Pages:
Author

Topic: [ANN][CDC] CLOUDCOIN REVIVAL. PoW/PoS. ** LATEST MARCH 2015 OPENSSL FIX ** - page 14. (Read 41371 times)

full member
Activity: 476
Merit: 100
sr. member
Activity: 322
Merit: 250
how can anyone deposit in to bter it saids cdc coin disabled when the fck are they goin to enable deposits again
sr. member
Activity: 275
Merit: 250
it was from getmininginfo and difficulty was up to 1.7 or thereabouts...
legendary
Activity: 2268
Merit: 1092
Someone is throwing over 170 MHs at it...

Where are you getting that figure from? If it's the "networkhashps" value in getmininginfo, that won't be correct, because it doesn't account for PoS blocks being generated separately to the PoW network.
legendary
Activity: 2268
Merit: 1092
Something is WRONG with CDC. I think there are multiple forks. Because wallet syncs for 5 mins, then it shows "out of sync", then it is 50 blocks behind and so on....No deposits showed on bter .  Which one is the official chain??

There are two issues with CDC that I'm going to address in this reply. I'm going to key out some possible solutions in the hope of provoking further discussion.

As mentioned previously, CDC places a lot of (I would say almost extreme) weight on PoS blocks. Around 10 hours ago I saw a massive 286 block correction on the official nodes. Someone generating PoS blocks managed to override/reverse nearly 3 hours worth of blocks minted on the network. Within a few seconds the height dropped from 284909 to 284623, and 14325.14 CDC disappeared from the money supply. That sort of thing is unprecedented AFAIK, and to be honest I didn't think it was even possible to correct that large a number of blocks or time period.

Problem 1: Chain can be influenced by someone holding a large balance who locks (or keeps closed) their wallet, then unlocks it. PoS blocks can be generated very fast by a client, sometimes every few seconds. The huge reversal was started by a client that generated 33 PoS blocks in 4 minutes. In addition, they seem to have started minting PoS at a lower height than the rest of the network, so their fork extended from a lower point in the chain, which is why the network reversal was ultimately so large.

A possible solution is to limit the allowable correction to a maximum number of blocks so such massive reversals are impossible. The limit does need to be chosen carefully, because if it's too small it will cause everyone to continue on their own private fork if there's some issue that causes them to be more than X blocks ahead (their client will refuse to snap back to the main chain)

Problem 2: Miners are sometimes dropping off the main chain, and need to resync. This has happened to me also. I've been struggling to understand why, and the reason hit me 30 mins ago: the client refuses to allow you to mine until you have at least one peer connected, but it doesn't actually care about whether that peer is properly synced. I think what is happening is that the first peer a client is connecting to is on the old chain, so the client starts allowing PoS and PoW mining, which has the potential to create a private fork once you start minting blocks. It won't happen every time, but there's enough peers sitting on the old chain that it's likely to happen at some point.

example:

1. new chain is at height 100
2. old chain is at height 20
3. your client is at height 75 when you start it
4. your client connects to peer on old chain, client permits mining even though new and old chain will never exchange/agree upon minted blocks
5. if you mint a block before connecting to a peer on the new chain, it will be minted at height 76 instead of height 101; you're now on a private fork

I think that segregating the network to ignore the old chain will fix the second problem, and will actually go some way to fixing the first (remember the client that did the huge reversal started minting PoS at a lower height). I am working on that code change right now.
full member
Activity: 476
Merit: 100
Release version 1.3 Please

1.3 is not going to be released any time soon, it is simply a future version number used on a dev/testing node. At this point there is really no difference to 1.2 (although there are some planned changes soon, to get away from the old v1.0 chain)

1.2 is the current release.


