For various (some dishonest) reasons I have been asked on numerous occasions to release the source code of programs available free from this website. To satisfy the genuine curiosity of those people who have asked for "the 'formula' for .... ", here are a few source code examples. Anyone who considers he understands the contents of these *.pas files is hereby granted permission to make full use of the information contained therein. If any future products should result no reference shall be made to me, the author, who is unable to provide any assistance in such matters.

Files are best read in non-proportionally-spaced text such as displayed by the mini word processor "NotePad".

To read or download a file click on underlined program name.

The number of possible ways in which a computer program can misbehave increases enormously with its size in bytes. A very large program is virtually certain to contain defects or bugs if only because of the impossibility of identifying and testing every one of these many ways. Debugging operations themselves are likely to introduce further defects. The importance of individual defects is uniformly distributed randomly between trivial and catastrophic.

A smaller program is a more reliable one. Unnecessary elaboration must be removed. Simplicity of source-code structure and understandability is the ultimate target. Small programs are better understood by both the programmer and during later maintenance. Testing will be more thorough and 'bugs' far less likely to remain. Consequently large programs must be composed of a number of small, independent, testable modules whose functions are precisely, unambiguously defined beforehand. Each module is then itself a complete small program.

The size of a complete program can greately increase due to built-in precautions to protect itself, the computer and its user from the consequences of accidental entry of 'illegal' data. For example, illegal data might result in the maths processor being asked to calculate the square root of a negative quantity. To omit protection might result in the program 'crashing', sieze-up of the computer's operating system and the need to press the re-set button. The number of possible ways in which illegal data can be entered is also very large. To lay traps, deal safely with them all and inform the user, is impractical and will itself increase the liklihood of bugs remaining in the finished product.

In G4FGQ's programs a bare minimum of protection is provided against entry of impossible data. Inevitably the user will make mistakes especially when unfamiliar with a new program. Mistakes are most likely to occur when entering numerical values. They fall into categories of incorrect measurement units, impossible physical proportions, ie, wire diameter greater than wire pitch, inclusion of non-numerical characters and using a comma for a decimal point. Obvious errors can be corrected, of course, by inspecting the data before pressing the Enter or (cr) key.

If you are running one of these programs in DOS where you have arrived from Windows, and it aborts, the system may not automatically switch you back into Windows. To return to Windows from DOS type EXIT (cr) against the DOS prompt. No harm is likely to come either to the computer or to the program as a result of being unceremoniously dumped back into the system.

To re-start a program which has just aborted while in DOS, from the same directory as the program is filed, just type the program's name against the DOS prompt followed by (cr). The .exe ending may be omitted. If you are not in the same directory then the name of the program must be preceded by the directory path to the program, the several items being separated by backslashes. Or you could change the current directory to that where the program is installed and then enter the program's name.

To Q(uit) or E(xit) from within a program do so by pressing the key indicated on the menu bar (Q or E) at the bottom of the screen. The option to quit is available only when the menu bar is visible. Pressing the quit key when the option is not available, if followed by the (cr) key, may indeed cause the program to close down but not in a very graceful manner.

The source-code language of G4FGQ programs is Turbo Pascal. Unlike BASIC and various spreadsheet calculators, when compiled into a stand-alone program, as they all are, the Pascal source text is inaccessible and cannot be modified. Programs can be run only as the author intended. Integrity and authenticity are preserved.

Subject matters are technical and are necessarily dealt with quantitatively. At the heart of each program is a set of mathematical functions and procedures. Properties of the 'parts' used to construct the model are stored within the program. Other procedures accept from the experimenter the details specifying the model. The program then proceeds with its analysis and its conclusions are displayed. Advice or warnings may be given to the experimenter as appropriate.

Entry of numerical input data obliges the experimenter to clearly visualise the system modelled and approach it logically. He will learn to ask himself the right questions. The relative importance of the various factors affecting behaviour of the system will be better appreciated. This conforms with one of the purposes of amateur radio activities: self-teaching in the art of radio communication.

But before the amateur who takes pride in being purely practical and ignorant of theoretical matters now leaves, don't go, these programs have been written with you in mind.

To use these programs you need no more than the ability to estimate the number of rolls of wallpaper required to decorate your shack. That's not maths - it's only arithmetic. Authors of articles published in radio magazines often confuse these activities.

Programs are intended to relieve the user of the labour and tedium not only of mathematics but of arithmetic too. But the quantitative visualisation of a model cannot be avoided.

If you have not yet mastered the sizes of enamelled copper wire in terms of milli-metres and still work in 'thou of an inch', or 'mils' you will be at a slight disadvantage - all the programs are in metric units. But down-load a program anyway and let me know how you get on with it. You'll be back.

Actually, most of my programs are more than 10 years old. Most of the basic research had been done and the mathematical procedures had been developed by then. Recent work has been to extend their range of application, to combine related subjects into single programs, to generally tidy up loose ends, and to make programs presentable and user-friendly.

At present, several more are in their final testing stages, being cross-checked for mathematical inconsistencies, bugs, etc. More effort is spent on later quality-control than on producing the first useable version of a program.

It's good practice to put a supposedly finished program on the shelf for a month or so, allow one's mind to forget it, concentrate on something else such as producing this web site, even spending some time on the 'bands'. It is surprising how previously unnoticed defects then become obvious. However, these discoveries do not improve one's self-confidence - how many other hidden boobs remain?

Until recently I have never actually finished a computer program. The self discipline of saying "It's done. Date it. Archive it. Make it available on the Net" has not come easily.

All programs ask for a set of numerical values which describe or define the matter in hand. Often the data are physical dimensions plus a choice of operating conditions. This initial array of input data is always on view at the top of the screen.

When a set of results has been computed it is displayed in the lower part of the screen. It is now a simple matter to select and change any one of the initial values while asking "What if I set it to this particular value . . . ?" Immediately the question is answered. A similar question can then be asked about any other input parameter.

Later programs, on the press of a key, have the facility to sweep input parameters slowly in fine increments over wide ranges while immediately observing the effects on output data.

There is no need to go back to the begining and re-enter a complete set of initial values although that particular option is available when needed. The response or behaviour of complicated systems when subjected to changes in inputs can be analysed and appreciated very quickly. Not only will programs be of practical use - by encouraging the user to become familiar with magnitudes and ranges of the essential factors they will also be of educational value.