That makes sense but how do you do that? If you release the source and binaries can't somebody else set up a node and have others connect to him, thus starting to mine? And if you release the source without the genesis block, when you launch don't you have to add it in and thus recompile the source / make new binaries?

Hello sunny,

You do realize what is going on here right? I've been trying to mine this for a few hours and have put a few things together. Consider:

1) I can't connect to the node you posted. Whenever I do connect, I immediately get disconnected. Ok, maybe there are too many connections (wow, very popular coin!)
2) primer- posts a "super node", 144.76.139.178:21316. I add it to my node list and the daemon immediately starts downloading blocks. Looks like his node is great!
3) Yet, if there are so many connections (CogPrimus posted that he had 997 peers), where are all the peers? I am only connected to the "super node" and maybe sometimes one other node.
4) Yet here are a bunch of us posting hashrates and clearly mining.
Conclusion: primer-'s super node is not well-behaved: it isn't giving us each other's peer info.

Further:

5) All of us are complaining about stale blocks. I've generated 5 or 6 blocks yet each one except the first has been a stale block.
6) Yet the blocks aren't being generated *that* quickly. Only one every 10-40 seconds. Someone is putting a lot of computing power here, which is fine, that is how PoW works, but the hash power isn't as if 997 separate parties were immediately interested in the coin! It is extremely unlikely that the stale would all be naturally orphaned.
7) Also, every once in awhile I get a burst of 5-6 blocks all at once.
Keep that in mind and take into account that just about everyone is connected to that "super node" only ("144.76.139.178:21316")

Conclusion? Someone is DOSing ShinyCoin!

Step 1: Flood the legitimate seed node with thousands of connections so that nobody can connect to it.
Step 2: Post a helpful "super node" for everybody to connect to.
Step 3: Don't advertise peers from the super node.
Step 4: When anyone sends you a block they mine, discard it.
Step 5: When you mine a block, send it to the seed node and propagate it to the network.
Step 5b: If someone mines a few blocks in a row, keep mining until you have enough ready to overpower them, then propagate yours all at once.
Result: You are the only one generating any coins!

At some point you can then stop DOSing the seed node, the network carries on, and you have a nice few hundred thousand shiny coins to play with if anything comes of the alt coin.

I strongly recommend addressing this by re-launching the coin, with multiple nodes, and a way to protect against this kind of attack.

Cheers

Rofl so literally the way to fix the broken pos was to change


to


And that took days to figure out?

Pro tip,
should have been
and that would also have worked.

Question though, if that was the fix (as this is the only significant change since April 27th), why didn't this commit fix it?

(I'm assuming the fix is up on this repository - is there something else to it?)

What source was this compiled from?

That wallet changes the master key for the centralized checkpointing. Essentially it gives control of the network to somebody other than whoever currently has control over the network.

Bump.

warren, a LiteCoin dev, said of Peercoin:


Why are broadcast checkpoints so indefensible? It would seem to make sense to have a protocol-level mechanism whereby the nodes agree not to go back past a certain number of blocks (say 10 or 20 blocks). What about the mechanism is flawed? It sounds like it's able to keep a network stable without requiring as much hashing power.

I don't follow. Which two decimal places? There's two fewer decimal places in PeerCoin than in BitCoin. CENT has four decimal places. So if you stake 1 cent you can get 1 hundredth of a cent, but if you stake 1 hundredth of a cent you can still get 1 ten-thousandth of a cent. It seems arbitrary.

I've gotten interested in PeerCoin recently and have been poking around the source code. I've found something unusual...

June 19th, 2012:

July 22nd, 2012 to today:

As of today, COIN is still 1 million in PeerCoin, while it's still 100 million in BitCoin. What gives? Some unit tests for PeerCoin even fail because of this (e.g. FormatMoney(COIN/10000000) predictably returns "0.00" instead of "0.0000001") . I haven't seen mention anywhere of PeerCoins only being divisible by 6 decimal places. The commit which changed this also sheds no insight.

Did Sunny decide to redefine 1 PeerCoin to be 1 million "satoshis" instead of 100 million "satoshis" at some point? If so, why?

EDIT: It looks like the coin was released in August, so the change was made before the coin was released. I guess PeerCoins are only divisible by 6 decimal places? What's the rationale behind this decision?

As I thought.

It would help. It means the network would verify that the money can only be sent to one of two addresses. A third-party would have no interest in hacking the keys because they still wouldn't be able to send the outputs to their wallet. They would have to hack the keys for the transaction and the keys for one of the destination wallets, at least, and the latter can be cold wallets.

Is the following scenario possible?

I want to send money to a script. This script will require knowing a private key to spend, as usual, but the script is only valid if the money is being sent to one of two addresses. That is, if the transaction has more than one output, it is invalid. If the transaction's output is a standard script to address A or to address B then it is valid. If the transaction output value is any less than x, it is invalid.

In other words, it seems the scripts serve to validate that you can spend the output, but then the output can be spent however the claimant(s) want. Is there any way to limit how the outputs are spent? All this without requiring a trusted third party of course.