Bitcoin Core 0.10.0 has been released

[etile]

watch-only addresses works, but show only "receive" and "generate", without "spending"

So balances more bigger then should be in real

How to fix?

[Mad_Max]

10 have same bug as was in 8.xx and 9.xx versions.

If you set -dbcache=xxx to high values like 256 or 512 Mb (min = 4, max=1024, default=100) client with a high probability will hang or crash at initial blockchain download (or just sync a few hundreds or thousand blocks after few days or weeks offline) trowing such error:

EXCEPTION: St9bad_alloc      
std::bad_alloc
bitcoin-qt.exe in ProcessMessages()      

2015-02-21 16:43:12 ProcessMessage(block, 536164 bytes) FAILED
      
bitcoin-qt.exe in Runaway exception  

In 2 varinats:
- just simple instant crash with such error
- no crash but client stop download new blocks (all other look fine), and if you close client (just normal exit, NOT process killing) it crash at exit with same error

With default settings (-dbcache=100) no such errors but at 100mb and current big blocks bitcoin core SO SLOW if data placed on HDD (not SSD) and killing disk and making computer almost unresponsive for other user activity (if it is not dedicated computer/server) while blockchain sync...

If this bug not fixed bitcoin will become  "SSD only" program soon. And it is a problem because of high volume of blokcchain data which keep growing even with some acceleration.  

P.S.
I sow this behavior on WinXP and Windows 7 32bit. Not tested on other systems yet.
And this error not simple run out of free RAM - in all cases i sow this error at least 500-1000 Mb RAM was still free before crash. So it is internal bitcoin (or may be levelDB ?) bug.

[gmaxwell]

10 have same bug as was in 8.xx and 9.xx versions.

If you set -dbcache=xxx to high values like 256 or 512 Mb (min = 4, max=1024, default=100) client with a high probability will hang or crash at initial blockchain download (or just sync a few hundreds or thousand blocks after few days or weeks offline) trowing such error:
[..snip..]
I sow this behavior on WinXP and Windows 7 32bit. Not tested on other systems yet.
And this error not simple run out of free RAM - in all cases i sow this error at least 500-1000 Mb RAM was still free before crash. So it is internal bitcoin (or may be levelDB ?) bug.

You're running out of address space.  A 32 bit program only has access to 4GB of address space. At least 1GB of that is used by the kernel.  Usually another GB is lost to libraries, stack, and other userspace overheads.  Leaving you with maybe 2GB of process address space. It then loses more from running multiple threads, each with their own stacks and heap arenas which use address space even if they're not using actual memory yet.  Then you get the process's own memory usage, and there just isn't much space left for huge caches.

We should probably cap the dbcache maximum on 32 bit hosts so that you don't get the surprises crashes at runtime... but sadly, I don't think it will be possible to accommodate your desire for a super large cache.   Have some patience though, the default cache is way more than adequate for a running node-- cranking it up only makes a real difference on the initial block download, which is a one time event.

[TurdHurdur]

Trying to build on Debian 6 LTS I get:

Code:

helpers/memenv/memenv.cc:65: error: ‘SIZE_MAX’ was not declared in this scope

Any idea how to remedy this?

My guess is you are running on a version of GCC which is at least outdated by 4 years of hard work. But if you dont post more details, I cant tell you more.

Think I fixed it, inserted the following at line 5 of src/leveldb/helpers/memenv.h and it compiles fine.

Code:

#ifndef __STDC_LIMIT_MACROS
#define __STDC_LIMIT_MACROS
#endif

[crazyearner]

Well seems if you got over 1k transactions in a mixture of wallets seems to crash client and had nothing but problems with v10 in the end  I had to delete wallet and rescan after and even this took like 12 times or more to finaly get done without it crashing ever time it scanned.

[cyberpinoy]

100 people working on the core and 0 people working on the demand for bitcoin, you guys are doing great work for your coin if your priorities weren't all messed up I think you should start doing some real work and maybe do something positive for your investors like ahh CREATE A DEMAND so people will have a reason, a purpose and a need to buy more bitcoins than they sell. STOP waiting for others to create a demand for your coin.

Not sure if serious or being trolled..  

