During the past several months (almost seven to be precise), I have had countably infinite attempts at getting some programs to work in our research group machines — some of you have read/responded to my previous entries (recent, not so recent) and some others have helped via email. Since VASP v4.6.28 successfully compiled (and executed) during Christmas 2005 and nothing like that happened in Christmas 2006, I was slowly beginning to doubt my belief in miracles, computer sentimentality and touch therapy - in a nutshell, first one refers to things falling into their appropriate places (like successful compilation and/or execution) around certain important days of the year while latter two mean that most computers are sensitive to who touches them.

After some correspondances, one of the developers - Alexandre Rocha - visited Michigan Tech during the last week and helped fix many issues - some related to compilation, some related to input file parameters and such. It was a pleasure working with and sharing knowledge with him (although I was the primary beneficiary of any/all such discussions), learning about many more things than I had originally anticipated…

With (L-R) Walter Nadler and Alexandre Rocha … while hiking/trailing towards Canyon Falls along Sturgeon River

All these means that our research group machine now has a properly working (parallel) version of SMEAGOL v1.0b. This in turn means that the same machine has properly working versions of MPICH1 MPI as well as parallel SIESTA v1.3f1p. In addition to these, planning (and implementing) expansion of our beowulf cluster can be undertaken before Summer semester begins.

Thanks again to all those who offered help, suggestions, comments - my belief in miracles (an Easter miracle this time!), computer sentimentality and touch therapy have all been restored. If anyone reading this is having similar issues, please do leave a comment using the Comment Form below and you will receive the working copy of arch.make and Makefile.