-loss?

Dia

Yes; in fact, the wallet format remains compatible as long as no new-style keys are generated. As soon as a new key in generated in 0.6.0 onwards, the wallet will not be compatible anymore with pre-0.6.0 clients.

Note that 0.6.0rc1 does not yet prevent old clients from opening such a new wallet, potentially corrupting the wallet. The next release candidates will do so.

[Diapolo]

I again tried to switch to 0.6 RC1 and GUI still freezes, even with EMET disabled for the BC client :-/.
Reverting to 0.5.2 instantly restores usability ... any ideas, how I can help to debug this? Or when will RC2 be release, so I can see if that one got fixed?

Dia

[Prayer]

Working on setting up a new home server on Ubuntu 11.10, and wanted to build bitcoind from source.  Ended up with the 0.6 code (didn't realize this until later).  Anyways, had fun playing with the import feature (I was banging my head about how to do this in 0.5.2).

As mentioned by another user, the import took a while.  Would it be worth considering threading this off in order to make the user experience a little nicer?

There seems to be something crossed in the getaccountaddress and getnewaddress RPC calls.  I generated a throwaway address at bitcionaddress.com for this demonstration (and masked it to keep anyone from accidentally using it):

Code:

Bitcoin Address: 1XXXXX7zMGHrs7oc1hCSwkNCmmfT8m2mSQ
Private Key: 5JMn6mZwW69NXXXXXXXXXXXXXXXXEgzyjD9NvywV3CzuVUT18wq

Then, I imported it, then got the list of accounts.  It's there!

Code:

me@chewy:~$ bitcoind importprivkey 5JMn6mZwW69NXXXXXXXXXXXXXXXXEgzyjD9NvywV3CzuVUT18wq Garbage
me@chewy:~$ bitcoind listaccounts
{
    "" : 0.00000000,
    "Garbage" : 0.00000000,
    "Prayer" : 0.14000000,
    "Test" : 0.00000000,
    "dunno" : 0.00000000
}

However, when I tried to verify the address on the imported key, it seems to have generated a new key/address pair:

Code:

me@chewy:~$ bitcoind getaccountaddress Garbage
1ZZZZZpXRbgb9dn3bfbwZb9YDKieqgnNaZ

Subsequent calls to getaccountaddress return the same address (no new addresses are generated).  However, you can verify that the 'account' now has two addresses associated with it:

Code:

me@chewy:~$ bitcoind getaddressesbyaccount Garbage
[
    "1XXXXX7zMGHrs7oc1hCSwkNCmmfT8m2mSQ",
    "1ZZZZZpXRbgb9dn3bfbwZb9YDKieqgnNaZ"
]

Having an extra address generated isn't a big deal, but I never asked for a new address to be generated.

I started looking into the code, but I'm not familiar enough yet to figure out what's going on.  I may play around with it tonight and see if I can figure out a fix, but it's been years since I've even looked at c++ and even then it wasn't much.

Also, if I may... shouldn't the client be able to refresh the wallet when a key has been imported?  It's rather annoying to have to stop/restart when importing a key before it shows up in the GUI.  Maybe this would be an argument for splitting server & client functionality, and use the GUI only to make RPC calls against the server.

Now, if my observations are worth anything, the address in my sig is valid.  I only have like .14 BTC now and want a bit more to play with =D.

[Pieter Wuille]

However, when I tried to verify the address on the imported key, it seems to have generated a new key/address pair:

Code:

me@chewy:~$ bitcoind getaccountaddress Garbage
1ZZZZZpXRbgb9dn3bfbwZb9YDKieqgnNaZ

Good point, I don't recall updating the account info when importing. I'll try to do that still before 0.6.0's release.

Also, if I may... shouldn't the client be able to refresh the wallet when a key has been imported?  It's rather annoying to have to stop/restart when importing a key before it shows up in the GUI.  Maybe this would be an argument for splitting server & client functionality, and use the GUI only to make RPC calls against the server.

Yes, that's a known issue, see the issue page. I'm not familiar with the GUI code, however.

[Diapolo]

Yesterday my pc crashed for no aparent reason (just  went like what seemed to be standby, then after 10 seconds rebooted). When i started bitcoin client it took about 2-3 minutes to load before the gui popped up.
After the gui shows it just freezes, can't click anything, task manager claims it's not responding, also RPC doesnt reply to commands.

I did a chkdsk but all seems fine there...

Also, all seems good when reverting back to 0.5.2.

Can't really explain it or reproduce the crash exactly, all i know is that 0.6rc1 now freezes whenever i start it. Used to work just fine before.

I should have read this thread from the beginning, as this is the behaviour I observed, too! Any info on this?

Dia

[D.H.]

It would be great if someone could create a windows installer for release candidates, like Luke-Jr did with the next-test branch here. Would bring in a lot more testers I'm sure.

[mcorlett]

It would be great if someone could create a windows installer for release candidates, like Luke-Jr did with the next-test branch here. Would bring in a lot more testers I'm sure.

Is this not satisfactory?
http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.6.0/test/

[Pieter Wuille]

I should have read this thread from the beginning, as this is the behaviour I observed, too! Any info on this?

Unfortunately, no. There have been a rather large number of bugfixes since 0.6.0rc1 however, and 0.6.0rc2 will be created very soon. I hope this problem is fixed by now.

[D.H.]

It would be great if someone could create a windows installer for release candidates, like Luke-Jr did with the next-test branch here. Would bring in a lot more testers I'm sure.

Is this not satisfactory?
http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.6.0/test/

Oops, sorry about that. I read through the first post a bit too quickly.