Boinc 5.10.7

[sierra_bound]

I think the problem lies in those default preference settings which are not the same as the ones you probably use in your WCG device profile. Some people may have been inadvertently saving those settings in the BOINC manager. That will really screw up your machine.

[njkid32]

After reading meshmesh's post in the other thread I went and checked. I had tried clearing the settings and it still didnt fix anything.

[sierra_bound]

I'm going to keep running the new version on a few machines, as long as there are no more crashes. Just curious to see if it has any positive effect on productivity. I'm stubborn.

[mcflurry4321]

I agree, I'm stubborn too. There is no way this release is that bad! Plus I like the pretty progress bars

[sierra_bound]

As a few people have suggested, this BOINC version may be a little more demanding on your machine. So if you already have a high overclock, especially with a quad-core, you may experience more problems. The only machine that hasn't given me any problems with this new version is an older dual-core AMD system that I will probably retire in the near future. It's not heavily overclocked.

[littleowl]

Thanks SB. I will get this downloaded and placed on my systems soon.

[Ashraf]

Well, I just noticed my CPU Usage History, the line in the grid looks funny. And I lost 5K points...

I switched back to 5.4.11, problem solved.

[meshmesh]

I don't think the use of new preferences feature is the reason behind the crashs. It is confusing and unnecessary, but should not result in this instability.

On the other hand, Boinc developers have decided to use a third party open source set of controls called wxWidgets rather than rely on the MS built in native library. They did this because they needed a common set of routines that are OS independent.

This new library is now part of boinmgr.exe but it does not involve boinc.exe
It is used to display list of projects, tasks, etc..
It is also used to show the Preferences checkboxes and controls.

So I wonder, whether Boinc client service would start from a clean boot and run without problem if left alone. Could it be the firing up of boinc manager that triggers the problems? Afterall, Boinc client hasn;t been significantly modified in this release. Boinc manager on the other hand has been changed a lot.

I changed one machine back to the older version that kept crashing every 5-10min and its been totally fine ever since..

If possible, can someone test on a machine that has been experiencing many repeatable problems:
install 5.10 release as a service
reboot windows.
Boinc service will run in the background
do not open boinc manager at all
use boincview only to check on the machine

Is Boinc running properly? Do the errors still occur?

Would be very interesting. maybe we can still use the new Boinc release after all.

[sierra_bound]

All I know is that after I cleared all the preferences, the Explorer crash problem went away on one machine. As I mentioned earlier, I think heavily overclocked systems are the most likely to experience problems with this new BOINC version. When you have a machine that is nearing the edge of stability, any demanding program can push it over the edge.

So far I have not seen any benefit as far as point production. But I've had the new version running for just a few days.

[littleowl]

It seems on my e6600 system it crunches the WU's a little quicker then it was with the older version.

[Okda]

ya and that is due to the higher benchmarks in the newer version

[sierra_bound]

For those of you unhappy with 5.10.7, there's now a beta (5.10.8) in testing. As always, use at your own risk.

http://boinc.berkeley.edu/download_a...8&type=win_new

Eventually a 6.0 version will be released. No timetable, though I'm guessing later this year. BOINC 6.0 will have a new graphics program. Also, it will be available only as a service on Windows. The single-user and multi-user modes will be eliminated.

[Okda]

why will they eliminate them ? i only use them

[hawkeyefan]

For those of you unhappy with 5.10.7, there's now a beta (5.10.8) in testing. As always, use at your own risk.

http://boinc.berkeley.edu/download_a...8&type=win_new

Eventually a 6.0 version will be released. No timetable, though I'm guessing later this year. BOINC 6.0 will have a new graphics program. Also, it will be available only as a service on Windows. The single-user and multi-user modes will be eliminated.

also 5.10.8 x64

[[XC] hipno650]

well i got an explorer crash on my 4200X2. but things seemed to have leveled out now. why would they get rid of the single user mode? i use that allot. cuz i do more than just fold on my rigs lots of them do gaming as well. and i like to run bionc when im browsing the web and stuff but not when gaming so i would like to still have the single user mode so i can just shut bionc off when gaming.

[njkid32]

Since I've installed as service all my machines are totally fine.. So, I'll be sticking to 5.10.7.

[linflas]

The X64 version seems to be much more responsive on Vista64 for me, no lag, or ghost boinc's in the task manager, thanks for the link.

[NaeKuh]

OMG i love this build!

My Q6600
6/25/2007 11:22:27 PM||Running CPU benchmarks
6/25/2007 11:22:58 PM||Benchmark results:
6/25/2007 11:22:58 PM|| Number of CPUs: 4
6/25/2007 11:22:58 PM|| 3519 floating point MIPS (Whetstone) per CPU
6/25/2007 11:22:58 PM|| 7670 integer MIPS (Dhrystone) per CPU

My X3220
6/25/2007 11:21:48 PM||Running CPU benchmarks
6/25/2007 11:22:19 PM||Benchmark results:
6/25/2007 11:22:19 PM|| Number of CPUs: 4
6/25/2007 11:22:19 PM|| 3288 floating point MIPS (Whetstone) per CPU
6/25/2007 11:22:19 PM|| 7500 integer MIPS (Dhrystone) per CPU


I need to go over to my friends house and load up this new build on her X3210 i gave her this weekend.

