Fletcher wrote:What we don't know is what the major differences are between and industry reliable physics engine such as WM and others like Solid Works Cosmos (built by thousands of engineers, physicists, and programmers) verses jim_michs bespoke programming. If his physics engine is more accurate and representative of the 'real world' interaction of masses in motion than the off-the-shelf physics engines then he is indeed a genius. Because nobody else saw what he saw as the deficiency or didn't recognize its significance to OU seekers.
FYI, as a former software-engineer I can safely say Jim doesn't necessarily needs to be a genius (as in: surpassing the intelligence of hundreds of engineers) to be able to program such a thing, while still able to produce a program surpassing the accuracy of any OTS physics engines. He "simply" has to meticulously program the machine for its task, so at least mildly above intelligence would suffice (he still could be a genius though).
What makes the task easier than perhaps expected lies in the things to be skipped:
*The highly complex drag/drop/connect/transform-capable user-interface (this saves at least 95% of the programming effort and bug-issues - this is not an exaggeration, it could be even higher)
*The multitude of design possibilities, scripting-possibilities, and information feedback: Just one program which basically animates the solution (perhaps some preprogrammed checkboxes, inputboxes, sliders etc.).
*The formula abstraction layer, because the formulas are already known for this specific task the program doesn't have to figure them out on itself.
*The real-time constraint: For an extreme complex situation the calculations could take hours to complete for just a one minute sim if that would be needed to get things right and keep the programming as simple as needed.
I think Jim is capable of such a thing....
...and still could have made a mistake (always a possibility).