In the last couple of days my qt client on win7 64bit is having problems syncing up. Has the number of full nodes dropped or am I just unlucky? Its also prone to crash since I did a resync.
It seems like an old miner has been running for a few days because it looks like there is an old fork being mined or something and I suspect that is contributing to the recent sync problems which several people other than yourself have reported. Just follow the suggestions I gave a few posts ago.
I've been running 'addnode 206.72.193.148 add' which sometimes helps but it still takes a long time to get any peers. Are there any other known good peers that people can use?
Looking in debug.log I'm getting a lot of the below messages for different peers. Have I ended up on the wrong fork?
2014-11-13 10:08:07 ERROR: AcceptBlockHeader() : extended invalid chain
2014-11-13 10:08:07 ERROR: ProcessBlockHeader() : AcceptBlockHeader FAILED
2014-11-13 10:08:07 Misbehaving: 58.20.250.60:8253 (0 -> 100) BAN THRESHOLD EXCEEDED
On the first issue could XCN pay for full nodes in the way it pays for miners? So say every 2000 blocks it could divide a special reward of XCN to all full nodes which have been up since the last reward. This would keep the number of full nodes from dropping as in other coins like BTC.
That's actually a decent idea but it's more complicated than it sounds, you need to develop a protocol for proving who is a seed node and how long they've been online. You also need a way to ensure that they are actually seeding other nodes with legit data. Then you need to alter the protocol to allow those types of special rewards to be accepted by the network. Not simple at all, but it is a good idea.
Yeah I realize it would be difficult to implement in a way which couldn't be cheated, it was something to think about to avoid the situation most coins have where all the coins go to the miners and none for the nodes or devs. Would people object if the protocol was changed so a percentage of the transaction fee went to a dev fund to pay for code improvements?
This reminds me, the CPU utilization is much higher than my blackcoin wallet even when it's staking. I assume this is either a coding issue or the M7 PoW lacks the 'ease of verification' which sha256 has? This could cause scalability issues later.
I'm unsure what you mean. Cryptonite doesn't use PoS, it's purely PoW. So if you're CPU mining than your CPU usage should be 100%, else you're not taking full advantage of your CPU.
Sorry I should have been clearer, when win64_cryptonite-qt_14091021 is just running as a full node verifying blocks it uses more cpu than my blackcoin client when it stakes, so it's verifying and sometimes creating blocks . I've seen it get upto 25% so my worry was that your M7 PoW was more difficult to verify than the standard bitcoin sha256. It should be possible to run nodes on very low spec machines to encourage as many people as possible to run xcn and leave it running in the background to increase the network security.
On the 2nd issue could we have an install and uninstall option or put all the config setting in a config file in the cryptonite directory rather than filling up the registry.
The registry changes only happen if you're using Windows, I'm not entirely sure why, but the same thing happens with Bitcoin.
If this get popular most people will be running this on windows ( unless we get a smartphone version) and being able to completely uninstall and then re-install could solve some of the sync issues.
