User:Kevin baas~enwiki/tex to cpp

intro

A TeX to C++ converter. Standard functions like hamiltions, calculating eigenvalues, etc. are actual functions, all operations and functions work. There is a most abstract class for all math variables, and one can specify more specific such as tensor or vector. Laplacian would be interesting - would make things like computational fluid dynamics, or simulations of reaction-diffusion spaces, "short" programs. Function & operator overloading is used to process the variables correctly. More abstract types are disambiguated via context, determined by the class of variable that overloaded functions return, if the class of the variables it takes as parameters and the TeX are insufficient. (tensor multiplication must return a tensor, etc.) derivatives/antiderivatives - the most basic is w/respect to "time" (which is not neccessarily real time, but a parameter specific for the "module" that determines processing order and scale). Dot notation and the variable t identifies time. All time derivatives/antiderivatives, etc, for a given module, are processed in parrallel (producing x', y', z'..., which overwrite x,y,z... when all computations for the iteration complete) each module has it's own "time", which may be a function of real time, and/or any variables within scope.

module scheduling

• priority levels
• like aynchronous distributed simulation

a visual TeX editor (instead of having to write in tex)

• for example, can select from context menu, matrixes, operations on matrix like transpose or inverse, etc.
• function keys switch alphabet of keyboard (english, latin, greek, hebrew)

c++ to tex (perhaps w/hints, or remarked code) a way to deal with structs; mixed(composite) classes (some strings, some tensors, etc.); object hierarcy

• i guess nothing really needs to be done here - just use dot notation, as usual.

philosophy of architecture

The task of the computer is to simulate any given complex dynamical system. Programs are specifications of complex dynamical systems.

modules

(execution/time-parameterized/entropic/information-dynamic) modules, determine program flow; "processing". a module is a member of a class

class subsystems {
  module fafd( time function, execution condition);
  module ffsd( time function, execution condition);
  
  tensor agsg, gfdg;
  ...
  }