[CF-metadata] stricter CF conventions

[Brian Eaton]

Hi Juerg,

The reasons for many of the decisions concerning required versus optional
metadata in CF is summarized in the last 2 paragraphs of section 1.1 in the
convention:

> This convention is designed to be backward compatible with the COARDS
> conventions [COARDS], by which we mean that a conforming COARDS dataset
> also conforms to the CF standard. Thus new applications that implement the
> CF conventions will be able to process COARDS datasets.
> 
> We have also striven to maximize conformance to the COARDS standard, that
> is, wherever the COARDS metadata conventions provide an adequate
> description we require their use. Extensions to COARDS are implemented in a
> manner such that the content that doesn't depend on the extensions is still
> accessible to applications that adhere to the COARDS standard.

The reason that units and the positive attribute are the required metadata
for determining a coordinate type, and axis and standard_name attributes are
optional, is for conformance to COARDS.

Brian

On Wed, Jul 30, 2003 at 12:38:31PM +0200, Schmidli Juerg wrote:
> Dear All
> 
> I think the CF-conventions could be made significantly simpler, 
> clearer and more precise without loosing anything of their 
> generality. The backwards-compatibility to COARDS makes the
> implementation of a general CF-compatible application 
> unnecessarily complicated, due to the different ways of 
> achieving the same goal (e.g. identification of the 
> coordinate axes). No large changes would be required.
> Changing some of the recommendations into requirements
> would suffice.
> 
> Consider, for example, the identification of the coordinate 
> axes and their type. Presently, there are three ways to 
> identify the coordinate axes:
> 1. units and positive attribute
> 2. axis attribute
> 3. standard_name
> 
> A simpler strategy, and easier to implement would be:
> - to identify the coordinate variables using the axis attribute
> - to determine the type of coordinate using the standard_name attribute
> 
> This would require the axis and standard_name attribute to
> be mandatory.
> 
> As with the present standard it would be easy to write
> netcdf files which are both CF-STRICT and COARDS compatible, 
> and to write applications which process CF-STRICT and 
> COARDS netcdf files. Thus CF-STRICT + COARDS = CF-PRESENT.
> 
> Best wishes 
> 
> Juerg
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata