|
|
|
|
|
The term "legacy code" probably evokes some romantic thought
of a beloved and ubiquitous piece of software passed from
generation to generation of programmers and users like a
treasured family keepsake. In the real world, however,
legacy codes and systems are often thought of as something
we're stuck with rather than something bestowed upon us. In
the case of scientific codes, we're going to use the term
to mean a code that may not necessarily be state-of-the
art in terms of the computer technology it employs, but is
still seen as a vital research tool by many scientists.
While the performance and functionality of the legacy codes generally holds up well, ease of use is another matter entirely. A great many of the computational chemistry programs in wide-spread use handle input and output in the form of flat text files. This means that to run a simulation, the scientist must work out what his beginning configuration should be, type it in, and then type in all of the simulation parameters. While this not may seem to be so bad at first blush, imagine creating an input file for a large biomolecular system with a few thousand atoms. To call this task tedious is a sizable understatement. Well, luckily for the new breed of simulation software users out there, there appears to be a major move toward integrating the existing codes with newer technology. This approach allows a user to interface with a much more friendly environment on the front end while still accessing the power of the original applications on the back end. One of the developing methods of accomplishing this feat is to fold the existing code inside of Java-based wrappers, and then use the extensive Java GUI classes (AWT or Swing) to build a sophisticated user interface. But beyond presenting a simple point and click program interface, Go To Page: 1 2 |
|
|
|