Dear All,
Couple of points on this thread.
The use of slashes as separators worries me. They have caused more side effects than anything else in my experience, particularly when then end up in a RESTful service URL. All it takes is a service call addressing the extended attribute for this to happen.
In many other communities the colon is becoming THE established namespace delimiter, such as the OGC standard URN syntax. SeaDataNet have followed this and the syntax has been recommended by MMI. Could the NetCDF side effect of the colon be addressed in some way?
Steve's inclusion in his posting about referring to an attribute by another name could be addressed through the ontology proposal that we discussed in Paris.
Cheers, Roy.
>>> Philip Bentley <philip.bentley at metoffice.gov.uk> 07/11/07 9:11 AM >>>
Hi John
Rather than invent another namespace separator - with potentially
unforeseen side-effects - might an alternative solution be simply to use
XML itself? For example:
char time(record=63, time_len=21);
time:long_name = "forecast date and time";
time:xml_attr = "<dc:title>Field Guide To CF</dc:title>
time:xml_attr = "<dc:creator>J R Luser</dc:creator>
or
char time(record=63, time_len=21);
time:long_name = "forecast date and time";
time:xml_metadata = "<dc:title>Field Guide To
CF</dc:title><dc:creator>J R Luser</dc:creator> ... "
However, I guess the logical end game, using this approach, would be to
put ALL the CF metadata in a single attribute!
char time(record=63, time_len=21);
time:xml_metadata = "<cf:long_name>forecast date and
time</cf:long_name><cf:standard_name>time</cf:standard_name><dc:title>Field Guide To CF</dc:title><dc:creator>J R Luser</dc:creator> ... "
Gulp!!
Regards,
Phil
>
>
> I could see mimicking XML namespaces somewhat, by adding a prefix to attribute names, but what should the seperator charactor be? Unfortunately the ":" wont work, since right now in CDL syntax we have:
>
> char time(record=63, time_len=21);
> time:long_name = "forecast date and time";
>
> suppose you want to add an attribute "title" and indicate its in the "Dublin Core" namespace. Possible examples:
>
> char time(record=63, time_len=21);
> time:long_name = "forecast date and time";
> time:dc_title = "forecast date and time";
>
> char time(record=63, time_len=21);
> time:long_name = "forecast date and time";
> time:dc/title = "forecast date and time";
>
> char time(record=63, time_len=21);
> time:long_name = "forecast date and time";
> time:dc#title = "forecast date and time";
>
> char time(record=63, time_len=21);
> time:long_name = "forecast date and time";
> time:dc at title = "forecast date and time";
>
> etc. At the moment the "/" seems the most natural with less possible conflicts.
>
>
> this would just be a convention, but we could start a "namespace" convention page, to minimize conflicts, just listing them:
>
> Prefix Convention XML namespace Owner
> dc Dublin Core "http://purl.org/dc/elements/1.1/" Dublin Core Initiative
> thredds THREDDS "http://unidata.ucar.edu/thredds/" Unidata
> ...
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.
Received on Wed Jul 11 2007 - 19:34:32 BST
