Grid computing center

log in

Advanced search

Message boards : News : Grid computing center

Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · Next
Author Message
Profile [AF>Amis des Lapins] Phil1966
     
Send message
Joined: 9 Aug 13
Posts: 6
Credit: 812,259
RAC: 0
Total hours: 2,907,771
Message 6020 - Posted: 25 Feb 2018, 12:02:54 UTC - in response to Message 6016.
Last modified: 25 Feb 2018, 12:04:10 UTC



Looks like you (and other crunchers) think you can define what is "cheating" on someone else server. Just think it twice (or more)...
SPOILER : no! your server, your rules ;-)

Secondly, you are not able to understand the validity of the rules, doesn't mean you're right.

Thirdly, you must learn to understand the point of view of others and especially put the form when you speak publicly on a forum. Questions and courtesy can help people to change their point of view. The form is very important in discussions, especially when one is in a weaker position.

+1

mmonnin
       
Send message
Joined: 22 Aug 16
Posts: 417
Credit: 1,927,837
RAC: 707
Total hours: 9,130,092
Message 6021 - Posted: 25 Feb 2018, 13:09:24 UTC - in response to Message 6016.


What fake data? The results are typically legit. The work units were ran and unaltered. The data got reported. How is that fake results? What you should be saying is that people weren't running projects the way a very small group of individuals think they should run them and label it as "cheating". Again, that concept is only in the minds of few as the results are no different from anyone else's results. Cheating is when someone doesn't do the work but still claims the credit.

I find that the results produced by this project are only a best guess anyways. It does not tell me if someone overclocked their CPU. It does not tell me if they ran multiple work units on each device which can impact its results. Running multiple work units on a video card is no different than running multiple clients for NCI projects. Why? because by your logic they are following those same rules. It doesn't matter if they are cpu intense or not. Every project stresses the system different. They run different. They use different resources. If maximizing my GPU to its fullest is ok by running multiple work units then I should be able to the same on my CPU. Regardless of CPU intensive levels. That isn't cheating. That is utilizing your hardware to its fullest for the sake of the project. Yes it is something point seakers look at. But it is something people who want to maximize their resources look at as well. Who are you to determine that something like that is "cheating". What gives you that right? "Science" was performed and was not "faked". WUProp may not be able to report the way it wants to. But it never has been fully accurate since the beginning. And it never will be unless it can gain all of those individual system difference. Even then, it would be too much data to filter through.


Looks like you (and other crunchers) think you can define what is "cheating" on someone else server. Just think it twice (or more)...
SPOILER : no! your server, your rules ;-)

Secondly, you are not able to understand the validity of the rules, doesn't mean you're right.

Thirdly, you must learn to understand the point of view of others and especially put the form when you speak publicly on a forum. Questions and courtesy can help people to change their point of view. The form is very important in discussions, especially when one is in a weaker position.


I think part of the issue that started this thread arose was due to what was considered 'cheating' changed several times w/o warning. So as we were playing by the rules it was then no longer acceptable. That happened 3 times.

damotbe
   
Send message
Joined: 22 Nov 14
Posts: 4
Credit: 5,235,494
RAC: 1,023
Total hours: 25,362,510
Message 6022 - Posted: 25 Feb 2018, 14:07:04 UTC - in response to Message 6021.

I think part of the issue that started this thread arose was due to what was considered 'cheating' changed several times w/o warning. So as we were playing by the rules it was then no longer acceptable. That happened 3 times.


I understand that. And one can understand that changing rules is a part of all systems when exploits/bugs are discovered regarding the primary goal of the system.

Primary goal of WuProp is not points or badges which are incentives. The goal of incentives in a system are to encourage good behaviours or to discourage bad behaviours. In WuProp points and badges encourage users to share their compute times and workunits properties to help the community with valuable data and it helps small or unknown projects.

But incentives alors work in another way. The community can gives incentives to the administrator, positive or negative, to encourage or discourage him. The Internet should not make us forget that real people are in front of their screens, the distance is just an illusion. Seb is a real person with a real job and he gives you a part of his free precious time. Give him positive incentives to continue working on something you like ! Be positive !

Profile [AF>Le_Pommier] Jerome_C2005
     
Send message
Joined: 28 Mar 10
Posts: 18
Credit: 617,942
RAC: 178
Total hours: 1,745,682
Message 6023 - Posted: 25 Feb 2018, 21:02:41 UTC

+ being not happy about what <some people> consider like "unfair change of rules" (in their point of view) was a fairly good reason to

- detach from the project since they didn't like it anymore, and stop wasting their own time here trying <whatever they were trying to do>

- develop their own boinc project with their own rules, to do something similar and/or better (in their point of view) and propose it to the boinc community

Cocagne
   
Send message
Joined: 23 Jan 11
Posts: 2
Credit: 642,222
RAC: 0
Total hours: 1,284,962
Message 6025 - Posted: 25 Feb 2018, 21:49:25 UTC - in response to Message 6023.

