Welcome to MilkyWay@home

Posts by Philadelphia

1) Message boards : Number crunching : New WU Length? (Message 4156)
Posted 15 Jul 2008 by Profile Philadelphia
Post:

I'd like cross-project fairness be implemented. For now the range of credits per work done differs on most projects, although most of the projects run more or less smack in the middle. There are just some that seem to worth themselves so scientific unattractive that they give credits beyond reason (That's at least Cosmo and Milkyway, perhaps some of the more obscure as well), others that have such a great confidence in their worthiness that they give sub-standard credits (like WCG and LHC on my puter).

The range of credits per hour and core on my puter is from 10 to 110, my (cheatfree) bench value would be 25, I get about 30 on average from most of the projects. The real low end are some alpha projects, where such stuff is supposed to happen. That's the fun of being ATA ;)


And regarding the own, traditional, points system at WCG: That's a fine example how much they like cross project fairness. They had this old points system implemented long before BOINC started for their own UD cient. Once they joined BOINC they started the alternative Credit system to stay on level with BOINC and not to grant their 7-times too high own points as credits.


I see you're pretty good with cut and paste, I thought some of that sounded familiar so I went back to cosmo and yes, word for word (words I put in bold), lol.

And please don't hijack this persons thread, start up another one regarding credits if you're so inclined.
2) Message boards : Number crunching : New WU Length? (Message 4153)
Posted 15 Jul 2008 by Profile Philadelphia
Post:
Old ones were 2:30, new ones are as well 2:30. Only the for the old ones it was minutes, for the new ones it's hours. So the same her as with niterobin.

What's the nearly same as well is the ridiculous amount of credits that gets tossed after completed ones, on the first one here even 10% more. One third to one quarter should be enough.



Saenger ....Are you the credit Police? What is your problem dude...1st Cosmo now here! Credits here were voted on...every vote counted to the credit we receive(including yours)....I thought credit discussion was done here....always at least one guy that likes to stir it up.Are you employed by DA?Apparently you perfer gestapoism to democracy...Please just go crunch projects with no credit and leave the rest of us in peace.....Grrrrrrrr


You beat me too it, we ran him off at Cosmo so he's here now, lol. Glad he's on SETI.GERMANY and not our team SETI.USA.
3) Message boards : Number crunching : No new WU (Message 3744)
Posted 12 Jun 2008 by Profile Philadelphia
Post:
Yes, WU's are available again :)
4) Message boards : Number crunching : No new WU (Message 3722)
Posted 11 Jun 2008 by Profile Philadelphia
Post:
No WU's in a long time, my 'puter is hungry for WU's.
5) Message boards : Number crunching : Validation anyone (Message 3679)
Posted 7 Jun 2008 by Profile Philadelphia
Post:
I haven't been able to upload for 5 or 6 hours :(
6) Message boards : Number crunching : No Work (Message 3495)
Posted 18 May 2008 by Profile Philadelphia
Post:
just got 20 wu through now


Nice, I just updated but nothing from the project :(

[edit] Just got my WU's too :)
7) Message boards : Number crunching : No Work (Message 3493)
Posted 18 May 2008 by Profile Philadelphia
Post:
M/W's been good the last couple of months or so sending out W/U, hopefully they will have more soon :)
8) Message boards : Number crunching : Best completion times (Message 3482)
Posted 14 May 2008 by Profile Philadelphia
Post:
My E6600 is consistantly running at 300s, I can't get it a weeeeee bit faster to get to 299s to break the 300 barrier :(
9) Message boards : Number crunching : 20 workunit limit (Message 3039)
Posted 5 Apr 2008 by Profile Philadelphia
Post:
thanks everyone for your imput, it was very helpful.
10) Message boards : Number crunching : 20 workunit limit (Message 3022)
Posted 4 Apr 2008 by Profile Philadelphia
Post:
thanks.
11) Message boards : Cafe MilkyWay : Word Link (Message 3014)
Posted 4 Apr 2008 by Profile Philadelphia
Post:
young
12) Message boards : Number crunching : 20 workunit limit (Message 3013)
Posted 4 Apr 2008 by Profile Philadelphia
Post:
thanks everyone :)

are the ones with minues numbers owed crunch time or the other way around. thanks again.
13) Message boards : Number crunching : 20 workunit limit (Message 2990)
Posted 4 Apr 2008 by Profile Philadelphia
Post:
Deadlines should be set so that the science of the project is done. Time critical tasks should have shorter and tighter deadlines than non-time sensitive ones. There is no reason that a BOINC client cannot cope with deadlines as short as 3 hours if the tasks can be done in less than that time. V5.10 should be able to cope with deadlines as short as several minutes. The reason for 3 hours is the default connection interval for older clients is 2.4 hours.

BOINC will eventually drop the LTD of a project with short deadlines low enough so that it is not a candidate for receiving work.

A useful modification to the server (if it is not already in place) would be to block work allocation to hosts that have a connection interval larger than the time to deadline of the task. The connection interval is a promise by the user that the computer will be allowed to attach to projects at least that frequently.

That said, I personally would hope for deadlines slightly longer than 12 hours as I have 6 computers that cannot be attached for 14 hour stretches and I would have to detach those. If the deadlines should be that short, then those computers should not be attached anyway.


Isn't there a way to go into one of the boinc files to change the LTD? It seems to me I read that in a thread a long time ago at SETI.
14) Message boards : Number crunching : new work (4/02) (Message 2987)
Posted 3 Apr 2008 by Profile Philadelphia
Post:
PS- Is it me or are the WU's getting longer?

I have the same feeling, but only a wee bit, some seconds maybe.


the WUs might be a little longer, because the star file is bigger. more stars to calculate the likelihood would increase the time of calculation.

basically, the WU does two parts, the first calculates an integral over the volume of space we're dealing with. this is pretty much fixed -- we can set how fine grained it is, but once thats fixed it should take the same amount of time.

the second part calculates the likelihood of the calculated model being a good match to the data. this is usually pretty quick as it's done in linear time over the stars in the stars file, and not nearly as computationally intensive as calculating the integral. so when we increase the star file you should see the WUs take slightly longer.

this is also the reason that the progress bar is a bit wonky, when it changes speeds that means it's stopped calculating the integral and started to calculate the likelihood.


Thanks Travis.
15) Message boards : Cafe MilkyWay : Word Link (Message 2986)
Posted 3 Apr 2008 by Profile Philadelphia
Post:
Post a word that is similar or related to the previous post/word. Starting this out with......

New
16) Message boards : Number crunching : new work (4/02) (Message 2963)
Posted 2 Apr 2008 by Profile Philadelphia
Post:
Thanks for the new WU's :)

PS- Is it me or are the WU's getting longer?
17) Message boards : Number crunching : More Work !!! Please :) (Message 2962)
Posted 2 Apr 2008 by Profile Philadelphia
Post:
WU's flowing nicely again, thanks :)
18) Message boards : Number crunching : Milestones (Message 2948)
Posted 2 Apr 2008 by Profile Philadelphia
Post:
Nice going everyone :)
19) Message boards : Number crunching : More Work !!! Please :) (Message 2947)
Posted 2 Apr 2008 by Profile Philadelphia
Post:
Out of WU's here :(
20) Message boards : Number crunching : Milestones (Message 2935)
Posted 1 Apr 2008 by Profile Philadelphia
Post:
Nice going everyone, congratulations!


Next 20

©2024 Astroinformatics Group