I've upgraded the wallet, it is working fine. Question I have is why you are adding setup for wallet to archive? It makes archive
noticably bigger but question is if anyone is actualy using setup to install MED wallet. Also, judging by file name it seems that
it installs older version of MED wallet (mediterraneancoin-0.8.5.1-win32-setup.exe).
For historical purposes, I'm seeing the following stats as of 10:50pm CST:
Difficulty: 1435.93694438,
Blocks: 8440,
Network speed: 121.805 gh/s
What formula are you using to calculate approximate hashrate? The one I'm using is (diff * 2^32 / 60) and it says that hashrate
was around 100 GHash. Command getmininginfo returns similar number:
16:20:39

getmininginfo
16:20:39

{
"blocks" : 9030,
"currentblocksize" : 1000,
"currentblocktx" : 0,
"difficulty" : 1395.64443691,
"errors" : "",
"networkhashps" : 98762175563,"pooledtx" : 0,
"testnet" : false
}
hi,
regarding the win32 build, the installer is part of the automatic gitian build; there is an error with the icons (from litecoin tree) and that is being fixed, so the next builds will have this aspect corrected.
We choose to go with the gitian builds because it's a very good build process and safe for the end-users since it can be replicated by anyone; the resulting binaries obtained with the gitian build should be exactly the same, bit for bit.
You are right, the resulting archive with the installer is quite large; in the future we might separate the installer from the rest in order to save some space. Also there is an issue with the naming of the installer (it is not being updated correctly during the gitian build), we will have to look into that.
Regarding how networkhashps is calculated, the function calculates the average in the same way as Litecoin does;
basically it does this:
uint256 workDiff = pb->nChainWork - pb0->nChainWork;
int64 timeDiff = maxTime - minTime;
return (boost::int64_t)(
workDiff.getdouble() / timeDiff);
simplifying a lot what the code does, pb is (usually) the current block, and pb0 is the block we want to compare to (its index calculated by the formula 'lookup = pb->nHeight % 2016 + 1' that is the time diff of about 1.4 days in case of constant difficulty ).
also the rpc command for getting the networkhash supports two optional parameters:
getnetworkhashps [blocks] [height]
so one can calculate the networkhashps specifying from which block (height) you want to start the estimate, and then of how many blocks you want to go back.