Nope not trolling,

Just tired of seeing the bitcoin developers concentrate so much on a core protocol that is literally utilized by less than 40% of the people using bitcoins right now, and out of the ones that do use it only about 10% of those actually care about bitcoin and do anything for bitcoin. SO you see the priorities of not only the Bitcoin developers, but the ones who actually use the bitcoin core is doing nothing for the value or the demand of bitcoin. So the investors we do have make little to no gains and the ones looking into bitcoin are turned away since there is no demand for bitcoin that would make it a good investment.

[Mad_Max]

You're running out of address space.  A 32 bit program only has access to 4GB of address space. At least 1GB of that is used by the kernel.  Usually another GB is lost to libraries, stack, and other userspace overheads.  Leaving you with maybe 2GB of process address space. It then loses more from running multiple threads, each with their own stacks and heap arenas which use address space even if they're not using actual memory yet.  Then you get the process's own memory usage, and there just isn't much space left for huge caches.

We should probably cap the dbcache maximum on 32 bit hosts so that you don't get the surprises crashes at runtime... but sadly, I don't think it will be possible to accommodate your desire for a super large cache.   Have some patience though, the default cache is way more than adequate for a running node-- cranking it up only makes a real difference on the initial block download, which is a one time event.

You sure this issue related ONLY to 32 bit systems? I mean it already know bug witch was tested and the cause of it clearly tracked to 32bit address limitations? I sow it one time even with ~2000 Mb free RAM... (4 Gb physical RAM, win 32bit, 3.5 Gb available due 32bit limitations, ~2 Gb free)

And my desire not for a super large cache, it about speed up bt disk work on HDDs. High dbcach help with it - not much, but helps.
In theory initial block download one time event, but in practice (i use btc about 2 years now) i already do this 3 times:
- my very first initial block download
- redownload after strange client crush (it was one of 0.8.x versions - not sure about exact reason, it not repeated and i upgrade ver soon) because of database corruptions
- redownload (actuality was DB reindex + some redownload) after big power flux in my city (result in sudden hard reboot) while client sync blocks from last few days - resulting in another blockchain corruption

2nd took > 1 day, 3rd (about 2 month ago) took about 2 days of heavy disk work!

Now I already know redownload usual not needed in such situations and -reindex may be enough, but if data on HDD (not SSD), -reindex takes almost the same time as full initial download, because HDD disk is bottleneck here, not Internet connections speed (if it not very slow of course? but 10-20 Mbit/s already enouhg to make HDD or CPU bottleneck, not network)
To avoid such situations in future i now keep backup of indexed blockchain- which almost doubled used space (~34 Gb working db + ~25 Gb zipped bacup)

And sometimes even sync of last one or few weeks blocks is like a pain in ass  (eg you need to check or send transactions fast, but realize what client was not synced from last week or two and need wait about hour before full sync)

[Newar]

Nope not trolling,

Just tired of seeing the bitcoin developers concentrate so much on a core protocol that is literally utilized by less than 40% of the people using bitcoins right now, and out of the ones that do use it only about 10% of those actually care about bitcoin and do anything for bitcoin. SO you see the priorities of not only the Bitcoin developers, but the ones who actually use the bitcoin core is doing nothing for the value or the demand of bitcoin. So the investors we do have make little to no gains and the ones looking into bitcoin are turned away since there is no demand for bitcoin that would make it a good investment.

You could stop looking at it as an investment. 

[develCuy]

This release has exciting features but ones I like the mos are the 'sendfreetransactions' setting, the fee calculation database, and the switch to  libsecp256k1 library. More security and more fair fees make a LOT of sense!

[gmaxwell]

You sure this issue related ONLY to 32 bit systems? I mean it already know bug witch was tested and the cause of it clearly tracked to 32bit address limitations? I sow it one time even with ~2000 Mb free RAM... (4 Gb physical RAM, win 32bit, 3.5 Gb available due 32bit limitations, ~2 Gb free)

Yes, and I explained why (the fact that 32 bit systems needs a lower limit was also well known in advance; Bitcoin imposes a hard cap of 1024 on 32bit systems... but it sounds like from your report that this is not low enough). Large cache settings (e.g. 4096MB) are extensively tested on 64 -bit hosts.

