How could UCB stop it from downloading work? As long as the program is project approved i don't see how they could block it.
I may be wrong but i just don't see how. I know they could block it froms SETI which is run by UCB but I don't see all the others. Especially if the program was better developed and maintained

---

4870 GPU
4870 GPU

This could be where the needed improvements in the Boinc Manager will be done and distributed by outside sources such as Crunch3r and others.

Well, if people developed 3rd party versions of the Boinc Manager in the way you suggest, would Berkeley allow them to download work? Would they be blocked?

You don't need to utilise BOINC Manager at all. Remember BOINC View? The main issue would be the BOINC client - but there are so many self builds out there in any case that unless one of them was really frackin with the projects I doubt they would even consider banning specific clients, but then the person would just need to alter the name of their build and away they'd go again. LOL sounds like a virus.....

I am not saying that it could not be done ... but just as a practical matter you would need 10-20 really good people to make a go of it ... and they would have to be really into it ... it is not just that BOINC is so big, but that there is so much that is messed up ...

Maybe its time to outsource boinc to India. :P

I notice now that the current wu's are requesting 11 credits for myself. The ones before this last cut were 7-8 credits. The de_s222 are running the same time with the .20 app. So that means we are now doing even more work for even less credits. So this was a 2-part credit cut without saying that the second part would happen. It would be nice if someone would have the guts to own up to these things before they happened.

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

I notice now that the current wu's are requesting 11 credits for myself. The ones before this last cut were 7-8 credits. The de_s222 are running the same time with the .20 app. So that means we are now doing even more work for even less credits. So this was a 2-part credit cut without saying that the second part would happen. It would be nice if someone would have the guts to own up to these things before they happened.

Don't feel bad... GPU Grid's tasks used to take 6-6:30 with an occasional outlier at 7 some hours ... now the outliers are at 6:30 and many are taking up to 9 hours ... GDF gave a mushy reply when I pointed this out ... so credit deflation marches on ...

Yes -- I pointed that out over there as well.

Don't feel bad... GPU Grid's tasks used to take 6-6:30 with an occasional outlier at 7 some hours ... now the outliers are at 6:30 and many are taking up to 9 hours ... GDF gave a mushy reply when I pointed this out ... so credit deflation marches on ...

---

Well I see it is not only me that is currently going what the flying F--k.
I got two new cards expecting my credits to go to 220,000-250,000 a day...
Yeah right...
It seems that my two 4850's are going to level off at 80-90K a day and my 4870 is going to be somewhere around 40-45K a day. What the hell, it should be double that...

What the frak happened? If the credits being granted fall in line with the math for the work what is the big deal?

Maybe ALL PROJECTS NEED TO BE TELLING SETI THAT THEY DO NOT RUN OTHER PROJECTS.

All projects have a standard to follow for granting credit. STICK THE HELL TO IT. This is damned ridiculous. Just because people are getting faster computers, and even faster video cards that can crunch does not mean that the slower people can tell the faster ones that they should be getting less credit for doing more work.

If this shit does not change I am done.

You want an idea to base credit on.
Ok fine... the simplist way possible...

You give out a 5000 sec long work unit.
The cruncher gets credit for doing a 5000 second work unit.
The actual time it took the crunchers system to do it does not matter, they still did a 5000 second work unit correct?
The computer that did it in what ever time still gave a valid calculation or response correct? Why should it matter if it was CPU or GPU, or opti-app based?
The project sets the work unit lengths. The projects decide on a value of credit per work unit second.

Is this not the simplest solution for all projects?
They all agree on a baseline of credit for a per work unit second.
So then the projects would be very inclined to create optimized apps. They would grant credit on a work unit length based standard.

Seems pretty damned straight forward and simple to me...

---

>You give out a 5000 sec long work unit
Try CPDN - 4,000 hour WUs available there.

---

Cheers,

PeterV

.

>You give out a 5000 sec long work unit
Try CPDN - 4,000 hour WUs available there.

