next up previous contents index
Next: 3.5.2 Bodytrans Functions Up: 3.5 Dynamically Loadable Functions Previous: 3.5 Dynamically Loadable Functions   Contents   Index

3.5.1 Potential Descriptors

The potential descriptor is used in orbit calculations and a few N-body programs. These are actually binary object files (hence extremely system dependent!!), and used by the dynamic object loader during runtime. Potentials are supplied to NEMO programs as an input variable (i.e. a set of keywords3.7). For this, a mechanism is needed to dynamically load the code which calculates the potential. This is done by a dynamic object loader that comes with NEMO. If a program needs a potential, and it is present in the default repository ($POTPATH or $NEMOOBJ/potential), it is directly loaded into memory by this dynamic object loader. If only a source file is present, e.g. in the current directory, it is compiled on the fly and then loaded. The source code can be written in C or FORTRAN. Rules and more information can be found in potential(3NEMO) and potential(5NEMO) The program potlist(1NEMO) can be used to test potential descriptors. See Section [*] for examples.


next up previous contents index
Next: 3.5.2 Bodytrans Functions Up: 3.5 Dynamically Loadable Functions Previous: 3.5 Dynamically Loadable Functions   Contents   Index
(c) Peter Teuben