[CF-metadata] Array Subscript Convention

[Brian Eaton]

All,

I think this is best handled by adding standard names since the issue is
how to describe a quantity.  I propose adding the standard names
index_start_0, and index_start_1.

Brian


On Wed, Jan 14, 2004 at 10:47:02AM +0000, Jonathan Gregory wrote:
> Dear Tom
> 
> > Is there a variable attribute for arrays of indices which indicate 
> > whether the
> > numbering convention goes  1:n  or  0:n-1  ?
> > I am building finite element meshes which require arrays of indices
> > to indicate the nodes contained in each triangular element.  
> > If the same array is used by fortran as by C the indexing will be off by
> > one unit.  
> 
> I think we should have a convention that numbering is always from zero. That
> is what we assume in section 8.2, for instance.
> 
> Best wishes
> 
> Jonathan
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata