... maybe uplinger would want to double the science weight from 1 to 2 now that there is the rich stockings (no not talking about socks), so that those running on mix have a slightly higher chance with the least common denominating randomizer :P

Sorry, but I think it is the way it is... not a democracy. Until the WU supply is secured for a longer period not seeing the weight/resource share is going to be lifted, nor the move from active intermittent to full active, then see the supply dry up within days and 1000 more breaking out in deep sobs. There is no guarantee and as it is, DDDT2 tasks get gobbled up so quick that the scientists are having all they can handle and that is to me very satisfactory.... they don't even have time to even work on FLU2 which demonstrates the difficulty of getting it perfect, and perfect is required from the results.

Fact: Just quickly flipped my selection to DDDT2 only and got 16, then flipped back to the mix of 4 sciences. They run incredibly quick on my Linux 64 bit client compared to my W7-32 bit duo... 3x faster almost... the stock is worth 0.4 days for 4 cores.

I have a dual core set at 5 days each in the buffer. It is currently running in high priority because it has 31 days of WU in the buffer. All the WU are estimated at about 8 hours although a few are finishing in the 2-3 hour range so I have aborted some since I am afraid they will not all finish in time.

It's when the total queue is so that the client computes that the last ones don't make it. Then HP mode indicates Earliest Deadline First (EDF) order and does away with any niceties about cross-project switching... it will stop fetching work until relieved, which can take until the very last task if the average of the remainder is no shorter.

Here the 7.5 day queue emptied in 5 days (CEP2 is consistently overestimated). The average DDDT2 task per WCG stats today processed in 2.75 hours. The 7 day mean is 2.97 hours. BOINC works by the rule that if a job takes longer than originally estimated, all the remaining queued jobs will be expected to run longer too and that move to increase TTC projections is much faster than going down again when a jobs process faster than estimated. Thus, you can see a client move in and out of a HP state regularly when large caches are maintained. 8 days cache only needs a few jobs running 25% longer to set off all alarm bells. If you happen to have a 6 hour power out that also goes into the equation... so keep it sane or become or sufferer of BOINCitis... not easily curable.

@adrianxw - i am glad you are crunching WU's for any project to be honest. anyone crunching jobs is fine by me no matter what they choose they are doing something with there computer

personaly i have concentrated on this project cause i like all the sub projects running here.



i dont understand what you mean by WCG is the important thing and science is a secondary..

science is the reason most of us crunch here... I will admit alot of us do like badge hunting and the like but there are also enough users who crunch ALL projects or crunch a specific project only and that is there choice.

WCG dose balance the science to make sure each project gets a fair share dont ask me how.

sure we have difference in opinion . Yes I asked what your thoughts were and they are important to me. cause again i think the projects here are important (i know a few people with AIDS) and I myself have had Cancer although we hope its gone now.

anytime a person says they have problems with WCG or Distributed computing or whatever i try to figure out why.. Main reason is so i can compare there questions/problems with my thoughts and you never know i might decide to swap to a different project (i have run alot in the past) heck i started with SETI before BOINC existed.

i am curently working for my DDDT-2 badge and i am glad i can get work units for it right now.. but besides that if you notice the badges you will see i work alot of difrent projects. And if one sparks my imagination i will work it for 1+cpu year maybe 2+ cpu years. If one dosnt spark my intrest i will crunch it a bit and go to bronze or silver badge (14-45 days with 1 cpu) and decide if i like the project or not later.

I like the badges myself it. and I like this project. and of corse i want to know why you have problems cause some times things are an easy fix and if i can convince you or anyone else to devote a little more time to DC or especially projects i am interested in it makes me happy.

thanks for crunching and i hope you continue no matter what project you choose