Hi Karl,
This seems reasonable. Since CF doesn't define the use of history as a per
variable attribute then an application should treat it like any other
unrecognized attribute. I'll change the conformance document to reflect
that.
Thanks,
Brian
On Wed, Nov 05, 2003 at 08:28:42AM -0800, Karl Taylor wrote:
> Hi all,
>
> In the conformance and compliance document
> (http://www.cgd.ucar.edu/cms/eaton/cf-metadata/conformance-req.html), it
> states that "the title and history attributes must only be used as
> global attributes." I thought our general philosophy was to allow the
> user to add any attributes he wanted to as long as they didn't conflict
> with those discussed in the CF-conventions. In files I am creating, I
> use history as a global attribute to describe the history that applies
> to all the fields in the file. But I want to use a similar attribute to
> describe in some cases how an individual variable was processed.
>
> For example, if one of the variables in the file is total precipition,
> it would be useful to note that the total precipitation in this file was
> computed as the sum of the large-scale precipitation and convective
> precipitation contained in a different file. I think this is part of
> the history and since it applies only to this variable, I would like to
> have the freedom to attach a history attribute to the variable without
> violating the convention.
>
> I'm not arguing here to make the "history attribute attached to a
> variable" part of the CF convention, but just to allow a user to include
> this attribute if he chooses.
>
> Any objections?
>
> Karl
>
> P.S. This information could also be reasonably included in a "comment"
> attribute attached to the variable, but I am reserving that attribute
> for another purpose.
>
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Fri Nov 14 2003 - 19:49:10 GMT
