PrimeGrid: "3-second" bug fixed

A bug affecting multiple Windows users on all LLR projects has been fixed. The problem was causing jobs to "complete" and claim success after about 3 seconds. The problem was corrected by increasing the amount of permitted memory for each workunit. It seems as if some computers were exceeding the old limit. I'm not certain why that limit was being exceeded since BOINC was claiming that about 7 MB was being used and the limit was over 10 times as high, but increasing the limit has fixed the problem. For more information or to discuss the problem, please see this forum thread.

More...