Dave, I applied all of the latest updates from the main trunk under subversion, and ran with your control.in, but none of the events looked unusual to me. In particular, the events you listed (84, 464, 493, 532) all had copious numbers of CDC hits. Can you remind me what you mean by no hits? Literally that, just a Monte Carlo header and no contents in the <hitsView> element? I have now initiated a complete rebuild of the exact release-2009-05-27 and will try that one as soon as I can get it to build. Right now, I am getting hit with "unrecognized Red Hat release" messages from the BMS make system. -Richard J. David Lawrence wrote: > Hi Richard, > > This problem with events occasionally showing up with no hits seems > to still be present, albeit at a much lower level. I reported on this a > few weeks ago and promised I'd send you the details needed to reproduce > it. I'm now finally getting around to it. > > I have attached a control.in file that exhibits the problem in > events 84, 464, 493, and 532 when run with release-2009-05-27. (Note > that this is release-2009-05-27 as of todays patch, not the original > release-2009-05-27). > > I should note that these events show the problem when running it on > Linux. I see the same problem when running on Mac OS X, but for > different run numbers. This behavior is different than when I ran across > this problem originally where the same event seemed to show the problem > on both platforms. I don't know if this is because I'm making a mistake > or something has changed with the random number generators used on the 2 > systems. If you don't see the problem for the event numbers I indicated, > just let me know and I'll look a little deeper. > > Regards, > -David > >
S/MIME Cryptographic Signature