Please remove "CompositeGrid" from your database

log in

Advanced search

Message boards : Number crunching : Please remove "CompositeGrid" from your database

Author Message
Michael Goetz
 
Avatar
Send message
Joined: 18 Apr 13
Posts: 31
Credit: 119,645
RAC: 0
Total hours: 113,979
Message 3969 - Posted: 9 Jul 2016, 22:51:20 UTC
Last modified: 9 Jul 2016, 22:54:47 UTC

"CompositeGrid" is a totally bogus "project". I should know -- I run it.

It's PrimeGrid's development server, and NOTHING on there is valid. It's only used (rarely) for testing, and nothing on there counts.

Nobody should be picking up badges for "work" done there.

Anyone running tasks from that system is contributing to global warming and their electric bill, but doing nothing else that's useful.

I wipe the database there regularly, or set it up with whatever oddball configuration I need. None of the information that WUProp is gathering from my test server is meaningful.

It's not a "beta project". It's used to test our BOINC server software.

Thank you.

Profile BilBg
Avatar
Send message
Joined: 20 Jun 12
Posts: 63
Credit: 82,991
RAC: 19
Total hours: 97,311
Message 4253 - Posted: 21 Sep 2016, 19:34:00 UTC - in response to Message 3969.

 
New "applications" from CompositeGrid are still announced in "[Beta] New / Intermittent applications" (22 times since 16 Aug 2016)
http://wuprop.boinc-af.org/forum_thread.php?id=351&postid=4234#4234
 
(I can read from first post they are "automatically published" but why?
It may mean somebody still runs them + WUProp?
)
 
____________



- ALF - "Find out what you don't do well ..... then don't do it!" :)

Profile Steve Hawker*
       
Send message
Joined: 31 Oct 12
Posts: 125
Credit: 642,446
RAC: 118
Total hours: 1,555,426
Message 4254 - Posted: 21 Sep 2016, 20:46:00 UTC - in response to Message 3969.

"CompositeGrid" is a totally bogus "project". I should know -- I run it.

It's PrimeGrid's development server, and NOTHING on there is valid. It's only used (rarely) for testing, and nothing on there counts.

Nobody should be picking up badges for "work" done there.

Anyone running tasks from that system is contributing to global warming and their electric bill, but doing nothing else that's useful.

I wipe the database there regularly, or set it up with whatever oddball configuration I need. None of the information that WUProp is gathering from my test server is meaningful.

It's not a "beta project". It's used to test our BOINC server software.

Thank you.


Just wondering why you don't use the power of the crowd to test whatever it is your testing. David Anderson uses the power of the crowd to test new versions of whatever on the BOINC test server and that achieves just as little science as you're claiming.

Yet testing is a valid activity. Whatever your purpose, a test IS useful else don't bother.

We're all more than happy to test systems, apps, configurations. For the most part we're adults capable of making our own decisions. Plenty of us are delighted to help out on the non-science and we'd like to help you.

Why not? What have you got to lose?

Michael Goetz
 
Avatar
Send message
Joined: 18 Apr 13
Posts: 31
Credit: 119,645
RAC: 0
Total hours: 113,979
Message 4257 - Posted: 22 Sep 2016, 1:42:32 UTC - in response to Message 4253.

 
New "applications" from CompositeGrid are still announced in "[Beta] New / Intermittent applications" (22 times since 16 Aug 2016)
http://wuprop.boinc-af.org/forum_thread.php?id=351&postid=4234#4234
 
(I can read from first post they are "automatically published" but why?
It may mean somebody still runs them + WUProp?
)
 


Because sometimes we test things. Sometimes privately. Sometimes publicly. Either way, however, if WUProp is running on the computer that is connecting to the test server, it's going to report those tasks to the WUProp server, which automatically posts forum announcements here.

Michael Goetz
 
Avatar
Send message
Joined: 18 Apr 13
Posts: 31
Credit: 119,645
RAC: 0
Total hours: 113,979
Message 4258 - Posted: 22 Sep 2016, 1:56:36 UTC - in response to Message 4254.

Just wondering why you don't use the power of the crowd to test whatever it is your testing. David Anderson uses the power of the crowd to test new versions of whatever on the BOINC test server and that achieves just as little science as you're claiming.

Yet testing is a valid activity. Whatever your purpose, a test IS useful else don't bother.

We're all more than happy to test systems, apps, configurations. For the most part we're adults capable of making our own decisions. Plenty of us are delighted to help out on the non-science and we'd like to help you.

Why not? What have you got to lose?


Sometimes we do just that. It's certainly helpful when we have something that we think is production-ready and want to give it a thorough shake out. Having real users access the server often reveals problems we didn't think to look for.

But if we've got something that we know doesn't work, and we're trying to fix it, having a crowd grabbing thousands of tasks when I only care about what happens to ONE task is harmful, not helpful. It hides the information I'm looking for in an ocean of irrelevant data, and it also means I need to make the server robust enough to serve an unexpectedly large number of tasks.

Not to mention the time wasted trying to figure out why there's thousands of tasks going out instead of just 1.


Post to thread

Message boards : Number crunching : Please remove "CompositeGrid" from your database


Home | My Account | Message Boards | Results


Copyright © 2017 Sebastien