• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

BOINC progress-report.

R

Rattledagger

Elite Member
After having solved the glut32.dll bug that largely contributed to a top of 7000 errors/day, it's now around 2000/day. Before core-client v2.28 was released it was around 400/day equals 10-12% error. Of course not linear, since 5300/day was 53% but still...

Seti-client v2.32 have just entered alpha, there they're trying to hunt down all the access-violations. These is the largest part of errors now.

Current seti-client for windows has much debug-code, with slows crunch-times to around 12h30m for AR 0.4 but is very dependent on angle-range since very high AR gives 4h50m. For reference, 3h40m in v3.03.

As I understand it, next beta-release of core client will not come before in May. If the seti-access-bugs is squished by then, this will most likely not be debug-builds...

As for release of BOINC, this will not happen before error-rate drops below 5%, and all high/critical "M1-Public Launch" in the taskbase is solved.
Afterwards they'll have to solve the high/critical "M2-Decommission Old Seti@home" that currently is to finnish programming separate gui for all platforms before old seti@home is killed off...
 
Thanks RD for the update.

Sounds like they are having a parsing problem.

Wish I was on the Design Team, would have that puppy knocked out in a flash. 😉
 
Are the problems in BOINC itself, or in the seti part of it? ie. could they release BOINC for other projects right now, if other projects had working crunchers?
 
Thanks for the update. Boy, the progress on this is underwhelming. 🙁
 
Support for other projects is M3, so it's a long way to go.
AFAIK except for the access-violations the bugs is in BOINC... Oh, 1 of these 15 giving same error was apparently in one of BOINC's lib-files...
A little count in taskbase shows 2 unspecified high/critical, 14 M1, 3 M2 & 3 M3. Not all of these is bugs. 😉


Dmcowen, since the sources is available, it's just to download & compile your own client & report/fix any bugs you find. 🙂
 
Originally posted by: Rattledagger
Support for other projects is M3, so it's a long way to go.
AFAIK except for the access-violations the bugs is in BOINC... Oh, 1 of these 15 giving same error was apparently in one of BOINC's lib-files...
A little count in taskbase shows 2 unspecified high/critical, 14 M1, 3 M2 & 3 M3. Not all of these is bugs. 😉


Dmcowen, since the sources is available, it's just to download & compile your own client & report/fix any bugs you find. 🙂
Click to expand...

That's true, I may check it out and see what I can see and do with it 🙂

 
Originally posted by: Smoke
Thanks for the update. Boy, the progress on this is underwhelming. 🙁
Click to expand...

well said smoke. enjoy SETI I as long as you can because when its done, the mass exodus to other projects begins!
 
Originally posted by: Unforgiven
Originally posted by: Smoke
Thanks for the update. Boy, the progress on this is underwhelming. 🙁
Click to expand...

well said smoke. enjoy SETI I as long as you can because when its done, the mass exodus to other projects begins!
Click to expand...

Yeah...kind of surprising given that is was slated to be done last month...oh well...keep crunchin'! 😀
 
Done last month? 😕

Anyway, the latest news is these:
"The next version of SETI@home will have a new graphical feature: when viewing in a window, you can use the mouse to rotate and zoom 3D objects."
"We fixed a bug that caused the same result to get sent to multiple hosts."

Also fixed a bug there an application unexpectedly terminates, and BOINC thinks successfully done crunching...


As for "mass exodus to other projects", anyone crunching seti@home not just for the credit should like BOINC, since a quick look on seti@home total wu returned & #wu generated shows every wu on average is crunched 6,46 times, while in BOINC it looks like seti aims on only 2 times and Astropulse 3 times... Of course some will discard wu and some bad crunching, and the client is probably a little slower, but still seti@BOINC will produce more science with only 40% of seti@home's crunching-power. The slow win-gui has contributed 43,89% of all results, so if most of these probably largely non-team-crunchers with no farms & no work-machines moves to BOINC it will already be more effective than seti@home. 🙂

 
You must log in or register to reply here.

TRENDING THREADS

Back
Top