And let see if they will be as successful as Seb...

zombie67 [MM]
             
Avatar
Send message
Joined: 30 Mar 10
Posts: 219
Credit: 8,804,177
RAC: 1,273
Total hours: 30,507,871
Message 6027 - Posted: 26 Feb 2018, 15:06:32 UTC

I support this project 100%, obviously. WUPROP has kept me BOINC-ing long after chasing credits became boring.
____________
Reno, NV
Team: SETI.USA

EG
 
Avatar
Send message
Joined: 16 Dec 16
Posts: 1
Credit: 154,003
RAC: 0
Total hours: 1,162,885
Message 6028 - Posted: 26 Feb 2018, 15:44:37 UTC

Like WUprop owes them something.... (more like they forgot the point)

Would be a shame if he really quits, he created a DB of hardware and it's capabilities even if it is only a guesstimate of project and hardware capabilities.
Invaluable to us people that push HW to the edge.

Chasing stars is moot compared to that.....

I don't run it anymore cause it was effecting my collatz times, I always saw the badges in it as a compilation of nothingness so I never felt the need to chase them....

Is it cheating? no, I don't think so. Does it go beyond the point/intent of the project? yes, it does.

His project he can run it anyway he likes.... but he should have stopped the practice before users grew dependent on it....

AS far as the dictator crap? if you don't like it don't run it, the base philosophy of Free Distributed Computing.

We are free to choose. That is as far from dictation as possible.
____________

Profile Conan
       
Avatar
Send message
Joined: 28 Mar 10
Posts: 565
Credit: 1,173,751
RAC: 186
Total hours: 3,241,486
Message 6029 - Posted: 27 Feb 2018, 22:02:18 UTC - in response to Message 6027.

I support this project 100%, obviously. WUPROP has kept me BOINC-ing long after chasing credits became boring.


Add me to that support list.

Conan
____________

Profile [AF>Le_Pommier] Jerome_C2005
     
Send message
Joined: 28 Mar 10
Posts: 18
Credit: 617,942
RAC: 178
Total hours: 1,745,682
Message 6033 - Posted: 2 Mar 2018, 8:24:21 UTC

Unfortunately it seems Seb got bored over the top, he is not answering us on the AF forum neither... for the moment things are like that.

On the other hand he has always been doing most things he does without talking much about it, we would have problems on the <various/many> services he is running (WUProp, SAM, AF forums, StatSeb website, signature system(s), etc), we would complain about them (shame on us) and suddenly they would get fixed and voila.

So let's cross fingers, touch wood, eat fortune cookies and wish him the best.

Profile EBM
 
Avatar
Send message
Joined: 10 Nov 16
Posts: 10
Credit: 99,645
RAC: 100
Total hours: 234,421
Message 6038 - Posted: 7 Mar 2018, 2:53:18 UTC

My situation: I run 2 instances of WUProp; I have 2 copies of BOINC running on my computer because I want my weaker graphics card to only run Sieving on PrimeGrid.

Anyhow, you do realize that even if you had every boinc project on one client, you only need one WUProp account. If I run Seti, WCG, Primegrid and milkyway for an hour and a half each, I get 1.5 hours of runtime for each project. If I am understanding this thread correctly, then you REALLY don't need 8 instances on a 4 core/2core HT host. You don't need a separate boinc client on your computer for every project you want to run.

Profile [AF>Le_Pommier] Jerome_C2005
     
Send message
Joined: 28 Mar 10
Posts: 18
Credit: 617,942
RAC: 178
Total hours: 1,745,682
Message 6040 - Posted: 8 Mar 2018, 17:36:42 UTC

Of course !

And I'm not expert in that part but I'm not 100% sure you need to run 2 boinc instance to deal with your PrimeGrid/GPU issue, I think this could be achieved through config files, but again I'm not knowledgeable enough on that part...

https://boinc.berkeley.edu/wiki/Client_configuration
(for instance see <exclude_gpu>)

https://boinc.berkeley.edu/trac/wiki/ClientAppConfig
(seems quite empty here)

I can ask at l'AF forum if you wish

Profile Coleslaw
         
Avatar
Send message
Joined: 11 Apr 10
Posts: 182
Credit: 8,164,885
RAC: 2,124
Total hours: 19,353,436
Message 6041 - Posted: 9 Mar 2018, 3:22:34 UTC - in response to Message 6038.
Last modified: 9 Mar 2018, 3:41:16 UTC

Anyhow, you do realize that even if you had every boinc project on one client, you only need one WUProp account.


Yes you only need one account. That isn't what is being discussed. What is being discussed is that people were running multiple instances of WUProp work units. What is happening is that people wanted all of their runtime hours to count. And I honestly don't think it picks up all of the clients with the one work unit.

So, if I have a rig with 250 clients running (which I do). Each of those clients have 4 work units that are Non-CPU intensive each. I would therefore be running 1,000 work units on that host. I look at my times here, and it shows me with a total of ~700 runtime hours for the day. That is across all hosts I have whether physical or virtual. So, I'm not positive if WUProp was crediting them based on the actual CPU time used or wall clock, but I think people were trying to guarantee they got all the runtime hours rather than just WUProp points. I'm thinking it was based off of wall clock because looking at the numbers, it would line up with total number of physical and virtual hosts rather than total number of clients actually ran in my case.
____________

Profile Yavanius
 
Avatar
Send message
Joined: 20 May 15
Posts: 32
Credit: 521,107
RAC: 247
Total hours: 540,563
Message 6046 - Posted: 10 Mar 2018, 19:38:21 UTC - in response to Message 6010.

