Bitcoin Forum

Official release build is now available http://www.ppcoin.org/ (via sourceforge)

What's in 0.3.0 release:

• Stake generation protocol upgrade (protocol switch March 20th)
• Qt UI support
• Fix compatibility with vanitygen (note: private keys dumped in v0.2 is no longer importable into v0.3.0, must dump again from v0.3.0 client)
• Miscellaneous bug fixes and improvements

The protocol upgrade in 0.3.0 includes a new algorithm to derive proof-of-stake hash modifier, the entity that scrambles computation for stake owners, which replaces the current proof-of-stake difficulty used as modifier in 0.2 protocol. The design was started late September last year, when I first began to realize the issues with using difficulty as modifier. Honorary mention also goes to Jutarul, who independently discovered and verified an issue with using difficulty as modifier and published on bitcointalk in December last year, while successfully executed a demo attack on the block chain. Other changes in the protocol include starting hash weight from 0 at the 30-day mininum age, and requirement that coinstake timestamp must equal block timestamp. Overall 0.3 protocol should significantly strengthen the proof-of-stake protection and resolve the current known vulnerabilities.

Upgrade instructions:

Upgrade is required before Wednesday March 20th. Nodes that have not upgraded before this deadline will disconnect from the main ppcoin network (hard fork).

To upgrade:
1) Before upgrade, you must first backup your wallet:
ppcoind backupwallet <destination_backup_wallet_name>
2) Note down your balance/stake/newmint amounts. Shutdown ppcoin
ppcoind stop
3) Inside your wallet directory, remove all files and subdirectories except for leaving wallet.dat and ppcoin.conf in place.
4) Download 0.3.0 and unpack/install.
5) Start up ppcoind/ppcoin-qt normally.
6) Use getinfo/about to confirm your ppcoind/ppcoin-qt version is now 'v0.3.0ppc-beta'
7) You should see block chain is being downloaded again. Please note during redownload of block chain your balance will be shown initially as 0 and change as block chain is being downloaded. This is expected. When block chain download completes your balance/stake/newmint should return to the same value before the upgrade.

My sincere appreciation to co-contributors of 0.3.0 release:

• Robert VanHazinga of Hartland PC (dreamwatcher) for the vanitygen compatibility fix
• Jutarul for demonstrating stake generation vulnerability
• EskimoBob for reporting issue fixed in 0.3.0

Cheers and have fun!

Sweet nice work and gtz on the new release :) am installing now.. will try on Windows 7, XP, Server 2008 and Ubuntu 12.04 so i'll let u know if I have any problems 8)

up and running on win7

Okay i still think that not talking about how the algorithm works isn't going to make the boogey man go away.

Can you explain publicly the new POS algorithm?

QT version ?

this

First of, congrats to the release! I can only imagine how much work has been put into it. Also your acknowledgement is somewhat flattering.

So far there has been a lot of secrecy about the development team and process of ppcoin. This has to be improved if this cryptocurrency shall be here to stay. It is perfectly possible to stay anonymous while performing a transparent development. Up to today you haven't stated a reasoning for your procedure.

If you have any design documents, which lead to the derivation of the 0.3.0 POS implementation please share them with the community. Also it would be great if you could explain HOW exactly the new implementation removes the weaknesses of a computational attack on the POS generation.

Also, please know that I have nothing against opaque development in general as long as the source code is published and can be verified. There are a lot of reasons why developers choose opaque development - the predominated one being lack of time to prepare documentation and engage in algorithmic design discussions.
However, I don't understand why it has to be that way here. Especially network security is a critical topic which benefits from a peer developed process, making the code stronger - not weaker.

Thank you

Thanks. Surely you have already touched upon the reasons of what you refer to as 'opaque' development, mostly, due to lack of resources, secondly, for security concerns. I only have time to discuss the design with trusted peers before release. I hope you can understand that there is lot of work involved and it's not trivial work to even understand the design and its intricacies. There is no separate document, I have put some comments into the source code, it's not long at all, only about 400 lines in kernel.cpp and some of it is preexisting code in v0.2. Interested parties can take time to look at it, and discuss it maybe in my disclosure thread. I'll try to answer some of the questions along the way.

Thanks, that should clear things up. Although I don't quite follow you on "security concerns". It seems that your understanding of cryptology got things backwards - at least from my perspective.

