Survey - what language are you coming from?

[Pablo Ripolles]

== Quote from Walter Bright (newshound at digitalmars.com)'s article
> Pablo Ripolles wrote:
> > little C but lots of Fortran 90/95 and Python, essentially for scientific
> > computing and visualization (numerical simulation)
> I'm glad to see Fortran represented here!

C was the first language I started with, I just played a bit with it.

I must say that I am mainly into the research and development of "good" and
"stable" designs for PDE solvers (Fluid Mechanics/CFD, Solid Mechanics/CSM,...).
In this context performance is critical.  It's so critical that the general
tendency is to let design related goals (maintainability, extendability,...)
completely aside.  I am sick of the often seen, unbearable, inflexible and old
fashioned "simulation codes" that sadly tend to become "legacy codes"... if ever
producing some legacy I choose legacy design.

Donald Knuth once stated "Premature optimization is the root of all evil".

The classical expectation that at the lower levels (leafs) of the "calling tree"
you barely need abstraction and therefore anything might be implemented (without
twisting design) using elementary language constructs just fails.

In this state of mind my journey has been:

Fortran 77 -> Fortran 90/95 -> Fortran 2003? (still waiting for acceptable
implementation) -> Python (Multiparadigm, abstract enough and brilliant ascetics
but too slow) -> D (Multiparadigm, abstract enough and FAST!)

PyD becomes crucial for the _late optimizations_ of my legacy Python designs.

Why not C++? I was reluctantly about to stick my heat into it when I discovered :D