Please tell me what "science" Grid Computing Center/Goofyx is doing?

Trying to find out if a monkey can type "Christmas" (app V2/V4) and how long it will take? https://en.wikipedia.org/wiki/Infinite_monkey_theorem


When we reach an infinite number of monkeys, we'll know. ;)

'Science' has always been objective. Fool's errands exist until something new is discovered, then suddenly it's a new field of study.

Profile Yavanius
 
Avatar
Send message
Joined: 20 May 15
Posts: 32
Credit: 521,107
RAC: 247
Total hours: 540,563
Message 6047 - Posted: 10 Mar 2018, 19:57:38 UTC - in response to Message 6041.

Well, WUProp only reports in every 6 hours. My logic might be off, but if you wanted to absolutely check, start WUProp at the same time on each server and then after 24 hours, check the report.

Incidentally, NCI doesn't always give you all 4 WUs. Sometimes it misses one or two. I don't know if it's just an instance or within a certain timeframe it happens. But times that 1 or 2 by a bunch of machines...

Apparently you don't follow the project as the project hasn't had a history of stability. GCC went down last year due to hardware failures and GoofyX has been trying to get things working again coupled with being hospitalized over the holidays. I think GCC is on hold till be can get NCI and CPU stable. Then he wants to combine them under one banner, but separate projects (like yoyo@home). Then maybe then GCC might make a return...but he hasn't said last I saw.

But, all that's mostly really beyond the scope of the original intent of this thread... :)

=Lupus=
     
Send message
Joined: 2 Apr 10
Posts: 8
Credit: 104,396
RAC: 40
Total hours: 263,703
Message 6048 - Posted: 11 Mar 2018, 11:56:22 UTC

My 2 Cent for the Goofy-NCI:

Seb, just kick it out of list. You dont have to worry about your stats anymore, and the ones trying to get a nice colored star for GCC-NCI... Well they can hate us, but they cannot do anything against it.

WUPROP is YOUR Project. YOU have the right to take whatever action to keep your databases' sanity well.

And if a host is only running monkey and WUPROP your database will ignore his/her monkey time and thus have an "empty" WUPROP - which can easily put out by a simple programming mechanic.

For me it is okay if you just "kill" goofyx-NCI from list, for the time being it abused by multiclienting.

For me, having more than one boinc client on ONE computer is considered cheating - but as sometimes the host identify process goes awkward (especially when your computing "farm" consists of exact the same computers - for example 20 PIs) banning people which do "most likely" have multiclient sometimes hits the wrong ones.

So, just to repreat myself: My opinion: Kick goofyx-NCI out of the list. just let wuprop ignore it.

MB Atlanos
 
Send message
Joined: 9 Apr 10
Posts: 2
Credit: 271,555
RAC: 17
Total hours: 109,480
Message 6074 - Posted: 31 Mar 2018, 3:59:21 UTC - in response to Message 6027.
Last modified: 31 Mar 2018, 4:00:00 UTC

I support this project 100%, obviously. WUPROP has kept me BOINC-ing long after chasing credits became boring.

Same for me. I sign the supportlist too.

greetings MB A

tito
     
Send message
Joined: 19 Apr 10
Posts: 8
Credit: 1,347,311
RAC: 81
Total hours: 3,658,146
Message 6233 - Posted: 29 Jul 2018, 13:42:33 UTC

It seems that Sebastien really stopped any work with wuprop and signature.statseb.fr
Does anyone have any informations?

Profile [AF>Le_Pommier] Jerome_C2005
     
Send message
Joined: 28 Mar 10
Posts: 18
Credit: 617,942
RAC: 178
Total hours: 1,745,682
Message 6234 - Posted: 29 Jul 2018, 15:36:32 UTC

Things not working ? like what ? (examples)

tito
     
Send message
Joined: 19 Apr 10
Posts: 8
Credit: 1,347,311
RAC: 81
Total hours: 3,658,146
Message 6235 - Posted: 29 Jul 2018, 16:03:36 UTC

Here all looks ok, but at signature.statseb.fr new project DHEP is missing.
Also new badges for GPUGrid publications are missing.
Maybe something more, but I can not find that :)

Profile [AF>Le_Pommier] Jerome_C2005
     
Send message
Joined: 28 Mar 10
Posts: 18
Credit: 617,942
RAC: 178
Total hours: 1,745,682
Message 6236 - Posted: 30 Jul 2018, 18:08:54 UTC

Yep DHEP is not considered yet, neither in SAM nor StatSeb, the announcement done by admin that his project is "live" is fairly recent, and it's holidays time around here too.

For GPUGRID I can't say, I'm not into it.

Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · Next
Post to thread

Message boards : News : Grid computing center


Home | My Account | Message Boards | Results


Copyright © 2024 Sebastien