Hi Ros,
I think that CF has to leave it open to users to encode long_name instead of standard_name where the standard_name has not yet been agreed.
In code, I often do hierarchical tests where I look for the "best name" in order of:
1. standard_name
2. long_name
3. name
4. title
5. short_name
6. id
In order to be backward-compatible CF probably needs to cope with old data that hasn't used either 1 or 2.
Ideally it would be good if it was compulsory to include either 1 or 2 (or both of them) and that in all cases units should be provided.
Cheers,
Ag
-----Original Message-----
From: cf-metadata-bounces at cgd.ucar.edu on behalf of Rosalyn Hatcher
Sent: Tue 17/10/2006 11:05 AM
To: cf-metadata at cgd.ucar.edu
Subject: [CF-metadata] Recommendation for use of standard_name attribute
Dear All,
It has been brought to my attention (Thanks Beate) that the CF Checker
only detects that
the standard_name attribute has been omitted if there is also no
long_name specified. (CF 3)
The omission of a standard_name means a consistency check on the units
can't be made
and so can mask potential units errors.
If there were a positive recommendation to use standard_names where
appropriate then the
checker could raise a warning that they have been omitted. The user will
then be aware
that there may be additional errors. I see a lot of effort being put
into expanding/maintaining the
standard_name table, I think the convention should encourage people to
use them.
Hence I'd like to propose an additional recommendation to CF section 3.3
something along
the lines of:
* All variables should use the standard_name attribute to describe their
content. Exceptions
are boundary, climatology and grid_mapping variables.
Regards,
Ros
--
------------------------------------------------------------------------
Rosalyn Hatcher
NCAS-Climate
Dept. of Meteorology, University of Reading,
Earley Gate, Reading. RG6 6BB
Email: r.s.hatcher at reading.ac.uk Tel: +44 (0) 118 378 7841
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu
http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Tue Oct 17 2006 - 04:32:18 BST
