Ivan Pribec

I would also try to stick with Fortran and C for the beginning. Not only can it be a reference implementation, but hopefully also a good example of what modern...

Having just said that, I would like to hear your opinions on are we willing to wrap any F77 code with modern interfaces? I can guess the opinion of @jacobwilliams...

Is there anything that could be done to improve the syntax highlighting? In fixed-form Fortran `C` in the first column indicates a comment, and the first 6 columns are reserved...

Are the METIS routines called from Fortran? If so, I wrote a module with the C bindings to be able and interface safely from Fortran: https://github.com/ivan-pi/fmetis

The matching C prototype can be found with the command `gfortran -fc-prototypes-external -fsyntax-only lsame_example.f` (use `>` to redirect output to a file) ```c #include #ifdef __cplusplus #include #define __GFORTRAN_FLOAT_COMPLEX std::complex...

Also note the attribute `pure` which is effectively the portable Fortran version of what the GCC `__attribute__((pure))` tries to achieve; it is also a weaker version of the `__attribute__((const))` added...

Here's a test driver for the old and new code: ```fortran C test_ascii.f program test_ascii use, intrinsic :: iso_fortran_env, only: int8 use, intrinsic :: iso_c_binding, only: nl => c_null_char implicit...

I'm curious what precisely do you mean by trouble? The routine would remain in fixed-form, it is standard Fortran 95. Besides improved readability, the behaviour of the routine is made...

Apologies if I over-reacted. So I understand that preserving the code `f2c`-compatible is the main reason limiting BLAS and LAPACK from introducing "new" (and arguably simpler) Fortran syntax? More generally...

So I just gave this a try with my system-installed f2c (Ubuntu 20.04), and found out that `f2c` supports `iachar`: ```fortran ! lsame.f ! compile with: gfortran -Wall -std=f95 -o...