Are we slowing down?

[ShoeLace]

http://www.seventeenorbust.com/stats/graphs/sysOverall.mhtml

are we slowing down? by my eyes we have been since teh start of april.

or is this maybe a byproduct of more people doing secondpass etc..


thoughts anyone?

[Keroberts1]

there is often a downward slope on the nd of that graph because users who don't have report intermediate blcks checked don't report any progress to the server until the test is completed. This means that when the test is complted it will raise all os the stats fro mwhen the test was assigned ot whe nit was completed. I don't know if I'm explaining this very well, maybe someone else can do it better.

[ShoeLace]

i understood that okay.. but that would imply there was some 200M cEM/s that hasnt reported in say the last 2 weeks.

does that sounds resonable to you? what sort of machine takes 2 weeks or more to complete a test? 1GHz p3 or other non SSE2 machine?

perhaps what i really need to see is number of tests assigned and returned graphed over a period of time.. (i know there are current values for that around..)

for instance the number of pedning tests at " 9:10 pm EST, 18 Apr. 2005" is "5148 tests" and in the last day we've complete "319 tests" at a system work rate of 1.67 G cEMs/sec (maybe i am lookign for a graph of system work rate)


other usefull info on http://www.seventeenorbust.com/stats/
such as 1069 users active in the last 2 weeks.



http://www.seventeenorbust.com/stats/graphs/sys24.mhtml

also anyone have thoughts on the 'spikes' on the weekly graph above. a lot of machines submitting at once? or some other statistcal anomoly?
i can only think of a number of machines doing small-n tests and then stopping.


its not really a problem.. i am just curious

[pixl97]

Its getting close to summer, about time to get out of collage, school, some jobs, etc... people go on vacation and forget about the DC client on there computer, so on and so forth.

Increasing numbers need active recruting.