[MacPorts] #66208: hdf5 @1.12.2_0+cxx+gcc12+hl: Port hdf5 is still broken (cxx_stdlib mismatch) after rebuilding it more than 3 times.

[MacPorts]

#66208: hdf5 @1.12.2_0+cxx+gcc12+hl: Port hdf5 is still broken (cxx_stdlib
mismatch) after rebuilding it more than 3 times.
------------------------+----------------------
  Reporter:  massonseb  |      Owner:  eborisch
      Type:  defect     |     Status:  assigned
  Priority:  Normal     |  Milestone:
 Component:  ports      |    Version:  2.8.0
Resolution:             |   Keywords:
      Port:  hdf5       |
------------------------+----------------------

Comment (by cjones051073):

 Replying to [comment:13 oliviermarti]:
 > My 2 cents :
 >
 > I use (and massonseb too) a library
 (https://forge.ipsl.jussieu.fr/ioserver/) that mixes C++ and Fortran code.
 We use it in several Fortran codes. The library needs parallel netcdf, on
 top of hdf5. We commonly use these code on Linux systems with Intel or GNU
 compilers.
 >
 > Months ago (don't remember when), I've installed openmpi, hdf5, netcdf
 and gcc12 with MacPort. All compiled with gcc12 variants. It installed
 smoothly. And I succeed to use the library and my code. So this is a
 solution that has worked.
 >
 > Mixing clang and fortran, I can compile, as massonseb, but the test
 cases of the library does'nt run. I've never succeed to mixed clang and
 GNU fortran. Is this really possible ?

 Mixing clang and gcc gfortran to build c/c++ and and fortran code indeed
 works just fine. Most ports in macports that need to build c/c++ and
 fortran infwct do this all the time. OpenBLAS is for instance one. If
 youmdo not believe me go check the build logs for it ;)

 So, again, I see no needmfor you to be forcing the use of gcc12 to build
 c/c++ here.
 >
 > That's why having back a working hdf5 and netcdf with gcc12 will be very
 useful for a least a few people.
 >
 > Thank's !

-- 
Ticket URL: <https://trac.macports.org/ticket/66208#comment:17>
MacPorts <https://www.macports.org/>
Ports system for macOS