[h5md-user] Update

[Pierre de Buyl]

Hi all,

I just re-read the whole spec and find it quite mature now :-) I just pushed a
few commits but mostly for clarification and consistency.

FYI, I started to implement a C library (for now, it is just a C file with a
header, as it is often easier to integrate into other programs) at
https://github.com/pdebuyl/ch5md

It can create a H5MD file with time-(in)dependent data (including a "append"
routine for the time stepping). A major issue that I see coming is that many
simulation programs have a layered system for accessing particle data and that
my routines require regular contiguous arrays.

Anyway, there is an example program that works. Documentation is quite poor at
this time.

I guess that it is time to ask for "final comments" to the list :-)

Cheers,

Pierre