View Full Version : I just remembered....
Mustard
08-23-2005, 12:21 AM
another *major* reason why I hate boinc..................... four days crunching seti, a large number of workunits crunched, and every single one of them is still *pending*
This pending crap really really really really ticks me off!!!!! To the point as a matter of fact that I'd just quit crunching if boinc was the only thing left to crunch.
PY 222
08-23-2005, 12:25 AM
Thankfully, there are still alot of other projects worth crunching that is better of.
Relax, have a cold one from the fridge. :cheers:
alpha
08-23-2005, 07:08 AM
Why is this a problem? You still get credited.
Mustard
08-23-2005, 11:34 AM
Why? Cause I don't care to crunch for four to seven days before I get credit on something........... I don't have a life... I'm a total stats whore
Why? Cause I don't care to crunch for four to seven days before I get credit on something........... I don't have a life... I'm a total stats whore
Number 1 reason why I dont participate in Boinc projects anymore.
Boinc the worst thing ever to happen to DC:
Fozzie
08-23-2005, 02:18 PM
Lexx you whore you have it bad mate.
The only way i survive is runing multiple projects.
Gotta get the fix. ;)
Mustard
08-23-2005, 02:21 PM
Actually, I'll probably have to eat a little crow on this............ :( Why? Because our local people here work on the gravity waves stuff , and one must support a locally based effort.... at least to some degree.
Combustion
08-23-2005, 02:53 PM
Originally posted by Lexx
another *major* reason why I hate boinc..................... four days crunching seti, a large number of workunits crunched, and every single one of them is still *pending*
You can estimate you pending credits by going to this page:
http://ztb.dyndns.org/pc.php
More pages (All BOINC Projects):
http://cjsoftuk.dyndns.org/boinc/PCM.php?action=acctid&nusers=1
http://cjsoftuk.dyndns.org/boinc/PCM.php
CC
Some pending credits never get turned into points.
I have pending credits over a year old that will never be credited.
Boinc is a lottery.
FoBoT
08-23-2005, 03:16 PM
thankfully i am only a moderate stats whore, so the delay doesn't bother me
much
:drums:
Mustard
08-23-2005, 03:20 PM
Originally posted by FoBoT
thankfully i am only a moderate stats whore, so the delay doesn't bother me
much
:drums:
a *moderate* stats ho?????? :) Either you a ho or you not a ho.......
Can't be just a little bit preggers you know!!!! :rotfl:
Mustard
08-23-2005, 03:24 PM
Originally posted by PCZ
Some pending credits never get turned into points.
I have pending credits over a year old that will never be credited.
Boinc is a lottery.
I wonder why that happens..... Theoretically it never should as the work unit should be re-issued after the time-out period if there weren't 3 results turned in. But then boinc doesn't stand for rocket science or anything of the sort. After the early early beta stage participation, I am totally amazed that it even runs at all. That was really a long long exercise, and painful too.
I suspect that was due to problems early in the project. When LHC ran out of work late last week, my pending credit gradually came down to 0 as the other results came in. Work has returned now btw :) And my clients have for the most part suspended their work on Einstein or Predictor and moved back to LHC seeing as I gave it the lion's share of resources.
My Einstein pending credit on the other hand is up to close on 5000 now and just keeps on growing.
Bok
Mustard
08-23-2005, 04:09 PM
hmmmm..... just tried getting work from LHC, and I keep getting communication deferred messages. :(
As for the einstein, are you getting any granted points in it, or do you just keep accumulating pending only?
I'm getting granted points, just not as fast as my pending is going up :)
Bok
I keep hearing the same complaints over and over again.
Not enough work. takes to long to get granted credit etc.
Well Folks Boinc has been running for quite a while now and it must be obvious that this is par for the course.
The plain truth is that all the projects are underfunded.
They don't have the hardware or the staff to do the job properly
With the possible exception of CP.
Anybody can jump on the Boinc bandwagon with a half arsed underfunded project and the DC community flock to it
The project can't cope and people get upset.
