A B.Cereus case of assignment poaching...

[bcp19]

One other thing to consider is that the person who 'poached' the assignments may have been legitimately given those assignment a long time ago, but if they were being worked on with an offline machine and they were not up to speed on extending them, they could have expired and then been reassigned. One thing that you cannot tell from primenet is if an exponent had ever been handed out before and 'returned'.

[tcharron]

I agree that old exponent processing could be the issue, but it isn't in this case.

I see that 11 of my (thousands of) TF assignments that did not receive credit when returned. For example, four of these (85755337,85205221,85200743, and 85473979) were completed by the same user before I returned them, and all of those were done with mfaktc around the same time, and to high bitlevels.

This is a small percentage of my TF work, but it does feel like some inefficiency in what should be a community effort.

[LaurV]

Then either try to communicate with the guy, eventually by a public post, or PM if you know his forum ID (assuming he is here on the forum) and see that one of you two give way to the other. The cleverest will give up. (old Romanian proverb, I believe in English is something like "Sometimes the smartest choice is giving up". I don't think that giving up should be your primary method for dealing with problems. But well...).

According with what I see here, people are still doing CPU factoring on 85M range, which is a totally waste of resources (in my opinion, don't throw the stone!). If few of us put our GPU farms on them, the bitlevel for whole the range will rise ten times faster. Better the people TF-ing there would do some P-1 with those CPUs whose clocks they waste.

edit: the four exponents you blame were completed by 3 different people, and two of them are working those 4 exponents since months, they raised the last 5 or 6 bits, I don't believe it has anything to do with poaching. For example the first expo and the last, those guys are reporting a bitlevel every 2-3 days since ages.

[bcp19]

Originally Posted by LaurV

Then either try to communicate with the guy, eventually by a public post, or PM if you know his forum ID (assuming he is here on the forum) and see that one of you two give way to the other. The cleverest will give up. (old Romanian proverb, I believe in English is something like "Sometimes the smartest choice is giving up". I don't think that giving up should be your primary method for dealing with problems. But well...).

According with what I see here, people are still doing CPU factoring on 85M range, which is a totally waste of resources (in my opinion, don't throw the stone!). If few of us put our GPU farms on them, the bitlevel for whole the range will rise ten times faster. Better the people TF-ing there would do some P-1 with those CPUs whose clocks they waste.

You have to consider one other factor here... any CPU with 256(hmm, not sure if it was Kb or Mb) of cache CANNOT do P-1. I was testing a Sempron<sp?> and it kept getting assigned TF in the 85M range due to the fact that it's cache was too low.