Re: [h5md-user] box data as part of trajectory/position

[Felix Höfling]

Am 16.11.2012, 21:48 Uhr, schrieb Peter Colberg  
<address@hidden>:

> Hi all,
>
> My subscription to the h5md-user list was disabled in August, due to
> the removal of an old email address. I have caught up with the list:
>
> On Tue, Sep 18, 2012 at 12:01:12PM +0200, Felix Höfling wrote:
> > Peter and Olaf: are you still following the list? Do you have an
> > opinion on that?
>
> I suggest the box information remains in /observables/box. I see no
> benefit from moving it to /trajectory, other than rendering existing
> simulation files incompatible with H5MD. I accept Pierre's earlier
> motivation that box is a macroscopic property.
>
> Peter

Hi Peter,

It's not so much about whether box is a macroscopic property or part of  
the trajectory data. My suggestions supports both views as the box data  
may equally exist in /observables and /trajectory.

The main motivation is the issue with potentially different sampling  
intervals of different particle groups, which remains unresolved if only a  
single box group is stored. In my proposal, the box information relevant  
for each snapshot of a particle group is stored, ensuring data consistency  
and simplifying the access of particle coordinates. Further, it avoids  
dependencies across different root groups.

I think these 3 benefits deserve reconsideration of your opinion. We  
should try making the H5MD data structures as clear and generic as  
possible with simple solutions for reading and writing.

BTW, if your old files shall be read by a future H5MD reader that expects  
the box information in trajectory, you may simply add a link from  
trajectory to observables. Such links are part of the proposal.

Felix