Message boards :
Number crunching :
Do I NEED to use BOINC 5?
Message board moderation
| Author | Message |
|---|---|
Dr. Dennis H. Chevalier Send message Joined: 7 Oct 01 Posts: 65 Credit: 16,622 RAC: 0
|
Dr. Chevalier R/ Al [/quote] Thanks for helping me out! doc Dr. Dennis Chevalier, PhD, FIPC, CPPP (Forensic Investigative Psychological Criminologist) The Consortium of Scientific Investigation and Research (CSIR) www.behavioral-sciences.org psychologist@charter.net |
|
Ulrich Metzner Send message Joined: 3 Jul 02 Posts: 1253 Credit: 13,565,513 RAC: 31
|
For someone like me who never has a big cache (.1 days), is there any reson I shouldn't be able to stick with 4.19? I kind of like it's simplicity. Just have a look here: http://boinc.berkeley.edu/download.php The working version for Solaris is still 4.19, although old. It is still working and i will use it until they tell me, i have to change. I also don't like at all the new version, it looks awkward, tremendous and horrible to me :P Aloha, Uli |
Keck_Komputers Send message Joined: 4 Jul 99 Posts: 1575 Credit: 4,152,111 RAC: 2
|
since the 'official' client is only @ 5.2.13 I referenced the older versions mainly so that someone already running that or later would know there is not currently a reason to upgrade farther in my opinion. BOINC WIKI BOINCing since 2002/12/8 |
Bill Michael Send message Joined: 4 Dec 03 Posts: 1122 Credit: 13,376,822 RAC: 100
|
since the 'official' client is only @ 5.2.13 Please note that the minimum version for some features is 5.2.8. It is still available, just not as the "recommended" which is at 5.2.13. Some people have valid reasons to run older versions. The lowest version that supports flops-counting is (I think) 5.2.6. There is no auto-update for BOINC - only SETI and the other project science applications. Auto-updating BOINC itself could be a security risk. There was never a BOINC V5.1.x, other than development versions; similarly, the 5.3.x are current development versions. Even middle digits = released versions (mostly). |
Clyde C. Phillips, III Send message Joined: 2 Aug 00 Posts: 1851 Credit: 5,955,047 RAC: 0
|
Maybe the Boinc Manager is the 5- series and the Boinc Daemon is the 4- series. ¡Qué confusión! |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0
|
Look more closely, I think what you will see running is the SETI 4.18 science application, not the BOINC management framework.
|
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0
|
Dr. Chevalier The machine that you have shows that it is running using 5.2.13 (returned workunits). From taskmanager the 4.18 that you see is the Science Application... The Boinc Core Client you installed (5.2.13) version information can be found in BoincManager on the "Help" Tab. "About BOincManager" {quote] I have 5.2.13 but when I do a Ctl-Alt-Del for the task list, I see Boinc 4.18 running. Whats the deal with that? It been the same through 3 clients I have upgraded from. Am I doing something wrong? Please advise as I am running 9 computers with dual processors on differant id's. thanks doc[/quote] R/ Al Please consider a Donation to the Seti Project. |
Dr. Dennis H. Chevalier Send message Joined: 7 Oct 01 Posts: 65 Credit: 16,622 RAC: 0
|
I have 5.2.13 but when I do a Ctl-Alt-Del for the task list, I see Boinc 4.18 running. Whats the deal with that? It been the same through 3 clients I have upgraded from. Am I doing something wrong? Please advise as I am running 9 computers with dual processors on differant id's. thanks doc Dr. Dennis Chevalier, PhD, FIPC, CPPP (Forensic Investigative Psychological Criminologist) The Consortium of Scientific Investigation and Research (CSIR) www.behavioral-sciences.org psychologist@charter.net |
|
PCRacer Send message Joined: 5 Jul 00 Posts: 3 Credit: 685,065 RAC: 0
|
since the 'official' client is only @ 5.2.13 http://boinc.berkeley.edu/download.php why reco these? where are they? are they beta/test versions? are they 'official' releases? is it necessary to have a specific client that will eventually auto update or will boinc give you a message when the hosts running 4.xx - 5.1.x are no longer compatable?? Every Dog Has His Day...I Wonder If I've Had Mine Yet? |
|
[BOINCstats] Willy Send message Joined: 4 Mar 01 Posts: 201 Credit: 152,243 RAC: 0
|
You will need BOINC v5.2.x to use an Account Management System when they become available. You'll need v5.2.3 minimum to use a full featured AMS. Join team BOINCstats |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0
|
BOINC View has a small bug where if it thinks the log file is corrupted it will not add more results into it. So, yes you can "lose" some work units. Since I roll my logs on a weekly basis this does not hit me too badly. If it is not adding them, shut it down, rename the log and restart it and it will create a new log, I THINK it will then "post" all the ones it had pending ... The next version should have logging to a MySQL database which, in theory, will improve matters. Though I will have to figure out how to do replication from where BOINCView logs to another database where I can play with the numbers more easily. |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0
|
BOINC is an evolving project. But the 3rd party apps (like BOINCView) also need updating to support new functionality in BOINC. e.g. you need BOINCView 1.2.2 or later for BOINC 5 support. The OP was specifically referring to liking the 'look' of BOINCMgr 4.x. Eventually he will need a BOINC > 4.x and BOINC and BOINCMgr come as a pair.
