I've had very poor performance with r24 built from git on linux. Lots of low difficulty shares and nonce issues concerning already being sent.
[2015-01-05 00:20:18] nonce 6537ae00 was already sent 2 seconds ago
[2015-01-05 00:20:26] nonce 47dd5f01 was already sent 2 seconds ago
I was only getting about 8 Mh/s effective hashrate at the pool with r24 as opposed to 10.5 Mh/s with r23. I'll be sticking with r23 for now.
Some days ago I added 2 nonce sending in quark, but it had some bugs (double sending) and I reverted the changeset.
Make sure that you build from the github version that is tagged with the "release 24" changeset. Commits before the tagged release could have sideeffects. And commits after are not tested properly and could have errors.
Yesterday I fixed a bug when running benchmarkmode (Run the benchmark for a while and it starts to print 100 times as much text to the commandline). Seems to work, but untested on the pool.
sp - i was mining on yaamp earlier and the site went down ... it is still down as far as i can see as the proxy wont reconnect to it ...
one thing i noticed was that ccminer retries for only 3 or 4 attempts then exits ... can this be changed? ...
i ask this due to teh many times a pool may disconnect for a few minutes at a time then reconnect ... if the miner is still trying ( which happens with much of the very earlyr versions ) then it would continue to mine ... but as it stands - if yaamp were to come back on only 3 minutes after it disconnects - then ccminer has already exited and wont reconnect ...
a setting? or an inherent thing programmed into ccminer? can i change it? ...
tanx mate ...
#crysx