So… C++? Delphi? Markov Chains?

I have a line on a job that involves porting some code that was originally written in R, then in Delphi, and now the researcher wants it re-written in C++, turned into multi-processor/multi-computer friendly (using MPI?), and turned into a plug-in for R. The program as it is now is pretty primitive – he apparently just puts a bunch of parameters into the actual Delphi code then recompiles and runs, and it outputs into a data file. Obviously the first step would be to have a wrapper program that gets the parameters from a data file, and later a wrapper that gets the parameters from however R passes them to plugins.

It’s been a while since I used C++, and the language has changed a lot since then. Name spaces, STL, Boost, auto_ptr, all this stuff is new to me. It’s going to take some frantic reading to get up to speed. Even worse, I have to read the existing code, which means learning a bit of Delphi/Pascal. And I’m going to have to find a decent IDE for C++ – although the consensus on StackOverflow seems to be to go back to the way I’ve always worked until I started using Eclipse last year: gvim, make, gdb, and a web browser open to the man pages.

Even better, the job would mean working from home. The dogs will be happy about that.