One of the things that you’ll have to pick up when studying Physics is programming. When I started my undergraduate education in the early 1980’s, the University of Heidelberg in Germany which I attended at the time, offered some programming electives which were terrible. You sat in front of a terminal that was connected to some kind of mainframe and tried to somehow convince the thing to run some code snippets. All output was text based and FORTRAN was the language of choice in those days and I was rather turned off by the whole experience. A couple of years earlier Commodore had introduced the C64, which spoke BASIC and even had an optional PASCAL compiler. By the time I did my undergrad studies in Physics I had already written several PASCAL programs that had basic graphics output. One of them was a numerical treatment of the gravitational three-body problem and orbits were shown of the computer screen. So going back to a text-only interface on a mainframe terminal was not appealing at all to me. Later in my training I used VAX/VMS computers for nuclear physics data acquisition, data analysis and event simulation purposes. My Ph.D. thesis was written on a VAX workstation using Tex/Latex and custom graphing and plotting software.

After deciding to work in industry (instead of academia) I found that PC’s were quickly becoming the dominant computational device (instead of Digital’s VAX, SUN, Silicone Graphics, etc.). For a couple of year in the early 1990s I was working as a project manager and had several programmers report to me. Partly out of frustration with the pace of progress on several software projects, I decided to get involved in programming again. The question was which programming language/compiler package best suited my personal tastes and the tasks at hand. I evaluated MS Visual Basic but did not like the fact that the code was not compiled but interpreted and that you needed to distribute a whole bunch of files along with your application to ensure that your program worked on a given PC. I did like the visual programing interface that MS VB offered. In those days Borland was the leading compiler manufacturer and I had looked into Borland’s Turbo Pascal and it’s (at the time) revolutionary Object Windows for DOS interface. Borland had just introduced the Delphi visual programming interface and this seemed like a tool worthy of evaluating. I was immediately hooked. Just like the original DOS compiler, Delphi code compilation was lightning fast. It’s visual programming interface made is simple and efficient to use. Tasks that took the programmers which were reporting to me weeks, I could accomplish is a day or two. The only drawback was that I was the odd man out, in the sense that everybody else around me used Borland’s C++ compiler (and later Microsoft’s Visual C++).

Off late, Delphi has been in decline, due partly to the ascent of C/C++ as the language of choice and partly due to Borland’s (now CodeGear) failures in marketing. I still use Borland Delphi whenever possible to accomplish simple and moderately complex programming tasks but also keep my C skills sharp in case I need to modify/port someone else’s code. Borland (CodeGear) has a historical programming site and this article captures the initial excitement that the Delphi development environment created. If you like to program in Delphi you’ll like to read this.