In article <q31pa2pc4cnpml2t7gh4fa2etmb8drctp6@4ax.com>,
 f/fgeorge <ffgeorge@yourplace.com> wrote:

<snip>

The newer version, the newest is 5.4.9, uses the actual time spent as 
a basis for requesting credit. MUCH more accurate, and in your case 
you will probably get more credits per unit using the newer version.

It's not the time spent that determines credit claims by S@h Enhanced, 
but an estimate of the number of calculations performed. If you look at 
the output-file text from one of your result pages you'll see the 
relevant stats.

BOINC clients earlier than v5.2.6 don't know how to report Flop-counts 
to the validator, so their 'old-fashioned' time-&-benchmark claims can 
be considerably higher or lower than intended. There's been talk of 
orphaning the older versions, but not until someone builds a current 
client for Solaris, which is apparently stuck somewhere around v4.4.

Since hosts running old clients are in the minority, though, most 
results from BOINC v5.2.13 and v5.4.9 are granted exactly what they 
claim, regardless of what types of system are 'partnered' in the quorum.

-- 
Odysseus