Not really, we just have different perspectives. As the maintainer of ppcoin I have obligations to not talk about unfixed vulnerabilities freely. It's a cryptocurrency in production, not just a research project that no investor loss would be involved. If you ask Gavin about his position on this matter he likely would have to tell you the same thing.

I agree with the sentiment to keep vulnerabilities confidential which can lead to immediate harm, until a solution is in place. However, suppression of reporting leads to a false sense of security - for the unaware. It would be great to have some sort of alarm in place. But this is an issue not restricted to ppcoin. It's not even restricted to cryptocurrencies or software. It's a general aspect of governance.

Excellent job ya'll.

"destination" - you actually need to add a destination path
$ ./ppcoind backupwallet /home/destination-dir-here/new_wallet_name.dat

Great! Now let's see how long until another vuln is discovered  ;D ;D ;D

Will be updating later on this afternoon =) gj.

Finaly !!!!!
Thanks and for great GUI

upgraded from source. Did not encounter any problems. Good job.
I'll compile ppcoin-qt later.

I upgraded cryptocoin explorer last night.

Install went smooth, and no issues so far.  :)

Note: if your wallet contains more than a couple thousand transactions, when you are redownloading block chain the Qt UI might appear locked up. This is a known issue due to the heavy-lifting of wallet scanning I believe but it's not really locked up and still doing the work. Just leave it to finish the redownload of block chain, or you can use command line getinfo to check its progress (getinfo response will be slower also in this situation).

If you have really huge/fragmented wallet causing redownload to be really slow, another option might be to redownload block chain with an empty wallet, then stop ppcoin and replace the empty wallet with real wallet backup.

... or not.  There's a difference between "unfixed vulnerabilities" and "half-baked design."

I think big decisions that affect the fundamentals of the design should be discussed in the open (see the current Bitcoin debate over raising the block size limit).

It's a whole gray area I am not really interested debating about this. You make judgement on bitcoin what should be discussed in the open what needs to be kept under wrap until it's patched, I make my judgement on ppcoin. Meanwhile, let's focus on getting the real work done.

Yes, but at the same time people can lead dishonest discussions of design decisions to spread uncertainty about new or different ideas. For example here (https://bitcointalk.org/index.php?topic=144471.0) where gmaxwell spreads uncertainty about ripple. Or this (https://bitcointalk.org/index.php?topic=140800.0) thread where I make a dishonest argument against PPCoin to see why Sunny King made some decisions which the Bitcoin developers think is a bad idea. Or the quoted post where the lead developer of Bitcoin spreads uncertainty against an altcoin attempting to fix its problems.



Anyways, congratulations Sunny King. I hope whatever you did fixes the problems you were having.

Done!

Nice have the qt now as default, just got 10 PPCs but worth keep updated xD

I accept donations xD

PKteaUnJXfRuZGFmBBh67cWu5ymnR9RBoF

Good job!

Hmm interesting  ;) Okay so you sound like you are okay with the difficulty adjustment formula now. Well if freicoin is interested in adopting it my offer to help still stand.

I think your difficulty adjustment is a pretty good demonstration that cryptocurrency is not as delicate as a lot of people claim it is. Also see the 'resurrection' of BBQCoin.

However, I'm not sure if it is a good idea to use your idea for a coin as large as bitcoin. I'm still inclined to think that using a average of many past points is better than an EMA for this application. Who knows, you may be right.

Anyways, congratulations. I haven't had the chance to see what the problem was. Programming is not my background, to be honest.

No problem, thanks galambo  :)

This kind of proves the point that the design should be discussed openly.

What is the problem with that? Do you fear that by discussing it publicly that someone may find an exploit in your algorithm? Gavin makes a valid point that the two he mentions in QUOTES are worlds apart.

not sure what all the secrecy of the design is about....but then again like you said...we'll just have to resort to the code which limits the participants in the discussion to only those who know how to read source code. In my opinion that is close-minded.

But to each his own.

Glad Gavin commented on this.

Reminder that upgrade to 0.3 is required before March 20th.

The only dishonesty here is your claim that I've been dishonest. As you can see in the thread, Joel has (generally?) agreed with every attack scenario on the ripple distributed consensus I've proposed. It's unclear how serious they might be because its unclear how validator trust would be configured. Joel points out that in practice it might be fine.

