"setgenerate true" for solo mining was removed in 0.13 - but it might still be useful for Datacoin. Is there a way to bring it back?
The internal miner was removed in
this commit.
Reverting the change will almost certainly lead to false virus detection.
We do not want to see Datacoin-Qt or the installer being quarantined nor do we want to pay protection money to the security suite vendors on a continual basis to whitelist us.
At least qt-i386 doesn't conflict with Windows Def. Can't speak for the others though.
What I found with regard to Microsoft specifically is limited to Microsoft Security Essentials -- and that is the outlier in the bunch. Microsoft Security Essentials tends to quarantine whether or not the internal miner has been removed.
Other suites like Norton, AVG or Avast will happily allow an install after the internal miner code has been removed.
What we found during a test over the past week - is that a 0.8 branch clone with the built-in miner triggers 22 warnings from virustotal whereas a 0.10 clone with the internal miner commit backported sees just 3 warnings.
More work is needed to have a comprehensive data set for how the various other anti-virus software reacts, but the pattern is already clear, speaking to me through the data; when a miner is builtin the masses get told to beware and the software gets removed.
Things like leaving the reference client alone w/ just a getblocktemplate and adapted miner for regtest should be fine.
It will be more maintenance to build in libraries from other projects which see updates more quickly than our own.
However -- everybody loves choices!
So we could maintain two flavors -- one with an expected trigger for experts where the A/V expectations are managed through trust --
and the other where the A/V expectations are managed by neutering the code to hold back a miner.
If we go that route then our website where the general public finds us - should deliver them an executable with great assurances it will be trusted by their security software.