Babylonian Language Confusion or Optimized Multi-Paradigm Programming?

In quant finance forums, the questions on the right programming language pops up quite frequently. Like Why Do Quants Like C++ . In my contributions, I emphasize on analzse-requiremenst-and-choose-the-adaquete-programming-paradigm (not the other way arond). It is not always understood that one-size-fits-all does not fit for integrated industrial scale applications.

And recently, I asked myself, what was the result of this approach in the UnRiskverse?
In lines of code, knowing that this is not saying much about comprehensive quality features. However:

Mathematica: 155.000
C++: 610.000
Java: 320.000
Java Server Pages: 50.000
SQL: 170.000
CSS: 4000
XML: 7000
others: 10.000

Yes, over 1,3 Mio (growing quickly, because from release to release, we are able to develop more UnRisk in UnRisk).

We are doing comprehensive tests for the next release now, and in a first view it might look difficult to do this in such a hybrid environment.
But we are producing Mathematica Notebook Logs from any action automatically and re-feed them. The same mechanism helps us in support to detect anomalies and errors unprecedented quickly.

No comments:

Post a Comment