Thanks :)

In the particular occasion i needed one..... normally i drink American size mugs (of Italian caffeinated strength in porcelain, not paper cup)..... they hated me when i made the coffee in the US office..... and I was always in very early to prevent the ditch-water servings (here comes trouble)

the HPF2 quorum is still 19 but the run times have been extended by a factor of 1.6

Just to explain to those that still suffer from their machine generating inflated benchmarks and thru no fault of their own end up claiming much higher than plausible for a given piece of hardware: The same rules of outlying apply to HPF2. If anyone claim is out-with certain variable bandwidths it will be excluded from the 'mean' calculation and credit granted from the remaining 'mean'. If extreme (hi) outlier, half of the credit is granted.

To the one recognising the below record:

la134_ 00051_ 14-- Valid 04/03/2007 03:07:46 04/03/2007 12:31:32 3.27 222.7 / 76.8

The average of the other 14 was 76.8, which is what was granted to all 15. The 222.7 or 68.1 per hour of computation on a single core is considered to be in the no-fly zone.

Recommend to verify the benchmark is within norm or obtain a stock copy for the relevant version of OS at Berkeley or here at WCG: http://www.worldcommunitygrid.org/reg/ms/viewDownloadBoinc.do?newMember=false

The 5.8.15 or better client for Linux is confirmed to have benchmark standards close to being on par with the windows version.

Hi Fullsoft,
You understand correctly. Each work unit of HPF2 is run with 20 small variations. A result that is very different from most other results is considered an error. So there is NO duplication.

The forum server is secured. However, we allow members to link to arbitrary URLs in their messages. This is standard for almost all bulletin boards.

Lawrence