H5Part development and discussion

[John Biddiscombe <biddisco AT cscs.ch>]

Achim
> - Experimental code added to define the step name and format of step number
>   as proposed/requested by John Biddiscombe from CSCS. This code is still
>   work in progress and the feature cannot be used in this release.
>   
That's great. Thanks. Incidentally, I'm sitting on a few changes in my 
local copy which I'd like to contribute
1) Builds/runs on win32
2) Added option to use INDEPENDENT_IO instead of COLLECTIVE_IO, I need 
this when using it on a system with a non-parallel file-system, using 
collective IO causes too large performance drop on our cluster.
3) Some fixes for usage in a stand alone process with no MPI (controller 
null checks)
4) Cmake support
5) Fix a crash when one process has zero particles to write, but others 
have some. Assumptions about data/mem space were wrong.
6) Added a function to query IsTimeStep Present, so that I don't 
overwrite an existing one if redoing the same time
7) some other minor tweaks I've forgotten about.

If I update my copy to work with your 1.3.2 (svn head), can I send you a 
large patch with all of it in, or would you prefer smaller patches which 
address individual issues (much more work)

JB