[Proj] Accurate algorithm for geodesic calculations

[Karney, Charles F.]

> > Ahh..., but it's not the case that anyone who can write passable C++
> > code can write equivalently decent C code.
>
> Of course, that is an absurd statement.  C very much a subset of C++
> and using C as an alternative is merely ignoring the superfluous parts
> of C++ that do little in facilitating most mathematical intensive
> procedures.

C may be a subset of C++.  However, people learning C++ as a new
language do themselves a favor by NOT learning about

    storing strings in char[]
    strcmp, strcpy, index, rindex
    malloc, free
    much of the C use of pointers
    much use of macros (like those for max/min and numeric constants)

because C++ provides strings, the STL, etc.  So, your assumption that a
C++ programmer can either (a) write code in the C subset of C++ or (b)
trivially strip his code down to a C core is simply not true.

> Be a saint and publish in pdf!