Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
1)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 122119)
Posted 1670 days ago by Vortac
There was no price increase, those are just usual daily oscillations. The price is steady at $0.01.
|
2)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 122094)
Posted 1671 days ago by Vortac
That's based on current pricing. Using the max price GDC has ever been it would be more like $270.
Well spotted. The incentive was $270 per day at one point in January - and no attacks on PrimeGrid were made (as far as we know). There were even no concerns about monetization in general. Surprisingly, things came to a head when incentives shrank to $12 per day.
Well, one has to strike a balance between user participation and incurred risks. Smaller incentives=less participation, but also less risk of cheating. Maximum security is achieved with zero users, but that's obviously impractical. PrimeGrid admins are apparently fine with less participation, although zero attacks were made with 20x larger incentives. To users not fully familiar with BOINC security (like me) that decision looks a bit too restrictive, but again, it's their project and they can decide to reduce those incentives even further, if they want so.
|
3)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 122080)
Posted 1671 days ago by Vortac
We discovered that, although it would not be easy, it could be done if you were willing to put enough effort into it. We concluded that if our BOINC credit was convertible in money, that might provide enough incentive to put in the effort.
Well, that attack you have discovered is almost certainly possible even without Gridcoin. I mean, simply not sharing statistics data with a third party surely won't mitigate such a dangerous threat? Meaning that, even without Gridcoin, PrimeGrid is still vulnerable?
Because, let's be frank, $12 per day is a negligible incentive and, as RagnarokDel has pointed out, that's all one could earn here, even if he hacks PrimeGrid completely and grabs all credits for himself. This whole discussion is hardly worth $12 per day, let alone hacking PrimeGrid through this new, undiscovered method.
|
4)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121945)
Posted 1672 days ago by Vortac
A note to all the "PG defenders": Please try to be kind to the people who feel like they have been wronged.
For some of them, this may cost them non-trivial amounts of money. For others, it's a repudiation of something they're very passionate about. Either way, it's easy to empathize with their positions.
So cut them a little slack, please. Vortac, for example, is not perhaps the epitome of etiquette today, but I think we can still afford to treat him the way we ourselves would like to be treated.
This isn't about who'se right and who'se wrong. Everyone should have an opportunity to express their opinions. Today this is even more true.
Thank you Michael, for your kind (albeit somewhat patronizing) words. But really, I must say, I am not visiting PrimeGrid forums so that people can empathize with me. Nor does anyone here need to worry about my financial situation, thank you very much. I usually visit PrimeGrid forums to engage in technical discussions, so I think there's no need to cut me some slack, especially if I am wrong about a certain technical issue.
The truth is, I am a PrimeGrid defender too and, any way you cut it, the project is about to lose some of the support it enjoys now (even if it's less than 35%). Shouldn't all true defenders be worried about that? Or shall we congratulate each other?
|
5)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121935)
Posted 1672 days ago by Vortac
why are you attacking so much?! I'm in team Gridcoin just like you and refer you to my original post
You wrote here http://www.primegrid.com/forum_thread.php?id=8272&nowrap=true#121899
It's slowly coming out on this thread that many Gridcoiners are after as much credit in as short a time as possible - up to and including hacking and cheating. The Collatz project was a victim of this, and I saw that some Gridcoiners abort Genefer21 tasks assigned to the CPU because of the time taken to complete. That's the threat that primegrid sees, hence the decision.
That's nonsense and FUD. There is absolutely no threat in aborting tasks and all work is double-checked with PrimeGrid, so it's impossible to hack the tasks like they did at Collatz. Get your facts straight.
|
6)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121931)
Posted 1672 days ago by Vortac
until many people do it the same way an the dc has the same wrong/cheated result
More wrong points. Even if half of the crunchers managed to 'hack' their tasks, that still means 50% failure rate on the double checks.
Really, it's very unfortunate that people are so quick on supporting bans, but apparently don't understand even the basic principles of work validation.
|
7)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121925)
Posted 1672 days ago by Vortac
My point was the slowest tasks might be perceived as prone to hacking due to the high amount of credit.
Well, it's another wrong point then. GFN tasks are issued with much less credits (proportionally) than shorter tasks like PPS Sieve. And even if someone managed to hack them, it would be a wasted effort because such hacked tasks cannot pass double-checking then.
|
8)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121918)
Posted 1672 days ago by Vortac
The credit makes perfect sense for the CPU version of the software. The GPU code simply gets the same amount of credit
Michael was kind enough to explain to us (in the Member Poll thread) that he can, as a PrimeGrid admin, adjust the credits anytime. If sieving is so far ahead, perhaps it's indeed time to adjust the credits and issue more of them to prime-finding projects, so that we can catch up on sieving? I guess credits make perfect sense in the context of actual FLOPS done, but it looks like that's not the only consideration here? Correct me if I am wrong please, my understanding of the project is surely not as good as yours.
|
9)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121914)
Posted 1672 days ago by Vortac
And I seriously doubt we're going to lose 35% of our computing power. But if we do, that power is probably concentrated in the PPS Sieve subproject. I will quite literally not live long enough to see that factoring put to use at PrimeGrid.
I've seen that argument before too. It's true, Gridcoin members mostly crunch PPS Sieve - because PrimeGrid admins decided PPS Sieve should yield most credits. If, for example, GFN tasks were rewarded equally (or more), they would be happy to crunch them too.
So, if sieving is so far ahead... why not adjust the credit system and start rewarding it with less credits? Instead, what we get is "we don't need that much computing power anyway, so good riddance to them". Makes very little sense to me.
|
10)
Message boards :
Number crunching :
New PrimeGrid Policy - Monetization of BOINC credit
(Message 121909)
Posted 1672 days ago by Vortac
The Collatz project was a victim of this, and I saw that some Gridcoiners abort Genefer21 tasks assigned to the CPU because of the time taken to complete. That's the threat that primegrid sees, hence the decision.
1. Collatz scenario cannot happen at PrimeGrid, because all results are double-checked here and always were (even before Gridcoin).
2. Aborting GFN-21 CPU tasks is obviously not cheating. Anyone can do it, aborting tasks is a legit BOINC operation. I would abort them too, why waste a CPU on a task which can be done much faster on a GPU?
The fact is: PrimeGrid is banning Gridcoin and losing ~35% of its computing power without ANY good reason. There was no actual cheating and no fake results EVER (with PrimeGrid), all such scenarios posted here were either made up or are impossible due to security mechanism which are already in place (like double-checking). The whole thing is so absurd and so devoid of any valid arguments that even aborting tasks is presented as a threat. I am very sorry to see such ineptitude here.
|
Next 10 posts
|