Actually _disclosing_ how a thing that you're asking people to use and put value it works is the most effective remedy to people spreading 'dishonesty' about it. Secrecy doesn't prevent someone who wants to make dishonest claims is not at all inhibited by secrecy. Quite the opposite, being secretive about it is an open invitatation for people to speculate that it works however they want to claim it works— both in more secure than reality and less secure than reality directions.

I certainly agree with not giving people exploit walkthroughs on systems people depend on... but that doesn't excuse swapping out a consensus algorithm and not explaining how its supposed to work. No point in arguing it— Sunny can shy the sunshine as much as he likes: everyone else is free to run far, far away.

If anyone wants to speculate on the changes: here is a diff to get you started (http://pastebin.com/59LWFBWN).

I do note that the "eventually be removed" centralized control over the blockchain identity are still there, even though the system has transitioned to mostly POS mining, and in fact it seems to have been made _much_ more central to the system: Now a centrally controlled clock must tick to advance the stake search. It also seems to me that whomever can create these modifiers could also use computation to choose ones that advantage their own stake.

Checkpoint still needs to be strong as that's how the network survived the vulnerability in 0.2. Also ppcoin market is still tiny it's easy to buy a bunch of coins and try controlling via proof-of-stake.

I am not aware of any centrally controlled clock that you are talking about. Could you expand on that?

Yes, if your block is the last selected block you can change 1 bit of the modifier (giving you two choices). It's a slight advantage, not enough to pose immediate threat like in v0.2 protocol.

As I understand it, PPCoin's original POS has a problem that the POS is really POW in disguise: miners could twiddle values fast and do computation to boost their PoS chances. This is a typical problem for PoS systems: "nothing is at stake", holding something doesn't translate into a finite resource you consume as you attempt to mine. Likewise, they could manipulate timestamps to improve their stake effectiveness.

You appear to have fixed this by constraining the values that can be twiddled for POS, so extra twiddling can't give you a gain over other miners. This produces a problem that no active PoS miners may be able to solve a block, so it appears that you've fixed that by using your checkpoint mechanism as a centrally controlled TICK that announces a modification value the PoS miners include in their PoS (CheckStakeModifierCheckpoints()).  So now the speed of PoS searching is ultimately limited by how fast your checkpoint clock 'ticks' by announcing new modifications.   Have I got that right?

[Edit: I pulled out a bunch of conclusions which are probably wrong, in the likely case that I'm wrong above.]

Of course, I could be totally off in space. Reverse engineering behavior from source takes some effort, and I did nothing beyond letting the diff scroll over my screen— I'm mostly only interested in what you're doing academically, and that would better wait for someone to produce a clear writeup of it.

CheckStakeModifierCheckpoints() is just used to make sure the stake modifier computation is deterministic and every node computes to the same value. It's only checked with a few blocks, like the hard checkpoint that bitcoin uses.

It's not actually part of the protocol itself. The reason it was there is because of the complexity of modifier computation there could be bugs that made it non-deterministic. I put it there just so if someone runs into it then I would know a bug exists with the modifier computation.

Each block stores its then-current modifier in its block index. When stake hash is computed, the coin (txout) generating the stake kernel must hash with the modifier that's 9 days into the future of the coin. There is no central clock involved.

Does the QT version also do the proof of stake calculation or do I have to run the daemon?

You can mine both types of blocks with the qt version, I can confirm this still works in 0.3.

Reminder that upgrade to 0.3 is required before March 20th.

Great coin!

I have some difficulty understanding how to install it in Windows. Could someone provide a simple guide for downloading and installing PPCoin in Windows?

I assume a GUI is included?

Reading up on this coin and getting more and more enthusiastic  :P

Yes Qt GUI is included in v0.3.0. For windows you just download ppcoin-0.3.0-win32-setup.exe and run it to install.

Reminder that upgrade to 0.3 is required before March 20th.

Is there going to be a client for windows or mac?

ppcoin-0.3.0-win32-setup.exe is for Windows. mac build is not supported right now.

hi how can we mine for it?

Just use any standard bitcoin miner. Largest pool is http://ppcpool.bitparking.com/pool

2 days left before protocol switch to v0.3. Please upgrade to 0.3.0 as soon as possible if you haven't done so.

