| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Chat Room
Thread: The PrimeGrid Cheating Scandal |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 53
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Maybe you could get your head up of the sand and stop trying to glorify PG over WCG in this cause. PG has lost 1 whole month of work while WCG has not lost even a second. It's not true. Before cheaters reveal we thought we have already done PPS Sieving 1 month faster than usual. Today we know we need 1 more month to achieve the level we thought yesterday we are at. Noone, except the cheaters, is hurt. All work done by others ordinar users accepted with no changes. And you are wrong about WCG has not lost even a second. Every Invalid task needs additional re-check so WCG already lost a time too. The only difference between PG and WCG - WCG paid (and continues to pay) credits to cheaters for their inventiveness, PG cut off ALL their credits (legitimate or fraudulent) in full! [Edit 1 times, last edit by Former Member at Apr 23, 2013 9:41:39 PM] |
||
|
|
twilyth
Master Cruncher US Joined: Mar 30, 2007 Post Count: 2130 Status: Offline Project Badges: 
|
Maybe you could get your head up of the sand and stop trying to glorify PG over WCG in this cause. PG has lost 1 whole month of work while WCG has not lost even a second. It's not true. Before cheaters reveal we thought we have already done PPS Sieving 1 month faster than usual. Today we know we need 1 more month to achieve the level we thought yesterday we are at. Noone, except the cheaters, is hurt. All work done by others ordinar users accepted with no changes. And you are wrong about WCG has not lost even a second. Every Invalid task needs additional re-check so WCG already lost a time too. The only difference between PG and WCG - WCG paid (and continues to pay) credits to cheaters for their inventiveness, PG cut off ALL their credits (legitimate or fraudulent) in full! One of us misunderstands how things work here. Invalid results get resent to other machines or just one other machine if there is no redundancy. I guess from the point of that particular work unit there is a delay, but look at this way. If the cheaters never participated, that wu would have had to wait in the cue for an available device anyway. So I really don't see how anything was lost from the project standpoint and therefrom from the standpoint of WCG.  As to the points issue, again . . . anyone who really cares about what we do here, simple doesn't give a rat's posterior.    |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Those individuals did join WCG and we have checked their results and all of their results are either pending validation, inconclusive or invalid. They have no valid results. Our validators have kept the data safe. They have no valid results, but they have 1M-2.5M BOINC credits for the last 3 months. :) This is not that important. The important is that there were no corrupted results which have to be re-run and also that techs found out how to manage such attacks  Cheers  Now.. Thats funny.  Maybe better to get the head up of the sand.. IF these people have managed to get credit for not valid work.. then they have gotten exactly what they want. And the reason they could do it is because the project actually can Not handle such attacks. Then when its noticed, it should be handled as PrimeGrid did; Fixing the system, and delete credit for not valid work. Thats the only way to say that one doesnt accept cheating. Well, I think all of us can choose our own path or way to the stars. Mine might be full of sand dunes in which I can time-to-time put my head, but it doesn't prevents me to see further than on the tip of my nose. Maybe you could get your head up of the sand and stop trying to glorify PG over WCG in this cause. PG has lost 1 whole month of work while WCG has not lost even a second. You are underrating WCG because cheaters have gotten some credits. So what? They can pave their itinera ad astra with them. You are here much longer than me to know, that WCG is the only project crediting its volunteers even for invalid / inconclusive tasks. This is fair, because it is not always the crunchers' fault when return invalids. But other project are blaming only us for all inconclusives. I am not sure what you mean with "Fixing the system", but I am guessing it has something to do with inability of changing usernames to include asterisks at the end of them... Are credits and asterisks your ultimate goals? Well, have a nice trip. I beg your pardon...?  