<- Click here to expand.
The program DIRAC2.BAS is the successor to DIRAC.BAS, and is the current effort directed at the computer simulation of wave-particle duality. You can download and view this program with QBasic, but you cannot run it. DIRAC2.BAS requires modules VEC4.BAS, DWAVEF.BAS and RANDOM.BAS. Microsoft QBasic cannot handle modules. You need QuickBasic 4.5 or Visual Basic for DOS (they do the job equally well).
VEC4.BAS contains a set of functions to do 4-vector arithmetic. You can download this program with QBasic. If you run it then you will see some local tests being performed to check that VEC4.BAS works correctly. Otherwise, you need QuickBasic 4.5 or VBDOS to link this module to DIRAC2.BAS.
DWAVEF.BAS contains a set of functions to do arithmetic on the Dirac wave function, which has four complex components or eight numerical components. Again, you can download it with QBasic and run it to do some local tests to check that all those alphas and betas really do anti-commute.
RANDOM.BAS deals with the generation of random numbers. It maintains a User Activity Log which times the user's presses of various buttons, and uses this to partially reseed the random number generator. This ensures that the randomness is 'really' random.
If you do have QuickBasic 4.5 or the equivalent available, then download each module with QBasic. Save it from within QBasic on a separate area of disk. Then use your QuickBasic to start DIRAC2.BAS as the main module, and load in VEC4.BAS, DWAVEF.BAS and RANDOM.BAS as auxiliary modules. Your .MAK file should end up like this:
DIRAC2.BAS
VEC4.BAS
DWAVEF.BAS
RANDOM.BAS
All the above modules have been compiled to DIRAC2.EXE, available here in ZIP form ...
DIRAC2 currently has a local randomness model installed, and we have been trying to simulate the capture of an electron in a potential well with this local model. This has been a failure, probably because we have no way of beating the Courant-Friedrichs-Levy condition, which is what we may need to be able to do in order to create a new eigenstate ex nihilo.
A nonlocal randomness model will be installed and we will try again. One would have thought that a local model of randomness would be adequate for the computer simulation of the capture of a consolidated wave packet in a potential well, but perhaps this is not the case: we need nonlocal randomness even for this. Actually showing that this is true would be a useful intermediate goal and this is what will be attempted next. Here we have something worthwhile to try which is within the range of an amateur using a Pentium-class computer. As a project it is also quite orthodox compared to other things proposed on this website.
When an electron is captured in a potential well, surplus energy is dumped as electromagnetic radiation. This is transverse radiation, and so we need to add transverse randomness to get the necessary broken symmetry. The electron ends up in a new eigenstate, which is a holistic process which requires nonlocal longitudinal randomness added to make it happen. So we have the doctrine that two randomisers are necessary, one transverse-local and the other longitudinal-nonlocal. Switch off either randomiser, and we cannot simulate the capture. Showing all this to be true is a feasible project which would be the first chink in the armour of quantum mechanics.
It has to be admitted that before we stated this doctrine that two randomisers are necessary in about September 1998, we were losing our way a bit. Now we have a clear stepping stone to cross, and the means to do it, so we can get on with the effort. If we succeed, then we will be able to do something that no one else can do, something essentially nonlocal, and we will have got round Niels Bohr's argument about the indivisibility of the quantum of action. Here is nonlocality in a tasty bite-sized chunk.