|
|
Dave Mickey Send message Joined: 19 Oct 99 Posts: 178 Credit: 11,122,965 RAC: 0
|
Just as a point of reference for Matt: I was in the same place you are - that is, asking "should I take the plunge beyond 4.19", when it was doing just fine. Never did any of the 4.x's above that. Sat back and just watched all the bugs of those higher 4.x's go by. I too didn't like BV dropping some of the completed units from it's log. Finally did take the plunge into 5.2.x (I think maybe .2 to start, recently upgraded to .13). I got 3 things from it: I was worried about the prospect of being too far behind when mandatory changes come along (perhaps with enhanced, but I don't really know). Maybe some big problem would arise, or I'd have to rush into it. I got the result duration correction factor which really simplified the cache size worry - my machines run a spread of < .5 to almost 1.0. Now I can give it just a single, real real-time number. But maybe if you sit at .1, you don't care. I do bigger caches than you. I got to upgrade BoincView (1.2.2), and I think this has fixed the missed completed result problem. Since I was using BV before 5.2.x, I haven't really even looked at the BoincMgr - I don't even start it. BV is great for monitoring multiple machines, and the current versions are a bit better than the one that works with 4.19. So, anyway, in my situation I've had no ill effects from going 4.19 to 5.2.13, here in my SETI-only world of slow computers. Don't know if the same opinion would hold in the multi-project case. Just my $.02, and worth every penny. Dave |
Lee Carre Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0
|
BOINC is an evolving project. that's also why everything is moduler, so that you can have 3rd party GUI apps, instead of the standard/stock one |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0
|
BOINC is an evolving project. Sooner or later a feature you really want or must have will come along. That or a new project or updated SETI that requires a specific higher version will appear. One day you will have to update so perhaps you need to just go with thye flow now.
|
Keck_Komputers Send message Joined: 4 Jul 99 Posts: 1575 Credit: 4,152,111 RAC: 2
|
The benchmarks were definatly changed after 4.19, I'm not sure if there were any major changes after that. 4.45 (and several around it) had the return results immediately bug that would often cause the logging programs to miss workunits or miss the final status. The new version will normally let work sit around in the 'ready to report' state long enough for the loggers to spot it. BOINC WIKI BOINCing since 2002/12/8 |
Lee Carre Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0
|
I've tried BoincView in the past, but I'm a perfectionist, and sometimes BoincView doesn't log a work unit in its logs. That bugs the heck out of me and made me stop using it. well, don't have it log any, then it's just an improved manager :) but the newer versions are much improved over the old(er) ones |
MattDavis Send message Joined: 11 Nov 99 Posts: 919 Credit: 934,161 RAC: 0
|
I've tried BoincView in the past, but I'm a perfectionist, and sometimes BoincView doesn't log a work unit in its logs. That bugs the heck out of me and made me stop using it. -----
|
Lee Carre Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0
|
Where there some changes in the methods of the version 5 benchmarks ? - On my test I noticed that dhrystone and whetstone are 30 % lower than on version 4.x.yes there are, at least for multi-CPU hosts, i noticed quote a large increase in both, and my claimed credit rose to what it should be too ;) I always had the or one of the lowest claims of credit, and my multi-CPU host is a dual 3.06GHz Xeon machine, so it's well worth upgrading as you'll also get nice features like the result duration correction factor, and networking improvements too, like i said previously, if it's just for GUI/cosmetic reasons then use BoincView |
|
Dotsch Send message Joined: 9 Jun 99 Posts: 2422 Credit: 919,393 RAC: 0
|
But, when Seti_Enhanced is "soon" released, it's in users own interest to upgrade to v5.2.6 or later, to get consistent claimed credit... Where there some changes in the methods of the version 5 benchmarks ? - On my test I noticed that dhrystone and whetstone are 30 % lower than on version 4.x. Also I noticed that the benchmark is not impactable by the compiler optimisations. |
©2020 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.