1.3 Can be released as a development version

1.2 is a can not use version
sr. member
Activity: 275
Merit: 250
Someone is throwing over 170 MHs at it...
sr. member
Activity: 275
Merit: 250
I fucking doubt it. Weeks mining down the fucking tubes...
legendary
Activity: 1232
Merit: 1000
I have 7 connections on my client. I hope I am on the right chain.
sr. member
Activity: 275
Merit: 250
No kidding, I lost another 50K. Someone has re-written the blockchain again. That's 230k lost due to this shit....
sr. member
Activity: 444
Merit: 250
Life is a bitch, get used to it...
Something is WRONG with CDC. I think there are multiple forks. Because wallet syncs for 5 mins, then it shows "out of sync", then it is 50 blocks behind and so on....No deposits showed on bter .  Which one is the official chain??
legendary
Activity: 2268
Merit: 1092
Release version 1.3 Please

1.3 is not going to be released any time soon, it is simply a future version number used on a dev/testing node. At this point there is really no difference to 1.2 (although there are some planned changes soon, to get away from the old v1.0 chain)

1.2 is the current release.
full member
Activity: 476
Merit: 100
Release version 1.3 Please
legendary
Activity: 2268
Merit: 1092
Good news! I have heard back from bter, and they are currently updating the daemon. Hopefully deposits will be re-enabled soon.

Trading is still possible if you want to buy CDC, or sell CDC you have previously deposited to bter. I have a few open buy orders in the book right now myself Smiley
legendary
Activity: 2268
Merit: 1092
V1.3?

Ask the author to publish

That's my node. Have to test new code somehow. Smiley

Now the node cannot update, the node in front of all the old nodes provide

Sorry, I don't understand this...
full member
Activity: 476
Merit: 100
    {
        "addr" : "203.20.114.252:63341",
        "services" : "00000001",
        "lastsend" : 1397209465,
        "lastrecv" : 1397209458,
        "conntime" : 1397209449,
        "version" : 60006,
        "subver" : "/Satoshi:1.3.0/",
        "inbound" : true,
        "releasetime" : 0,
        "startingheight" : 284185,
        "banscore" : 0
    },


V1.3?

Ask the author to publish

Now the node cannot update, the node in front of all the old nodes provide
legendary
Activity: 1232
Merit: 1000
I deleted everything but my wallet.dat. Downloaded the blockchain again. It seems that I only mined 9000 CDC on the official chain. So lame I thought I had over 100k CDC. ohh well at least I am fully synced now.
legendary
Activity: 1148
Merit: 1000
legendary
Activity: 2268
Merit: 1092
Do you have the 2 seed nodes? I redownloaded the blockchain and lost 180k and now stuck at 282136. Not sure I could have been on a private fork since I have been running the client using only the new blockchain since you posted it...

Have you tried quitting and restarting? I've found in my testing that sometimes the client seems to become fixated with peers on the older chain, which means the sync will never advance. Seems it's a bit like pot luck when starting up; occasionally it takes a few tries. (Deleting peers.dat may also help - it will re-add the two seed nodes at next startup). As more v1.0 clients upgrade to the new version (and others join for the first time with v1.1/v1.2) this should become less of an issue. It is annoying - if it does continue I can always add code which will deliberately ignore the older clients. I'm trying not to segregate the network (since v1.0 clients can still use the newer chain) but it may have to come to that.
sr. member
Activity: 275
Merit: 250
startingheight  are not the same, the subver is not the same, this is the reason why all Orphan block?


http://cdc.okaypool.com/index.php?page=statistics&action=blocks

There are two forks of CDC. That's why there are different starting heights.

228898+ is the old chain which is stuck (PoS only, no PoW). Only older clients v1.0 (/Satoshi:0.7.2/) will be on this chain, but newer clients can use it to partially sync.

282957+ is the newer chain, which all clients can use: v1.0 (original), v1.1, v1.2.

CDC is hard to mine because PoS fights with PoW; a single PoS block can orphan the end of the chain that contains PoW blocks. Mostly this is just one or two blocks, but sometimes it can be 10+ blocks that are orphaned and replaced with the PoS block. CDC is not the only coin which is PoS heavy. Smiley

Do you have the 2 seed nodes? I redownloaded the blockchain and lost 180k and now stuck at 282136. Not sure I could have been on a private fork since I have been running the client using only the new blockchain since you posted it...
Pages:
Jump to: