Re: [h5md-user] fields of observable group

[Peter Colberg]

On Mon, Aug 29, 2011 at 04:09:21AM -0400, Pierre de Buyl wrote:
> >>>On Fri, Aug 26, 2011 at 04:17:27PM +0200, Felix Höfling wrote:
> >>>>If there are no objections, I will change "samples" to "sample".
> >>>
> >>>No objections from my side.
> >>>
> >>>As I proposed earlier (in private mail before this list existed),
> >>>it would be convenient to rename “coordinates” to “sample” (or
> >>>“samples”) in the trajectory group, which avoids the necessity
> >>>of special handling of trajectory versus non-trajectory data.
> >>OK for me too for "sample" everywhere. This will simplify H5MD
> >>codes :-)
> >>
> >>Pierre
> >I'm still a bit unsure, whether singular or plural is nicer. This
> >has been discussed in length before, e.g.,
> >http://stackoverflow.com/questions/338156/table-naming-dilemma-singular-vs-plural-names
> >
> >A useful rule is mentioned there, which we might adopt for H5MD
> >(and have done so already, I think):
> >
> >In his book "SQL Programming Style," Joe Celko suggests that a
> >collection (e.g. a table) should be named in the plural, while a
> >scalar data element (e.g. a column) should be named in the
> >singular.
> >
> >That means a dataset is named in singular (e.g., 'sample'), while
> >a group is named in plural (e.g., 'observables').
> >
> >Another useful guideline may be this one:
> >http://justinsomnia.org/2003/04/essential-database-naming-conventions-and-style/
> 
> A coherent naming scheme would then be the following ? (with the
> capital S lowercased, indeed)
> /observableS/energy/{sample,step,time}
> /trajectorieS/colloid/position/{sample,step,time}

/trajectory is not a collection of independent trajectories, but the
trajectory of the whole system. Therefore singular is consistent with
the H5MD naming scheme (which we have to put in writing in the draft).

Peter