what do you think this is about? "Glorify Pg over WCG"? I am sorry you yet again need to point your finger at the wrong thing here. I did not glorify anyone over anyone else. But pointing out what is going on - and why it should be dealt with - even if you dont want to see it. "Underrating WCG Because cheaters have gotten some credits. so what?" Again.. I dont underrate WCG - just pointing out whats going on here. ;) Why are you so busy defending WCG ? No one here blame WCG as a project for anything.. Its not WCG "fault" that they get attacked by people who wants to cheat. And when so happens, one should think any project takes it serious. I also see that you have the need to drag in talk about astra and asterisks.. as a way to put me down. I dont know why you think that is needed. And I dont know what you think you will achieve with it either. If WCG doesnt care if cheaters gets credit, it would be fine with an official statement, so that we all get to know it is totally accepted. but I dont take your saying as an official WCG statement.. so until I see one, I still have good hopes for another outcome. [Edit 1 times, last edit by Former Member at Apr 23, 2013 10:13:26 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Ok, let look at 1 WU lifetime:
DSFL_ 00070-17_ 0000045_ 0821_ 0-- 625 Invalid 31/03/13 16:05:24 31/03/13 16:43:45 0.64 19.7 / 48.0 <--- cheater DSFL_ 00070-17_ 0000045_ 0821_ 1-- 625 Valid 31/03/13 16:06:10 01/04/13 07:33:22 5.14 96.3 / 96.0 DSFL_ 00070-17_ 0000045_ 0821_ 2-- 625 Valid 01/04/13 07:34:55 04/04/13 11:26:24 5.84 95.6 / 96.0 This WU could be validated 01/04/13 07:33:22, but "thanks to" cheater, 2nd task was resent only 01/04/13 07:34:55 How long WU have had to wait in the cue for an available device? 2 minutes 23 seconds. The result: this WU was validated only 04/04/13 11:26:24. You may say that without cheater this wu could be validated 04/04/13 too. You'll be right, but not at 11:26:24, and, for example at 10:52:24 How many time WCG losing "thanks to" cheaters? On the average the total time WUs are at cheater side. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Ok, I'll honestly answer to your question.
----------------------------------------I very often hear from ardent supporters of WCG that PG pays "exaggerated" credits for meaningless prime finding. Let leave aside the question of meaning/meaningless of PG efforts, let focus at "exaggerated" credits. If you don't care points/credits earning, you have to admit WCG pays even more "exaggerated" points/credits, because you pay for nothing. [Edit 1 times, last edit by Former Member at Apr 23, 2013 10:40:43 PM] |
||
|
|
twilyth
Master Cruncher US Joined: Mar 30, 2007 Post Count: 2130 Status: Offline Project Badges:
|
Ok, let look at 1 WU lifetime: DSFL_ 00070-17_ 0000045_ 0821_ 0-- 625 Invalid 31/03/13 16:05:24 31/03/13 16:43:45 0.64 19.7 / 48.0 <--- cheater DSFL_ 00070-17_ 0000045_ 0821_ 1-- 625 Valid 31/03/13 16:06:10 01/04/13 07:33:22 5.14 96.3 / 96.0 DSFL_ 00070-17_ 0000045_ 0821_ 2-- 625 Valid 01/04/13 07:34:55 04/04/13 11:26:24 5.84 95.6 / 96.0 This WU could be validated 01/04/13 07:33:22, but "thanks to" cheater, 2nd task was resent only 01/04/13 07:34:55 How long WU have had to wait in the cue for an available device? 2 minutes 23 seconds. The result: this WU was validated only 04/04/13 11:26:24. You may say that without cheater this wu could be validated 04/04/13 too. You'll be right, but not at 11:26:24, and, for example at 10:52:24 How many time WCG losing "thanks to" cheaters? On the average the total time WUs are at cheater side. I think I've already admitted that any given individual wu would have been delayed in both scenarios haven't I? The scenarios being with and without cheaters. So I'm afraid I fail to see the point you're trying to make. Ok, I'll honestly answer to your question. I very often hear from ardent supporters of WCG that PG pays "exaggerated" credits for meaningless prime finding. Let leave aside the question of meaning/meaningless of PG efforts, let focus at "exaggerated" credits. If you don't care points/credits earning, you have to admit WCG pays even more "exaggerated" points/credits, because you pay for nothing. I think some responses may be missing, but IDK. If I'm mistaken, just ignore that comment. I say that only because it's not clear to me what you're responding to. Anyway, as has already been pointed out, the motivation for providing credit in the case of invalid and inconclusive wu's is that this is not always the fault of the device doing the work. I know that I can a machine that consistently produces good results for all projects but may produce bad ones frequently for one specific project. Other times I get a rash of bad results. I have no idea why this happens. If WCG didn't credit, I'm not sure I would care but then I have so many rigs I barely notice anyway. