Kevin Paul

Sorry for the delay. Yeah. It's not a decorator yet. But it could be easily transformed into one. As for taking the client in as input, I would assume that...

Ok. I'm looking at this a bit today, again. First, I have created a mock-up of the solution I described above in https://github.com/dask/dask-mpi/pull/110. I'd appreciate any comments on that solution...

Now that I've had some experience playing around implementing #110, I've discovered a new way of accomplishing the same things. And I like this approach even more. The original intent...

@betatim I agree with you entirely. My intention is to make the `execute()` function part of the public API of the `dask_mpi` package. So, you will still have an `"initialize()`-like...

Incidentally, the `mpirical` package (https://github.com/kmpaul/mpirical) is the closest success I've ever had at trying to accomplish something like "run this function with MPI". It _still_ uses `mpirun` or `mpiexec` under...

Great. Then I understand you. And that should be doable with the `execute()` function. I hope that is satisfactory.

NOTE: The `execute()` tests are not complete. All of the CLI and `initialize()` tests should be ported to use the new `execute()` method, but when the CLI is modified to...

Note that even with the merge of #89, these errors still occur.

One potential solution to this is to *not* use an `atexit` handler to shut down the cluster. Instead, require that the user shutdown the cluster *manually* at the end of...

## Possible API changes If we have to break backwards compatibility, then I want to suggest some other changes that might work with the suggested changes above. With @joezuntz's changes...