Bindings to the full hdf5 were supposed to be in the example set for bindings-DSL. It doesn't use pkg-config, though, and hdf5 developers didn't like the idea of adding support. [...]
That seems strange to me - pkg-config is such a useful system, and "support" for it is incredibly easy to add and practically zero- maintenance. [...]
I have to start asking for forgiveness. In fact, PETSc (http://www.mcs.anl.gov/petsc) developers refused pkg-config, not HDF5. I investigated several libraries for numerical and massive data processing, and most didn't had pkg-config support. I started asking about it at pkg-scicomp-devel debian list, and then Tollef (maintainer of pkg-config): http://lists.alioth.debian.org/pipermail/pkg-scicomp-devel/2009-September/00... http://thread.gmane.org/gmane.comp.package-management.pkg-config/345 After learning from them that pkg-config files belong to upstream packages, not distribution packages, I asked PETSc guys. Unfortunately, their petsc-maint mainling list doesn't seem to have an external archive. But Matthew Knepley, from PETSc, said that "we do not use that system. It's unreliable, and in my opinion, one of the worst designs possible." They (Matthew and Satish Balay) suggested that we "use the makefile targets which give the include and library information." I'm not sure how practical that is with cabal, but I gave up before trying harder and decided to focus on packages with pkg-config already available. I think 'hmatrix' package on hackage uses a script to check for dependencies, but I didn't read enough to check how portable that method can be.
Currently I've got a "template" hdf5.pc file in the source tree which can be customized and dropped into the appropriate directory. It's a lot less manual than it ought to be, but it's at least a lot less ugly than hard-coding my development machine's include and lib paths.
Maybe it's worth contacting HDF5 guys about that. Best, Maurício