This morning  I performed upgrade. Coinotron is using version 0.3.0 now.

Final reminder: less than 1 day before protocol switch. Upgrade as soon as possible to v0.3.0 if you haven't done so. I will be watching the switch on the network tomorrow.

Upgraded.

What is the exact switch over time?

Exact switch time is
Wed 20 Mar 2013 17:20:00 UTC

Is there anything strange about PPC block 38018 to 38040?  Someone generated like 20 blocks of Proof of Stakes in around 10 minutes, using 250/500 PPC as stake.  Is it related to the old vulnerabilities where you can speed up PoS block generation, discovered by Jutarul?   ???  We are suppose to upgrade to the new version by today too.

The upgrade hasn't happened yet. About 14 hours to go.

The new protocol proposed by Sunny should "smoothen out" the effect of putting offline stake online. However, the fact that these rapid block generation cycles occur may not be a problem of the POS scheme itself, it's rather an outcome of non-competitive mining. Obviously, we need a more aggressive POS mining landscape. Maybe the incentives are too low right now?

The fast generation of PoS block is still ongoing, about 55 out of last 60 blocks, and in the amount of 250/500 PPC, so I presume it's the same person.  Is this normal or something that can be safely ignored?  Or in Jutarul's word, this can actually be "good" for PPC, attracting more people (including potential offenders) to generate PoS blocks and thus making the PPC network more secure, and not someone exploiting the vulnerability by hastening the cycle?

This is me. It's a wallet I created by sending approximately 250-500 PPC every 5-10 minutes over the period of a month. The actual amount and time depended on how much PPC I had available. I kept this running on a 24x7 node and it averaged about 150 blocks immature constantly after about 40 days. The machine it was on had a hard drive failure and it's been offline for a week or so. Having recovered the wallet from backup I've fired it up again and I guess it's catching up. For the curious the ppcoind node this runs on uses about 80% CPU idling - I assume that POS stack code takes a bit of CPU while cycling through so many transactions.

Thanks for the details on how the system actually works.  As most of us would not have had enough PPC or the time dedicated to test it empirically.

Hi Chris, when you generate these blocks which version of the client were you running? Is it the latest v0.3 client?

The protocol has been switched over to 0.3 today. The situation should improve gradually. Like Jutarul said, stake participation rate might have also contributed to the problem.

The average age of stake is meant to be eventually closer to 90 days than 30 days. In 0.3 protocol a change was made so that the weighting starts from 0 at the 30-day minimum age. This should help strengthen proof-of-stake protection when difficulty is still low.

How does a newb like me use GUIminer to simply and quickly with minimal typing and -flags start mining PPC? There are no guides in the Wiki, and all other guides are for bitcoin directories....

Is there a way to see how old the coins are in each address and/or how long they've been sitting?

I will be following up with PPC, and I like PPC better than litecoin and ripples now.

so, my favorites are now;  bitcoin, ppcoin, ripple, litecoin; in the same order. :)

This blog was informative,

http://themisescircle.org/blog/2013/06/24/the-proof-of-work-concept/

OK I am screwed I Lost several PPC  14.393888 PPC It seems I never updated the Client before March 20th which I could of sworn that I did.
I am getting a Unexpected Visual C++ Runtime Error twice in a row and than the Client Closes.

Upgrade it then

PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
PANIC: fatal region error detected; run recovery
DB_LOGC->get: LSN: 1384/5911106: read: DB_RUNRECOVERY: Fatal error, run database recovery
DB_TXN->abort: log undo failed for LSN: 1384 5911106: DB_RUNRECOVERY: Fatal error, run database recovery
PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
PANIC: fatal region error detected; run recovery

That appears to be a database error.

I would backup my wallet file (you should have backups already) then move the data directory elsewhere while the program is closed and redownload it all from the beginning then copy the old wallet back into place and restart the client. That's just how I'd approach it.

Maybe someone else in here can give you better advice on repairing the db, I'm not sure if / how that can be done.

Ok well I don't know how to do that either any help from Sunny King would be appreciated seeing in as he is the genius behind PPCoin

Maybe you can just send me the 14+ PPC I had in the Wallet and I will send you the Wallet File Sunny what do you think?

Good luck with that, the last time Sunny King posted anything on bitcointalk other than a 'weekly update' was November 13, 2013.

That's right - two full months ago.