Re: GSL and (visual) c++

[Brian Gough <bjg at network-theory dot co dot uk>]

Hi,

Andreas Krueger writes:
 > a) Is there something similar in c++ ?

There has been some discussion about this in the mailing list archive
http://sources.redhat.com/ml/gsl-discuss/ --- nobody is writing
anything.

 > b) How do I use it in general c++?

See the section 'Compatibility with C++' in the manual for some
information.  Basically, it should work from C++ directly as extern
"C" { .. } is present in the header files.

 >    Why is it platform specific for non-Windows systems?
 >    (as said on http://sources.redhat.com/gsl/)
 >    Isn't c/c++ independent of the OS?
 >    How can I implement it into my "Visual-C++"-environment?

It should work ok with any ANSI C/C++ compiler.  The platforms listed
are ones where it has been tested and reported to work.  There was
some discussion on the list about how to compile with Visual C++
(search for 'visual' on the list archive).

 > c) Can I use only parts of it? e.g. statistics only

Yes. This is usually possible, but there are dependencies between some
parts. Look at the header files to check.

 > P.S.: I couldn't find any support for "error propagation",
 >       perhaps you can also implement my "class measure<T>":
 >       www.andreask.de/cpp/measure

There are some functions in specfunc/,  e.g. gsl_sf_multiply_err_..,
gsl_sf_exp_err_..., etc. I'm not sure if they are a complete set.

If you'd like to write a C version under the GPL we could consider
including it or pointing people to it. These things are less
convenient in C due to the lack of operator overloading of course.

regards
Brian Gough