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I think I've already admitted that any given individual wu would have been delayed in both scenarios haven't I? The scenarios being with and without cheaters. So I'm afraid I fail to see the point you're trying to make. If I'll save total task durations for users who received _1 and _2 tasks, without cheater this WU would have the next lifetime: DSFL_ 00070-17_ 0000045_ 0821_ 0-- 625 Valid 31/03/13 16:05:24 03/04/2013 19:56:53 5.84 95.6 / 96.0 DSFL_ 00070-17_ 0000045_ 0821_ 1-- 625 Valid 31/03/13 16:06:10 01/04/13 07:33:22 5.14 96.3 / 96.0 So without cheater this WU would have been validated at 03/04/2013 19:56:53 04/04/13 11:26:24 - 03/04/2013 19:56:53 = 15:29:31 It's WCG time loss. I know that loss of time from all WUs with cheaters Invalid tesks can not be combined. This example is my comment to your phrase PG has lost 1 whole month of work while WCG has not lost even a second. [Edit 1 times, last edit by Former Member at Apr 23, 2013 11:36:16 PM] |
||
|
|
twilyth
Master Cruncher US Joined: Mar 30, 2007 Post Count: 2130 Status: Offline Project Badges:
|
I think I've already admitted that any given individual wu would have been delayed in both scenarios haven't I? The scenarios being with and without cheaters. So I'm afraid I fail to see the point you're trying to make. If I'll save total task durations for users who received _1 and _2 tasks, without cheater this WU would have the next lifetime: DSFL_ 00070-17_ 0000045_ 0821_ 0-- 625 Valid 31/03/13 16:05:24 03/04/2013 19:56:53 5.84 95.6 / 96.0 DSFL_ 00070-17_ 0000045_ 0821_ 1-- 625 Valid 31/03/13 16:06:10 01/04/13 07:33:22 5.14 96.3 / 96.0 So without cheater this WU would have been validated at 03/04/2013 19:56:53 04/04/13 11:26:24 - 03/04/2013 19:56:53 = 15:29:31 It's WCG time loss. I know that loss of time from all WUs with cheaters Invalid tesks can not be combined. This example is my comment to your phrase PG has lost 1 whole month of work while WCG has not lost even a second. But you haven't proved anything. I'm not sure why you don't see that. OK, so that particular wu didn't get validated until much later. So? Was any time lost by devices actually crunching work units? No. Was the eventual completion of the project delayed in any way. No. Again, I don't see your point. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Was any time lost by devices actually crunching work units? No. Agree. Was the eventual completion of the project delayed in any way. No. Disagree. If one of the cheaters systematically invalidates a huge amount of tasks, the project has been delayed. Why do I think this is happening systematically? Because I see that during the last 3 months one of the (PG) cheaters had one the biggest WCG PPD... which (could) means a huge amount of invalidated tasks... and you name this phenomenon "eventual".  How many of yours have PPD = 1.4-1.6M ? I've knew you don't care about credits, but these credits are an equivalent of invalidated tasks... |
||
|
|
twilyth
Master Cruncher US Joined: Mar 30, 2007 Post Count: 2130 Status: Offline Project Badges:
|
Was any time lost by devices actually crunching work units? No. Agree. Was the eventual completion of the project delayed in any way. No. Disagree. If one of the cheaters systematically invalidates a huge amount of tasks, the project has been delayed. Why do I think this is happening systematically? Because I see that during the last 3 months one of the (PG) cheaters had one the biggest WCG PPD... which (could) means a huge amount of invalidated tasks... and you name this phenomenon "eventual". [img]http://stats.free-dc.org/charts/userdailysing...bwcg&name=671843&[img] How many of yours have PPD = 1.4-1.6M ? I've knew you don't care about credits, but these credits are an equivalent of invalidated tasks... Please explain to me how the project was delayed. If it's so obvious, you should be able to explain this to me as if I'm a 10 year old (believe me, I act like it at times). |
||
|
|
|