From: schuelaw@carrot.whitman.edu (Albert Schueller) Newsgroups: comp.parallel.mpi Subject: Re: A real puzzler? Date: Fri, 16 Jul 1999 22:56:31 GMT Organization: Posted via RemarQ, http://www.remarQ.com - The Internet's Discussion Network Message-Id: References: <7mo4l8$knk@dfw-ixnews8.ix.netcom.com> Reply-To: schuelaw@whitman.edu User-Agent: slrn/0.9.5.4 (UNIX) Xref: ukc comp.parallel.mpi:5348 On 16 Jul 1999 20:28:24 GMT, Gene Wagenbreth wrote: >The best way to diagnose a problem is to shorten your code as much >as possible such that the problem still occurs. You ought to be >able to post a very small code that exibits the bad behavior. If >not, at some point you will find that if you delete a certain >statement or group of statements, the problem goes away. The >content of that group of statements is a major clue as to what >is going on and how to fix it. > >Sorry if its a pain, but thats why programming is so frustrating. >You go to do something straight forward that you think will take a >day, and run into a problem that takes 3 days to fix. But you >gain information that is useful in the long run. You're right of course, I just haven't been able to get back to it. Essentially you've outlined my strategy for tracking this thing down myself. Mostly, I was hoping that my brief description of the problem would ring a bell with someone and they'd just pipe up saying, "Oh yeah, I ran into that and here's what I was doing wrong..." When I get back to it next week, I'll be sure and post the solution of this problem. Albert