Message boards :
Graphics cards (GPUs) :
*development* BOINC 6.6.33 is out
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
For those like like to test and/or live on the leading edge a new development build is out. BOINC v6.6.33 for Windows, Windows x64, Linux, Linux 64 and Macintosh. If you have any problems post them on the BOINC alpha mailing list. Change Log for 6.6.32 (not released or on any download list!) - Mac MGR: Eliminate erroneous references to wxMac-2.8.7 in XCode (now building with wxMac-2.8.10). - SS: Add new optional boolean xml tag <default_ss_first>; if true, then launch default screensaver graphics app first, else run science graphics first. - MGR: Remove references to ViewNews, it won't appear in this release. - fix display problem when talking to old clients - lib: return proper error codes from boinc_rename() and boinc_mkdir() - client: Haiku support (from Urias McCullough?) - client: include plan class in other_result list in sched request - client: unify code for writing soft link files - Mac MGR: Fix per-thread-locale logic on Mac; weak-link sqlite3 for OS 10.3.9 compatibility. - client: if scheduler request didn't request work, don't report 0 tasks - MGR: In Async RPC logic, change calling order for wxCondition and wxMutex to match examples in literature Change log for 6.6.33 - Mac: delete source files ViewNews.cpp, ViewNews.h from XCode Project - client: include plan class in other_result list in sched request (for resource-specific jobs-in-progress limit) - client: show "est. delay" correctly in work fetch debug msgs - client: show times correctly in rr_sim debug msgs - client: in "requesting new tasks" msg, say what resources we're requesting (if there's more than CPU) - client: estimated delay was possibly being calculated incorrectly because of roundoff error - client: add a 1e-6 slop factor in deciding if a resource is fully utilized. |
|
Send message Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I'll throw in the 1st error... It won't (for me at least) terminate the science apps when exiting. You can stop them by clicking Advanced -> Shut down connected client, but exiting the manager doesn't work. Same as 6.6.31 in this regard. BOINC blog |
Paul D. BuckSend message Joined: 9 Jun 08 Posts: 1050 Credit: 37,321,185 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
I'll throw in the 1st error... This is an on-again off-again problem in that the developers cannot make up their minds how they want the system to function. So, we have this confusion ... The trend has been to make it so you have to explicitly shut down the client or it will run regardless ... so you may as well get used to going into the advanced menu and turning off the client when you want it shut down. {edit - add} Oh, and I may have found the error that could cause zero second work fetches to occur ... Richard Haselgrove has been chasing this one for awhile and I think I may have found the logic flaw that he exposed. Now we wait to see how much coding UCB will do to avoid touching the line I pointed to as the probable cause of the error ... the last time they rebuilt an entire module ... where we are still digging out bugs ... |
©2025 Universitat Pompeu Fabra