Michael Schlottke-Lakemper

Results 366 comments of Michael Schlottke-Lakemper

I am all for consistency and descriptiveness in naming things. Since consistency is broken anyways (it would have to be `comm_set_errhandler!` if I'm not mistaken about the naming scheme), I'd...

Interest yes, but no capacity at the moment 😞 For me to write this section, I'd first have to find the time to set up and try out how to...

Re/ the first option: Not sure if we need provide this with the described level of automation. I think the ability to generate the `const.jl` file (and put it in...

This sounds like a potential solution. However, I strongly suggest that if we use this path, the ability to compile a C program to figure out the constants should be...

As just discussed during the monthly Julia for HPC call (thanks also to @mkitti @giordano @williamfgc for the discussion): Maybe it would be sufficient for now to restore the ability...

> I'm not opposed to it, but I do wonder as to the utility: are there more ABIs in the wild? We had this discussion yesterday as well. For the...

As a follow-up questions, was it intentional to remove all typed pointers for https://github.com/JuliaParallel/MPI.jl/blob/e4479cfe16f1fdbbdab7423eca03c98764b0c671/src/consts/openmpi.jl#L245-L248 While in `mpi.h` the first two should be pointers to `int`/`Cint` ([ref](https://github.com/open-mpi/ompi/blob/01d5a86c6021a9610b512e06679c11ba8974a833/ompi/include/mpi.h.in#L554-L557))?

> > * `MPI_Aint` to be of `ptrdiff_t` type, which is implementation defined (Q: isn't that a potential issue that it can vary?) > > > > * `MPI_Fint` to...

> Would you mind making a PR on the existing const files? What do you mean exactly? One that fixes the types I've pointed out above?

It's purely a matter of personal preference, but I'd give the current version (which just introduced another breaking change) some time to mature for a while. If there are no...