Oh I'm still trying them after all these years. Someone has to support the work in Climate Change after all ;)

---

Indeed, if they gave out longer work units the time spent on each by systems relative to each other wouldn't change. The granulity of mutations would change though, but I'm not sure what effect that would have. (the validator would have to let more finished WUs build up on average to generate the larger WUs, but that shouldn't lead to problems as long as it can handle it)

What I'd rather see is them using more advanced methods to analyse the data, though. Precision has been tweaked as much as possible to ensure errors for the current model are minimized - which means the model itself is now the bottleneck. I wonder how much of a difference it would make to look at strips of the sky as strips of a globe instead of a cylinder?

Well I see it is not only me that is currently going what the flying F--k.
I got two new cards expecting my credits to go to 220,000-250,000 a day...
Yeah right...
It seems that my two 4850's are going to level off at 80-90K a day and my 4870 is going to be somewhere around 40-45K a day. What the hell, it should be double that...

What the frak happened? If the credits being granted fall in line with the math for the work what is the big deal?

Maybe ALL PROJECTS NEED TO BE TELLING SETI THAT THEY DO NOT RUN OTHER PROJECTS.

All projects have a standard to follow for granting credit. STICK THE HELL TO IT. This is damned ridiculous. Just because people are getting faster computers, and even faster video cards that can crunch does not mean that the slower people can tell the faster ones that they should be getting less credit for doing more work.

If this shit does not change I am done.

You want an idea to base credit on.
Ok fine... the simplist way possible...

You give out a 5000 sec long work unit.
The cruncher gets credit for doing a 5000 second work unit.
The actual time it took the crunchers system to do it does not matter, they still did a 5000 second work unit correct?
The computer that did it in what ever time still gave a valid calculation or response correct? Why should it matter if it was CPU or GPU, or opti-app based?
The project sets the work unit lengths. The projects decide on a value of credit per work unit second.

Is this not the simplest solution for all projects?
They all agree on a baseline of credit for a per work unit second.
So then the projects would be very inclined to create optimized apps. They would grant credit on a work unit length based standard.

Seems pretty damned straight forward and simple to me...

It's better to stop crunching MW an move to Collatz.

---

So move then, Beserk, and stop clogging up this board whingeing about it.

---

Cheers,

PeterV

.

It's better to stop crunching MW an move to Collatz.

NO NO, Not Collatz, Not My House, You Guyz are doing just Great here ... :P

>You give out a 5000 sec long work unit
Try CPDN - 4,000 hour WUs available there.

I also run CPDN. Maybe you missed my stats sig.
That was not the point.
The point is if you run a X second work unit, you should get credit for running a X second work unit no matter what project supported application was used to do it.

---

So move then, Beserk, and stop clogging up this board whingeing about it.

i moved on days ago.

---

So move then, Beserk, and stop clogging up this board whingeing about it.

i moved on days ago.

So hurry back and keep the faith :P

---

So move then, Beserk, and stop clogging up this board whingeing about it.

i moved on days ago.

And get even LESS credit. Talk about cutting your nose off to spite you face....go figure.

So move then, Beserk, and stop clogging up this board whingeing about it.

i moved on days ago.

And get even LESS credit. Talk about cutting your nose off to spite you face....go figure.

I prefer to run Rosetta which didn't use to be anywhere near this in credits is now closing in with all of the reductions here but still a few times less (so far). Some times less is better.

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

So move then, Beserk, and stop clogging up this board whingeing about it.

i moved on days ago.

And get even LESS credit. Talk about cutting your nose off to spite you face....go figure.

I think it is more of why tolerate constantly changing rules...
I have already moved one of my machines to collatz as well.
I could care less about credit. However if I am going to donate time, I would like something for it. Constantly having credit standards lowered because of people bitching and moaning that you are getting more credit, but ignore the fact that you are getting more work units done.
The project admins that continuously collapse to these people's constant bitching pisses me off even more and makes me less inclined to help their project.

---

By utilising CPs latest ap, you are getting MORE credit than you were before.

I still say credit parity across projects for standard apps is a valid goal for the BOINC devs. Yeah it sucked a bit when the credit was reduced, but then within days CP released another app and bingo actual credit per unit time went up!

You really are crying about spilt milk.