On Oct 2, 2009, at 1:22 PM, Eric J Korpela wrote: > On Wed, Sep 30, 2009 at 9:42 PM, Paul D. Buck <[email protected]> > wrote: >> Yep, that is part of the definition... how is Eric's deflationary >> multiplier that debases this an improvement? > > You continue to claim that the credit multiplier is deflationary. I > don't see how you reach this conclusion. The multiplier is remarkably > stable, and any instability almost entirely due to variations in how > the data is analyzed rather than variations in the host machines or in > the BOINC clients that they run. It's certainly a far more stable > means of granting credit than the "benchmark x time" method since its > the median of all machines rather than the lower claim of two machines > or the median claim of three.
Wasn't me that said so ... it was you (and I quote): "There are also some fairly major drawbacks, and that is why I think it won't be implemented as is. First, it's immediately deflationary. The average host that connected to s...@home today earns 292 credits per day, which is significantly more than 100. There would be strong (perhaps irresistible) demand for projects to keep their current credit scheme for as long as possible. " http://setiathome.berkeley.edu/forum_thread.php?id=45024&nowrap=true#705408 And it seems to me skimming this this thread: http://setiathome.berkeley.edu/forum_thread.php?id=48296&nowrap=true#787179 It is continuing to act in this manner. I cannot prove that it is still acting deflationary because I don't do SaH and most of my computers are new since I last did do SaH so I would have no basis for comparison. There are also the corrosive effects of creeping run time like what is happening at GPU Grid where the run time to complete has been climbing but not the payment. By the way, the average host should be earning more because it is very likely 2.92 times more powerful than the computer that was used as the baseline for the 100 CS per day. Again, I cannot prove that because I don't have one of those old machines. But, if the clamp is to hold the "average" host to 100 CS per day, well that too is deflationary ... because the "average" host will slowly increase in power. When I get my next round of updated to my computers it is likely to be with 50% more cores and slightly higher in clock speed ... add that into your average ... and I push everything down ... Anyway, Eric, your words, not mine ... Again, I would be less bothered by this were some consistency applied. As I have stated in the past, if there were as much vocal and obvious pressure applied to projects that under pay as there is to those that overpay I would be right there with you ... but there isn't ... if there were, I would have seen the bumps in earnings ... If I switch my computing power from project to project I can see significant swings in earnings demonstrating we are not solving that problem either. And lastly, we do not have a good historical database/history but it sure seems that I have to keep adding newer and faster computing elements over the months to keep my average daily earnings in the same range. This problem was gushing evil spirits back in BOINC Beta and we tried to get it fixed back then, it wasn't ... if it had been, it would have been easy. Only a couple of projects running and a reasonably tight community ... now not only don't you have that you have years of the legacy of neglect and what to some seems to be double dealing on the part of the projects. Spend some time on the boards, almost any board and the unhappiness is palpable at times. It is easy to dismiss it as the rants of just a few malcontents like Paul D. Buck ... but all I have to do is point to the fact that your product cannot hold but 1/6 of the people that try it ... 1.8M sign up and all you/we can keep are 320K? No, I am not saying that the 5/6 leave because of credit woes, nor is it as someone suggested because I told one joke... rather I suspect because of something very simple ... "we" don't listen ... and I could go on for hours ... to no avail ... By the way, if you want, I will rebut the other post you made where you suggested that I take hold of some problems and solve them and I would get grateful support ... give me some time and I will tell you how those efforts worked out ... But, I best stop here, people with a history as long as mine can have it used to prove anything at all ... _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