All the time we keep giving time and resources to Boinc it won't get fixed and it surely needs fixing.
Please
If you are involved in a Boinc project and it runs out of work or screws you for credit don't post about it we already know this is going to happen.
alpha
08-24-2005, 04:24 AM
Originally posted by PCZ
The plain truth is that all the projects are underfunded.
They don't have the hardware or the staff to do the job properly
I hear what you are saying and understand why people are frustrated, but I still feel the need to reiterate that this isn't BOINC's fault.
If the BOINC driven projects were as reliable as distributed.net and folding@home, we probably wouldn't be having this discussion (or any of the other BOINC-hate ones).
Just for the record, I'm not a BOINC developer or in any way affiliated with Berkeley. ;)
alpha
It is not just the projects which choose to use the Boinc platform that are at fault.
A lot of what folks hate about boinc was written in as a kneejerk reaction to widespread cheating that went on in SETI Classic.
The lead developer has said in interviews that contributers are not to be trusted.
Although he stopped short of actually saying that given the chance they will all cheat that is what he meant.
Unfortunatey there are quite a few fanboys on the Boinc forums ready to defend
it at all costs.
Anyone who makes negative comments is shot down without a fair hearing.
Anyone who starts crunching a Boinc project today won't see any appreciable credit for at least a week or more and slowly build up large amounts of pending credit.
Some of this pending credit may never actually turn into credit.
If you are unfortunate enough to have your results compared to a couple of idiots running DC on PC's 'overclocked to the MAX' then you get nada because they screwed up the WU.
This isn't your fault why should you suffer.
Also
Some of the projects have very short deadlines for work to be returned.
Seems fair enough but do they reciprocate, do they ell, you have to wait ages for credit.
Some of the projects restrict the amount of work you can download a day, might be OK for a uni processor CPU but try keeping a duallie or quad fed on that starvation diet.
If you run out of work in one project of course the other projects you have set up will fill in the gap.
Nice thought but in reality they will all be screwed except CP.
Now I have nothing against CP but it is a very long term commitment, the WU's take ages to complete and in all probability most never get completed.
I just had a dabble with Boinc to see if anything had changed in the time I've been away from it.
Sadly nothing has.
linux users are still getting screwed over for credit and the projects still run out of work :(
I spent a few minutes reading posts on the predictor forums but had to come away as i was sick of seeing posts in every thread crowing about a stupid wiki.
Angus
08-24-2005, 04:51 PM
Originally posted by PCZ
I just had a dabble with Boinc to see if anything had changed in the time I've been away from it.
Sadly nothing has.
linux users are still getting screwed over for credit and the projects still run out of work :(
I spent a few minutes reading posts on the predictor forums but had to come away as i was sick of seeing posts in every thread crowing about a stupid wiki.
Why does a DC program need a 'wiki' that approaches 800 pages to be understood?
The whole queue management thing is crap. Just give me the ability to DL the queue size I want and let me manage it. If deadlines go past, who cares? Just resend it to someone else - there's lots of people looking for WUs.
Mustard
08-30-2005, 12:24 PM
Well, er uhm, unh, ah...... actually there are a couple of the boinc ones that seem pretty solid to me with no problems to day, einstein, sztaki, and xtremlab. Of course it's a little early on xtremlab as they don't have the validator running, but other wise it keeps on running. No work issues, and the same with the other two.
Seti-II on the other hand seems to be designed as a test to see how much crap a project can hand out and see how many people will stick with it..... Then after reading that interview with David Anderson, he sure appears to be out of touch with reality when it comes to what his system can handle in number of clients. Wants to double it and it can't even handle what they have. :jester:
Predictor I can't get running on my linux boxes.... says I don't have enough space, which is *still* a recurring problem. Too bad as I'd like to do the protein stuff.
So it appears that there are institutions other than Berkeley that are going great guns with Boinc..... but Berkeley (Seti) still isn't.
Powered by vBulletin® Version 4.2.4 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.