In theory initial block download one time event, but in practice (i use btc about 2 years now) i already do this 3 times:
- my very first initial block download
- redownload after strange client crush (it was one of 0.8.x versions - not sure about exact reason, it not repeated and i upgrade ver soon) because of database corruptions
- redownload (actuality was DB reindex + some redownload) after big power flux in my city (result in sudden hard reboot) while client sync blocks from last few days - resulting in another blockchain corruption

Getting corrupted is a separate issue, that shouldn't happen. Any issues causing that should be fixed. FWIW, if you end up corrupted a reindex will work which tends to be faster and less IO intensive because it leaves the blocks in place (doesn't have to write the blocks again)... not a reason to not fix the corruption on unclean shutdown, but perhaps it's a helpful bit of advice for you.

Well seems if you got over 1k transactions in a mixture of wallets seems to crash client

Can you try restating this? What does "in a mixture of wallets" mean?  I've personally tested Bitcoin core with wallets with tens of thousands of transactions without issue, but I'm not sure precisely what you're doing, or what you mean by crash.

[tee-rex]

I've been trying to download the new client, but whenever I try I get only "The connection was reset. The connection to the server was reset while the page was loading" error. What might be wrong? I'm using the latest Firefox.

I'm scared to download torrent.

[dserrano5]

I'm scared to download torrent.

Why? You can verify the hash and the PGP signature of the SHASUMS file so you are certain that your copy is good.

[cyberpinoy]

Nope not trolling,

Just tired of seeing the bitcoin developers concentrate so much on a core protocol that is literally utilized by less than 40% of the people using bitcoins right now, and out of the ones that do use it only about 10% of those actually care about bitcoin and do anything for bitcoin. SO you see the priorities of not only the Bitcoin developers, but the ones who actually use the bitcoin core is doing nothing for the value or the demand of bitcoin. So the investors we do have make little to no gains and the ones looking into bitcoin are turned away since there is no demand for bitcoin that would make it a good investment.

You could stop looking at it as an investment.  

Well what are we supposed to look at it as a currency? A transaction process? Its not a currency, and its value as a transaction protocol is not even utilized or promoted by the developers. As for now it is an investment, and since the developers who have made a but ton of money from the investors have seemingly already got paid you (and they) could say its time for them to get off their ass and do thier job. if not for the people who looked at bitcoin as an "investment" where would bitcoin be? NOWHERE, if not for the people who looked at bitcoin as an "investment" where would the developers be financially, most likely worse off than they were 6 years ago. Maybe what you and the bitcoin developers need to realize is they created the monster that is now bitcoin, they made it an investment out of greed. If not it would never have become one would it? So the bottom line is they wanted to make money and steered its real reason of creation from a transaction protocol to an investment, now its time to sleep in the bed they made. Maybe now we have a glimpse of why Satoshi left huh?

And FYI they cant even do their job right 100 people worked on the update and I still have 1 hour transaction times on companies using the bitcoin core.

How can we get Satoshi back??

[Boussac]

Nope not trolling,

Just tired of seeing the bitcoin developers concentrate so much on a core protocol that is literally utilized by less than 40% of the people using bitcoins right now, and out of the ones that do use it only about 10% of those actually care about bitcoin and do anything for bitcoin. SO you see the priorities of not only the Bitcoin developers, but the ones who actually use the bitcoin core is doing nothing for the value or the demand of bitcoin. So the investors we do have make little to no gains and the ones looking into bitcoin are turned away since there is no demand for bitcoin that would make it a good investment.

You could stop looking at it as an investment.  

Well what are we supposed to look at it as a currency? A transaction process? Its not a currency, and its value as a transaction protocol is not even utilized or promoted by the developers. As for now it is an investment, and since the developers who have made a but ton of money from the investors have seemingly already got paid you (and they) could say its time for them to get off their ass and do thier job. if not for the people who looked at bitcoin as an "investment" where would bitcoin be? NOWHERE, if not for the people who looked at bitcoin as an "investment" where would the developers be financially, most likely worse off than they were 6 years ago. Maybe what you and the bitcoin developers need to realize is they created the monster that is now bitcoin, they made it an investment out of greed.

A transaction network based on a transaction protocol. That's what Bitcoin is.
Then you have investors who speculate on its mass market adoption and lobbyist who insist on calling it a currency which is only one application of the network.
Meanwhile the technology and the ecosystem are making good progress thanks to the developers. Imho an investor is simply a middleman, a time intermediary between now and mass market time.

