There are some outstanding clients with higher benchmarks, the old problem, so any fixed credits per WU or some other solution without a higher quorum planned?

---

How about to allow only >5.5.1, this will reduce some of 5.5.0 unofficial clients?

How about to allow only >5.5.1, this will reduce some of 5.5.0 unofficial clients?

I wouldn’t much like that: I’m running (stock) v5.4.9, and because of the problems I’ve had with later versions on another Mac, I’m very reluctant to upgrade this one. It makes appropriate claims as requested by the SETI@home apps, for example, as do all versions since 5.2.6 or thereabout.

I would rather see the work replicated, which serves as a deterrent to a variety of abuses. Fixed credit might be a simpler solution at the back end—I guess they have their reasons for the quorum of one— which I’d be perfectly happy with as long as it’s roughly in line with the nominal benchmarks or cross-project average.

Unwanted specific BOINC versions can be excluded from the scheduler, by adding this code to the scheduler request.

---

Jord.

The BOINC FAQ Service.

How about to allow only >5.5.1, this will reduce some of 5.5.0 unofficial clients?

I wouldn’t much like that: I’m running (stock) v5.4.9, and because of the problems I’ve had with later versions on another Mac, I’m very reluctant to upgrade this one. It makes appropriate claims as requested by the SETI@home apps, for example, as do all versions since 5.2.6 or thereabout.

I would rather see the work replicated, which serves as a deterrent to a variety of abuses. Fixed credit might be a simpler solution at the back end—I guess they have their reasons for the quorum of one— which I’d be perfectly happy with as long as it’s roughly in line with the nominal benchmarks or cross-project average.

The reason we have a min quorum of 1 is that our search method is highly resistent to failures/bad results, so theres really not much reason to replicate work. We're also very alpha at the moment so we're working on a simple validator and figure out some credit strategies that will hopefully make everyone happy. We might bump the quorum up to 2 in the future, but right now it's more useful to us to have as many results back as possible (good or not).

How about to allow only >5.5.1, this will reduce some of 5.5.0 unofficial clients?

I wouldn’t much like that: I’m running (stock) v5.4.9, and because of the problems I’ve had with later versions on another Mac, I’m very reluctant to upgrade this one. It makes appropriate claims as requested by the SETI@home apps, for example, as do all versions since 5.2.6 or thereabout.

I would rather see the work replicated, which serves as a deterrent to a variety of abuses. Fixed credit might be a simpler solution at the back end—I guess they have their reasons for the quorum of one— which I’d be perfectly happy with as long as it’s roughly in line with the nominal benchmarks or cross-project average.

The reason we have a min quorum of 1 is that our search method is highly resistent to failures/bad results, so theres really not much reason to replicate work. We're also very alpha at the moment so we're working on a simple validator and figure out some credit strategies that will hopefully make everyone happy. We might bump the quorum up to 2 in the future, but right now it's more useful to us to have as many results back as possible (good or not).

Stay at quorum of 1, its fast and not wasted of time, all other solutions are welcome!

How about to allow only >5.5.1, this will reduce some of 5.5.0 unofficial clients?

I wouldn’t much like that: I’m running (stock) v5.4.9, and because of the problems I’ve had with later versions on another Mac, I’m very reluctant to upgrade this one. It makes appropriate claims as requested by the SETI@home apps, for example, as do all versions since 5.2.6 or thereabout.

I would rather see the work replicated, which serves as a deterrent to a variety of abuses. Fixed credit might be a simpler solution at the back end—I guess they have their reasons for the quorum of one— which I’d be perfectly happy with as long as it’s roughly in line with the nominal benchmarks or cross-project average.

The reason we have a min quorum of 1 is that our search method is highly resistent to failures/bad results, so theres really not much reason to replicate work. We're also very alpha at the moment so we're working on a simple validator and figure out some credit strategies that will hopefully make everyone happy. We might bump the quorum up to 2 in the future, but right now it's more useful to us to have as many results back as possible (good or not).

Stay at quorum of 1, its fast and not wasted of time, all other solutions are welcome!

I'm seeing some 5.5.0 versions claiming twice the credit on the same amount of time processed on some computers. With a quorum of 1 and since the WUs all last about the same amount, my opinion would be for a fixed credit. As this issue is repeated on every project, I will won't make any more posts on this.

---

Stay at quorum of 1, its fast and not wasted of time, all other solutions are welcome!

+1!

An increase in quorum = 50% reduction in crunching speed. No one wants that, and should be avoided unless the science absolutely requires it.

---

There are some funny top computers with 50-100k benchmarks, oh guys ^^, I hope that fixed credits coming soon!

This computer: 2202 is running 1000+ sec. and getting an avg of 2.25 credits.

This computer: 2206 is running 1000+ sec. and getting an avg of 2.35 credits.

This computer: 2195 is running 500+ sec. and getting an avg. of 2.87 credits.

This computer: 2198 is running 480+ sec. and getting an avg. of 2.67 credits.

This computer: 2199 is running 660+ sec. and getting an avg. of 2.67 credits.

Computers 2195, 2199, 2198 are all core2duo, 2 e6600 quads and one e6750
Computers 2202, 2206 are P4's

The core2's are running Vista and the P4' are running XP

Note the discrepancy between times, credits, cpu's, if credit is assigned by cpu time invested(work accomplished) the P4's are really low. If credits are per WU then why are there any discrepancy?

i just ask me if those fake clients do send bogus results and have effect to the science or do they just claim more credits.

if its just credits then let the kids play for the "highscore" its their energy they waste.

if it have effect to the science then the guys from the project should take action asap.

---