[linflas]

wow, mine to, went up two thousand points on drystone, and about 600 on Whetstone

6/25/2007 11:39:51 PM||Running CPU benchmarks
6/25/2007 11:39:51 PM||Suspending computation - running CPU benchmarks
6/25/2007 11:40:22 PM||Benchmark results:
6/25/2007 11:40:22 PM|| Number of CPUs: 2
6/25/2007 11:40:22 PM|| 3453 floating point MIPS (Whetstone) per CPU
6/25/2007 11:40:22 PM|| 10403 integer MIPS (Dhrystone) per CPU
6/25/2007 11:40:23 PM||Resuming computation

[[XC] Malkec]

wow, mine to, went up two thousand points on drystone, and about 600 on Whetstone

6/25/2007 11:39:51 PM||Running CPU benchmarks
6/25/2007 11:39:51 PM||Suspending computation - running CPU benchmarks
6/25/2007 11:40:22 PM||Benchmark results:
6/25/2007 11:40:22 PM|| Number of CPUs: 2
6/25/2007 11:40:22 PM|| 3453 floating point MIPS (Whetstone) per CPU
6/25/2007 11:40:22 PM|| 10403 integer MIPS (Dhrystone) per CPU
6/25/2007 11:40:23 PM||Resuming computation

Holy @ that benchmark


My 5.10.7 install is stable as a rock. I always use service installation and haven't had a boinc related problem in ages.

[NaeKuh]

wow, mine to, went up two thousand points on drystone, and about 600 on Whetstone

6/25/2007 11:39:51 PM||Running CPU benchmarks
6/25/2007 11:39:51 PM||Suspending computation - running CPU benchmarks
6/25/2007 11:40:22 PM||Benchmark results:
6/25/2007 11:40:22 PM|| Number of CPUs: 2
6/25/2007 11:40:22 PM|| 3453 floating point MIPS (Whetstone) per CPU
6/25/2007 11:40:22 PM|| 10403 integer MIPS (Dhrystone) per CPU
6/25/2007 11:40:23 PM||Resuming computation

omg i bow down on your MIPS score.

[Movieman]

Holy @ that benchmark


My 5.10.7 install is stable as a rock. I always use service installation and haven't had a boinc related problem in ages.

What the heck are the specs on this machine?
A C2D at 4000+ on linux maybe? Impressive.

[linflas]

What the heck are the specs on this machine?
A C2D at 4000+ on linux maybe? Impressive.

E6600 @ 3.6, 4GB D9GMH @ 800mhz 4-4-4-8
Boinc 5.10.8 X64 install on Vista 64 Ultimate

This mornings is a bit slower

6/26/2007 9:34:28 AM||Running CPU benchmarks
6/26/2007 9:34:59 AM||Benchmark results:
6/26/2007 9:34:59 AM|| Number of CPUs: 2
6/26/2007 9:34:59 AM|| 3546 floating point MIPS (Whetstone) per CPU
6/26/2007 9:34:59 AM|| 10117 integer MIPS (Dhrystone) per CPU
6/26/2007 9:35:00 AM||Resuming computation

[sierra_bound]

Just a few words about the CPU benchmark. A 10% increase in your benchmark scores does not mean your daily point production will go up by a similar percentage. The older versions of BOINC basically have a bug that prevented the correct reading of scores, especially for systems using C2D. This affected only the integer calculation. To be quite honest, I have not noticed any increase in point production since switching a few machines to 5.10.7.

[Movieman]

E6600 @ 3.6, 4GB D9GMH @ 800mhz 4-4-4-8
Boinc 5.10.8 X64 install on Vista 64 Ultimate

This mornings is a bit slower

6/26/2007 9:34:28 AM||Running CPU benchmarks
6/26/2007 9:34:59 AM||Benchmark results:
6/26/2007 9:34:59 AM|| Number of CPUs: 2
6/26/2007 9:34:59 AM|| 3546 floating point MIPS (Whetstone) per CPU
6/26/2007 9:34:59 AM|| 10117 integer MIPS (Dhrystone) per CPU
6/26/2007 9:35:00 AM||Resuming computation

That's really surprising to see that high a benchmark at that speed.
The whet makes sense, the Dhry looks very high for 3600mhz.

Just a few words about the CPU benchmark. A 10% increase in your benchmark scores does not mean your daily point production will go up by a similar percentage. The older versions of BOINC basically have a bug that prevented the correct reading of scores, especially for systems using C2D. This affected only the integer calculation. To be quite honest, I have not noticed any increase in point production since switching a few machines to 5.10.7.

Agreed, just the "potential" for higher scores when we luck out and the other people in the quorum are also on 5.10.7..
Where I see the real potential is in the slower machines. My DX3600 and DX2400 are doing MUCH better.
Funny to call them slow machines but in todays world, they are..