[dooglus]

Nope not trolling,

Just tired of seeing the bitcoin developers concentrate so much on a core protocol that is literally utilized by less than 40% of the people using bitcoins right now, and out of the ones that do use it only about 10% of those actually care about bitcoin and do anything for bitcoin.

Yeah, sometimes it seems like the bitcoin-core developers are focussed on developing bitcoin-core.

Something must be done about this crazy situation.

In an unrelated note I heard that Microsoft has huge numbers of developers who do nothing but develop Microsoft software. Maybe you should email them and get it sorted out.

SO you see the priorities of not only the Bitcoin developers, but the ones who actually use the bitcoin core is doing nothing for the value or the demand of bitcoin. So the investors we do have make little to no gains and the ones looking into bitcoin are turned away since there is no demand for bitcoin that would make it a good investment.

Do you not understand that having a functional client is absolutely essential, and that without it the network just doesn't work? We can't all use SPV clients - what would they connect to?

I think where you're going wrong is in thinking that "we" have or are looking for "investors".

Anyway, if there is no demand for bitcoin why is it trading at over $200 per unit?

[countryfree]

I had been using online wallets and blockchain-free software for the past 18 months, but since this was a major release, I decided to give it a try. I was amazed to see the free space on my HD goes from 339 to 305 Go in about 10 and a half hours. I had never downloaded so much data in so little time!

[ChuckBuck]

Appreciate your great work

Yes awesome work guys

100 people working on the core and 0 people working on the demand for bitcoin, you guys are doing great work for your coin if your priorities weren't all messed up I think you should start doing some real work and maybe do something positive for your investors like ahh CREATE A DEMAND so people will have a reason, a purpose and a need to buy more bitcoins than they sell. STOP waiting for others to create a demand for your coin.

And before you code and bitcoin developer lovers get all bent out of shape, I have been in bitcoins for a long time, mined them traded them,own informational websites, facebook groups, held seminars and never have I ever used the bitcoin core wallet to much work on that core and not enough for the demand. The core wallet is to confusing to people without coding experience, to understand it, and god forbid if they are not super experienced and only have a command line to use they wont get past the git clone, the developers have spent way to much time on the core development that is simply put, NOT userfriendly. Unless its in a graphical interface, it deters users from even using bitcoins so adoption and adaptation is more complex. the value of bitcoin continues to fall with every new release. its time to get your priorities together, because in another 5 years if things keep going this way, there will be no more bitcoins

This should answer your non-technical concerns regarding the Bitcoin Foundation's priorities:

http://blogs.wsj.com/moneybeat/2015/02/24/bitbeat-for-bitcoin-some-good-press-finally/

Finally get some ads and publicity that'll help people become familiar with Bitcoin, thus directly pushing technology to meet the users needs (making Core and Bitcoin more user friendly).

[Hamuki]

Will get this version later today.

Thanks!

[pecintapantai]

what is bitcoin core?
hmmm 

[Mikestang]

I've been running Bitcoin Core 0.10.0 stable for several days, but this morning I woke up to an error window telling me to see debug.log.  I've had errors before, and always just restart the program and no problem.  Well this time I restarted it and the loading screen comes up and it went through the block verification so fast as if there was no blockchain on my computer, then the program crashed.  I repeated the step and same thing, no error window either time.  I reinstalled 0.10.0, reset the options, pointed it to my data directory, but it still crashes at the loading window with no error message.  Here is a debug log from an attempted start of the program:

Code:

