* Rob Latham <[email protected]> [2012-05-24 16:01:12]: [...]
See, now you have a list of options, and you show them to someone like Dries and he says either "you forgot about XYZ" or "oh, i see now, all of those are gross"
That's a bit harsh :( My opinion is simply that, if the only thing we get from putting it inside the MPI library is to avoid the progress thread/poking issue, it might be better to try to provide a solution for the progress thread instead. After all, even if it is built into the MPI library, there will very likely be no progress unless somebody is calling into MPI. And on the other hand, if this can be done equally well on top of MPI, it should be done on top as that is the fastest way to provide this new functionality to the general public (without having to wait for MPI-3.1) Dries