2015-02-27 14:55:26 Bitcoin version v0.10.0 (2015-02-13 09:55:11 +0100)
2015-02-27 14:55:26 Using OpenSSL version OpenSSL 1.0.1k 8 Jan 2015
2015-02-27 14:55:26 Using BerkeleyDB version Berkeley DB 4.8.30: (April  9, 2010)
2015-02-27 14:55:26 Default data directory C:\Documents and Settings\Mn\Application Data\Bitcoin
2015-02-27 14:55:26 Using data directory D:\Documents and Settings\Mn\Application Data\Bitcoin
2015-02-27 14:55:26 Using config file D:\Documents and Settings\Mn\Application Data\Bitcoin\bitcoin.conf
2015-02-27 14:55:26 Using at most 125 connections (2048 file descriptors available)
2015-02-27 14:55:26 Using 2 threads for script verification
2015-02-27 14:55:26 Using wallet wallet.dat
2015-02-27 14:55:26 init message: Verifying wallet...
2015-02-27 14:55:26 CDBEnv::Open : LogDir=D:\Documents and Settings\Mn\Application Data\Bitcoin\database ErrorFile=D:\Documents and Settings\Mn\Application Data\Bitcoin\db.log
2015-02-27 14:55:26 Error: Couldn't open socket for incoming connections (socket returned error An address incompatible with the requested protocol was used.  (10047))
2015-02-27 14:55:26 Bound to 0.0.0.0:8333
2015-02-27 14:55:26 init message: Loading block index...
2015-02-27 14:55:26 Opening LevelDB in D:\Documents and Settings\Mn\Application Data\Bitcoin\blocks\index
2015-02-27 14:55:26 Opened LevelDB successfully
2015-02-27 14:55:26 Opening LevelDB in D:\Documents and Settings\Mn\Application Data\Bitcoin\chainstate
2015-02-27 14:55:26 Opened LevelDB successfully
2015-02-27 14:55:26 LoadBlockIndexDB: last block file = 0
2015-02-27 14:55:26 LoadBlockIndexDB: last block file info: CBlockFileInfo(blocks=125316, size=135455381, heights=0...119964, time=2009-01-03...2011-04-24)
2015-02-27 14:55:26 Checking all blk files are present...
2015-02-27 14:55:26 LoadBlockIndexDB(): transaction index disabled
2015-02-27 14:55:26 LoadBlockIndexDB(): hashBestChain=00000000660aaa727ebc610ed069e23abd28f516c614de894e1492994df78edc height=5350 date=2009-02-23 15:53:18 progress=0.000041
2015-02-27 14:55:26 init message: Verifying blocks...
2015-02-27 14:55:26 Verifying last 288 blocks at level 3
2015-02-27 14:55:26 No coin database inconsistencies in last 289 blocks (293 transactions)
2015-02-27 14:55:26  block index             406ms
2015-02-27 14:55:26 init message: Loading wallet...
2015-02-27 14:55:26 nFileVersion = 100000
2015-02-27 14:55:26 Keys: 0 plaintext, 201 encrypted, 201 w/ metadata, 201 total
2015-02-27 14:55:26  wallet                   15ms
2015-02-27 14:55:26 Reindexing block file blk00000.dat...
2015-02-27 14:55:26 Block Import: already had block 00000000c937983704a73af28acdec37b049d214adbda81d7e2a3dd146f6ed09 at height 1000
2015-02-27 14:55:26 Block Import: already had block 00000000dfd5d65c9d8561b4b8f60a63018fe3933ecb131fb37f905f87da951a at height 2000
2015-02-27 14:55:26 Block Import: already had block 000000004a81b9aa469b11649996ecb0a452c16d1181e72f9f980850a1c5ecce at height 3000
2015-02-27 14:55:26 Block Import: already had block 00000000922e2aa9e84a474350a3555f49f06061fd49df50a9352f156692a842 at height 4000
2015-02-27 14:55:26 Block Import: already had block 000000004d78d2a8a93a1d20a24d721268690bebd2b51f7e80657d57e226eef9 at height 5000
2015-02-27 14:55:27 UpdateTip: new best=0000000083be2e513d76f87ebdf93310040ecb5ec14556bbe2b635bad65bdbdb  height=5351  log2_work=44.385884  tx=5419  date=2009-02-23 15:56:37 progress=0.000041  cache=1

{several more of these entries, deleted to save space}

2015-02-27 14:55:28 UpdateTip: new best=00000000507fcfe6138eaf4ebbb1766337d78bbbda1f5e15fba2c822d77ec01c  height=5762  log2_work=44.492626  tx=5830  date=2009-02-27 10:53:38 progress=0.000044  cache=411
2015-02-27 14:55:28 mapBlockIndex.size() = 5764
2015-02-27 14:55:28 nBestHeight = 5762

Any assistance would be appreciated, cross posting this in the beginners and help section, too.